USF Libraries

The responses of the west Florida shelf to forcing at tidal, synoptic, seasonal and inter-annual time scales

MISSING IMAGE

Material Information

Title:
The responses of the west Florida shelf to forcing at tidal, synoptic, seasonal and inter-annual time scales
Physical Description:
xxvi, 255 leaves : ill. ; 29 cm.
Language:
English
Creator:
He, Ruoying
Publisher:
University of South Florida
Place of Publication:
Tampa, Florida
Publication Date:

Subjects

Subjects / Keywords:
Ocean circulation -- Florida -- Gulf Coast   ( lcsh )
Continental shelf -- Florida -- Gulf Coast   ( lcsh )
Dissertations, Academic -- Marine Science -- Doctoral -- USF   ( fts )

Notes

General Note:
Includes vita. Thesis (Ph.D.)--University of South Florida, 2002. Includes bibliographical references.

Record Information

Source Institution:
University of South Florida
Holding Location:
Universtity of South Florida
Rights Management:
All applicable rights reserved by the source institution and holding location.
Resource Identifier:
aleph - 029279851
oclc - 51851251
usfldc doi - F51-00026
usfldc handle - f51.26
System ID:
SFS0040036:00001


This item is only available as the following downloads:


Full Text

PAGE 1

Examining Committee: Office of Graduate Studies University of South Florida Tampa, Florida CERTIFICATE OF APPROVAL This is to certify that the dissertation of RUOYINGHE in the graduate degree program of Marine Science was approved on July 1, 2002 for the Doctor of Philosophy degree. Major Professor: Robert H. Weisberg Ph.D. Member: Mark E. Luther, Ph.D. Member : Gary Mitchum, Ph.D Member. Wilton Sturges, Ph.D. Member. John J. Walsh, Ph.D

PAGE 2

THE RESPONSES OF THE WEST FLORIDA SHELF TO FORCING AT TIDAL, SYNOPTIC, SEASONAL AND INTER-ANNUAL TIME SCALES by RUOYINGHE A dissertation submitted in partial fulfillment of the requirements for the degree of Doctor of Philosophy College of Marine Science University of South Florida Date of Approval: July 1, 2002 Major Professor: Robert H. Weisberg, Ph.D.

PAGE 3

Copyright by Ruoying He 2002 All rights reserved

PAGE 4

DEDICATION To my dear family whose love, encouragement and supports are the most valuable assets in my life

PAGE 5

ACKNOWLEDGEMENTS I would like to thank my advisor, Professor Robert Weisberg for his strong scientific insight, continuous encouragement and high standards. I enjoyed every moment of his company and thank him for being the kindest mentor a graduate student could wish for. I gratefully acknowledge my Committee Members Professors Mark Luther Gary Mitchum, Wilton Sturges and John Walsh for their excellent guidance and fresh ideas that have greatly enriched my research experience. I would also like to thank Prof. Edward VanVleet for his commitment and always support during my graduate program. This dissertation would have been impossible to create without the support of many people In particular, I thank the staff and students of Ocean Circulation Group. I thank my family for their consistent support and confidence on me and for their love, understanding and encouragement throughout my study and research. This work is supported by the Office of Naval Research, grant #14-35-00130787 and the National Oceanic and Atmospheric Administration, grant# NA 76RG0463. In addition, supports from USF Marine Science Knight Oceanographic Fellowship, Robert Gerral Fellowship and Paul Getting Fellowship are greatly appreciated.

PAGE 6

I thank Drs. G. Vargo, F. MullerKarger, C. Heil, S. Meyers and L. Zheng of University of South Florida, Drs. G. Mellor, L. Oey and T. Ezer of Princeton University, Dr. J. Pullen ofNRL, Drs. P. Newberg, andJ. Mesias of Oregon State University, Drs. R. Signell and C. Denham of Woods Hole USGS, Prof. R. Reid and Dr. M. Howard of Texas A&M university, Drs. B. Leben of University of Colorado and Dr. R. Gasparovic of John Hopkins university for kindly providing me the research tools, data and helpful discussions.

PAGE 7

TABLE OF CONTENTS v LIST OF FIGURES vi ABSTRACT XXlll CHAPTER 1. INTRODUCTION 1 1.1. Present State of know ledge on the west Florida Shelf Circulation 3 1.2. Motivations and Objectives 5 References 9 CHAPTER 2. TIDES ON THE WEST FLORIDA SHELF 12 2.1. Abstract 12 2 .2. Introduction 13 2.3. Observation 15 2.3.1. Tidal Height 15 2.3.2. Tidal Current 19 2 3.3. Phases of Tidal Height and Tidal Currents 22 2.4. Hydrodynamic Model 24 2.5. Modeled Elevations and Currents 28 2 5.1. General Features 28 2.5.2. Tidal Residual Current and the Lagrangian Transport 40 2.6. Bottom Stress and Turbulence Mixing 41 2.7. Summary 48 References 52

PAGE 8

CHAPTER 3. A LOOP CURRENT INTRUSION CASE STUDY ON THE WEST FLORIDA SHELF 3 .1. Abstract 3.2. Introduction 3.3. Data 3.3.1. Hydrographic Data 3.3.2. Satellite Data 3.3.3. Current Data 3.4. Numerical Simulations 3.5. Summary and Discussion References CHAPTER 4. WEST FLORIDA SHELF CIRCUALATION AND TEMPERATURE BUDGET FOR 1999 SPRING TRANSITION 4.1. Abstract 4.2. Introduction 4.3. Model and Forcing Field 4.3.1. Model 4.3.2. Atmospheric Forcing 4.3.3. Lateral Boundary Forcing 4.4. Model and Data Comparisons 4.4.1. Sea Level 4.4.2. Currents 4.5. Mean Circulation 4.5.1. Flow Fields 4.5.2. Temperature and Salinity Fields 4.5.3. Lagrangian Drifter 4.6. Temperature Budget 4.6.1. The Temperature Equation 4.6.2. Depth-averaged Balance 11 55 55 55 56 57 62 65 72 74 79 81 81 82 85 85 89 90 91 91 91 97 97 101 114 117 117 117

PAGE 9

4.6.3. Vertical profiles of the term by term balance 124 4.6.4. Term by term contribution to the seasonal change in SST 129 4.7. Summary 131 References 134 CHAPTER 5. LOCAL AND DEEP-OCEAN FORCING CONTRBUTIONS TO ANOMALOUS WATER PROPERTIES ON THE WEST FLORIDA CONTINENTAL SHELF 137 5.1. Abstract 137 5.2. Introduction 137 5.3. Observations 141 5.4. Model Simulation with Local Forcing Only 147 5.4.1. Motivation 147 5.4.2. Model and Forcing Fields 150 5.4.3. Results 153 5.5. Model Simulation with Local and Idealized Loop Current Forcing 168 5.5.1. Motivation 168 5.5.2. Model and Forcing Fields 170 5.5.3. Results 171 5.6. Discussion 181 5.7. Summary and Conclusion 188 References 195 CHAPTER 6. WEST FLORIDA SHELF CIRCULATION AND TEMPERATURE BUDGET FOR 1998 FALL TRANSITION 6.1. Abstract 6.2. Introduction 6.3. Model and Forcing Fields 6.3.1. Model 6.3.2. Atmospheric Forcing 6.3.3. Lateral Boundary Forcing iii 197 197 198 200 200 202 202

PAGE 10

6.4. Model and Data Comparison 204 6.4.1. Sea Level 204 6.4.2. Currents 204 6.5. Modeled Seasonal Mean Circulation 210 6 5 .1. Circulation Fields 210 6.5 .2. Lagrangian Trajectories 221 6.6. Scalar Fields and Temperature Budget 223 6.6.1. Temperature and Salinity Field 223 6.6.2. The Temperature Equation 226 6.6.3. Depth-averaged Balance 227 6.6.4. Vertical Profiles of Term-by-term Temperature Balance 229 6 6.5. Across-shelf Transects for Seasonal Mean 232 6.7 Summary and Conclusion 235 References 239 CHAPTER 7. SUMMARY 241 APPENDICES. THREE-DllviENSIONAL TEMPERATURE DIOGNOSIS IN THE PRINCETON OCEAN MODEL ABOUT THE AUTHOR iv 250 End Page

PAGE 11

Table 2.1. Table 2.2. LIST OFT ABLES Comparisons of observed and computed tidal elevation at reference sites. Normalized RMS Tidal residuals showing contribution of non tidal fluctuations to the sea level variability at nine tide gauges spanning the study domain .. Table 2.3a. Greenwich phases of the major tidal constituents at the St. Petersburg tide gauge and at the 12 ADCP stations (left hand columns), plus the relative times between high water and at St. Petersburg and maximum (shoreward directed) semi-major axis tidal currents at 12 ADCP stations (right hand column). Phase differences are converted to time difference using speeds of M2, S2, 01 and K1 of 29.98 deglhr, 30.00 deglhr, 13.94 deglhr and 15.04 deglhr, respectively, where "+" and"-" indicate time lags and leads, respectively. Table 2.3b. The time lag between high tide at St. Petersburg and the maximum (shoreward directed) semi-major axis tidal current at 12 ADCP stations determined by the maximum lag-correlation coefficient between the composites of M2, S2, 01, and K1 constituents. Table 2.4. A comparison between observed and modeled tidal ellipse parameters v 18 19 23 23 39

PAGE 12

LIST OF FIGURES Figure 1.1. The west Florida shelf and its associated local and deepwater forcing. 8 Figure 2.1. Model domain and observational locations. The 9 tide gauges (denoted by stars) are South Pass, Waveland, Pensacola, Panama City, Apalachicola, Cedar Key, St. Petersburg, Naples and Key West. The 12 moorings for velocity data (denoted by triangles) are: ASl (47m), TSl (31m), TS2 (47m), TS3 (46m), TS4 (63m), TS5 (142m), TS6 (296m), EC2 (50m), EC3 (30m), EC4 (20m), EC5 (10m), EC6 (10m). 16 Figure 2.2. Vertical profiles of the velocity hodograph ellipses semi-major axis amplitudes and orientation (measured counterclockwise from east) for EC designated moorings. The thick solid and dashed lines denote M2 and S2 respectively. The thin solid and dashed lines denote 01 and respectively. 20 Figure 2.3. Vertical profiles of the velocity hodograph ellipses semi-major axis amplitude and orientation angles (measured counterclockwise from east) for TS designated moorings. The thick solid and dashed lines denote M2 and S2 respectively. The thin solid and dashed lines denote 01 and respectively. 21 vi

PAGE 13

Figure 2.4. The model grid used for the regional tidal simulation 26 Figure 2.5. The distributions of tidal amplitude (upper panel) and phases (low panel) along the model open boundary for M2 S2 K1 and 01 constituents (see legend provided). The abscissa is the grid index that includes both points over land and water. The land points are marked so only the points over water are shown. 29 Figure 2.6. Comparisons between modeled and observed amplitude (left panels) and Greenwich phases (right panels) for M2 S2 K1 and 01 constituents at the nine coastal tide gauge locations. Crosses (circles) denote modeled (observed) values. Bulls eyes are where they overlay. 31 Figure 2.7. Modeled co-amplitude and co-phase (relative to Greenwich meridian) distributions for M2 S2 01 and K1 Solid (dashed) lines denote amplitudes (phases), and the contour interval for amplitude and phase are indicated in the lower left comer of each panels. 33 Figure 2.8. The amplitude ratio between the principle semi-diurnal and diurnal tides. 36 Figure 2.9. A comparison between the observed (left panels) and modeled (right panels) M2 S2 01 and K1 tidal current hodograph ellipses at 12 moored ADCP locations. 38 Figure 2.10. The distribution over the entire model domain of the modeled M2 S2 01 and K1 tidal current hodograph ellipses. 42 Figure 2.11. Eulerian (upper panel) and Lagrangian (low panel) Vll

PAGE 14

representation of the tidal residual surface circulation. Figure 2.12. The spatial distribution of mixing properties inferred from the M2 tidal currents. From the top to bottom are the friction velocity, the resistance coefficient rand the log layer thickness. Units are provided in each panel, as are bathymetric contours (light lines) for 20m, 50m, 1OOm, 200m, 1 OOOm and 2000m isobaths. Figure 3.1. West Florida shelf geometry, locations of the hydrographic casts (denoted by dots), and moored ADCP measurements (triangles denote upward looking and circles denote downward 43 46 looking ADCPs). 58 Figure 3.2. Figure 3.3. Figure 3.4. Figure 3.5. Across-shelf (Sarasota transect) distributions of temperature, salinity, density and Chlorophyll fluorescence sampled on June 6th (the left panels) and June 28th (the right panels). Calculated across-shelf (Sarasota transect) geostrophic currents (assuming zero at bottom) along the Sarasota transect on June 6th (the upper panel) and June 28th (the lower panel). The contour interval is 5 em s-1 and southward currents are denoted by solid lines. 60 61 June 6th 2000 A VHRR SST obtained (with permission) from the Applied Physics Lab at J. Hopkins University. The 75 m and 200m isobath contours are overlaid to show the LC front penetration onto the WFS. 63 Time series of A VHRR SST maps from June 1st to July 6th 2000 obtained (with permission) from the Applied Physics Lab at J. Hopkins University. The 75 m and 200m isobath contours viii

PAGE 15

are overlaid to show the LC front penetration onto the WFS. 64 Figure 3.6. Time series of Topex/Poseidon and ERS sea surface height maps from June 1st to July 6th 2000 obtained (with permission) from the Colorado Center for Astrodynamic Research. The 75 m and 200 m isobath contours are overlaid to show the LC front penetration onto the WFS. 66 Figure 3.7. Time series of near-surface currents observed on the WFS from June 1st to July 6th 2000. Thick dark arrows denote currents at the 150m isobath; thin dark arrows denote currents at all other isobaths; and gray arrows denote the coastal winds at Venice, FL. The depth contours correspond to the 20m, 50 m, 75 m and 150m isobaths. 68 Figure 3.8. Time series of near-bottom currents observed on the WFS from June 1st to July 6th 2000. Thick dark arrows denote currents at the 150m isobath; thin dark arrows denote currents at all other isobaths; and gray arrows denote the coastal winds at Venice, FL. The depth contours correspond to the 20m, 50 m, 75 m and 150m isobaths. 69 Figure 3.9. The June 28th depth profiles of the ADCP observed currents (thick light lines), the thermal wind calculated baroclinic geostrophic current (thick dark lines), and their difference (thin dark lines) at the 150 m (left hand panel) and the 75 m (right hand panel) isobaths. 71 Figure 3.10. Sea surface height (solid line) and associated geostrophic current (dash line) distributions along the WFS shelf circulation model open boundary. 75 ix

PAGE 16

Figure 3.11 Depth-averaged current maps from the idealized LC and wind forced model experiments. Panel A is for the LC only case. Panel B is for the LC plus upwelling favorable (0.1 N m"2 ) wind case. Panel C is for the LC plus downwelling favorable (0.1 N m"2 ) winds case. Thick lines along the open boundary indicate the sea surface height perturbation region. The depth contours correspond to the 20 m, 50 m, 100 m, 200 m, 1000 m, and 2000 m isobath. Figure 4.1. Figure 4.2. The regional model grid (upper panel) and bathymetry and station locations (lower panel). Sea level comparisons are with Florida tide gauges at Pensacola, Apalachicola, St. Petersburg, and Naples. Velocity comparisons are with acoustic Doppler current profiles from instruments moored at the 50m, 30m, 25m, 20m, and lOrn isobaths (1-6). Temperature is described along transects I-IV, and the temperature budget is diagnosed at Stations A, B, C, and D. Comparisons between modeled (bold) and observed (thin) sea level at Pensacola, Apalachicola, St. Petersburg, and Naples as quantified by a squared correlation coefficient, along with the 76 88 NCEP wind velocity sampled at station A. 92 Figure 4.3. Comparisons between modeled and observed currents at the 50m isobath (mooring CM2) sampled at depths of 3m, 25m and 40m, along with the NCEP wind velocity sampled at station A. Each vector current time series is accompanied by its seasonal mean east and north velocity components (left hand couplet), and each model/data comparison is quantified by its squared complex correlation coefficient, phase angle (or X

PAGE 17

Figure 4.4. Figure 4.5. Figure 4.6. Figure 4.7. Figure 4.8. Figure 4.9. angular deviation of the model vector from the data vector measured counterclockwise), and regression coefficient (right hand triplet). 94 Comparisons between modeled and observed currents at the 25m isobath (mooring NA2) sampled at depths of 3m, 12m and 20m, along with the NCEP wind velocity sampled at station A. Quantitative comparisons are as in Figure 4.3. Comparisons between modeled and observed currents at the lOrn isobath (mooring EC5) sampled at depths of 2m, 5m and 8m, along with the NCEP wind velocity sampled at station A. Quantitative comparisons are as in Figure 4.3. Comparisons between modeled (bold) and observed (thin) seasonal mean velocity vectors and hodograph ellipses at mid depth for all six mooring locations on the WFS between the 50m and lOrn isobaths. Modeled seasonal mean velocity vectors for the depth averaged and near-surface, mid-water column, and near-bottom sigma levels, k=2, 10, and 20, respectively. Modeled seasonal mean vertical velocity component fields (converted to the z-plane) sampled at the near-surface, mid water column, and near-bottom sigma layers, k=6, 12, and 16, respectively. Bold lines indicate upwelling. Thin lines indicate downwelling. Sea surface temperature and sea surface salinity fields at the end of spring 1999 model simulation (May 31, 1999). xi 95 96 98 99 102 103

PAGE 18

Figure 4.10. Modeled seasonal mean velocity vectors for the depth averaged and near-surface, mid-water column, and near-bottom sigma levels, k=2, 10, and 20, respectively, from the model twin experiment forced with wind stress only. Figure 4.11. Sea surface temperature at the end of spring 1999 simulation (May 31, 1999) for the model twin experiment forced by wind stress only. Figure 4.12. Evolution of the monthly mean, depth averaged velocity vectors for March, April, and May relative to the spring 1999 seasonal mean. Figure 4 .13. Modeled temperature sections sampled on March 15 across transects originating at DeSoto Canyon, Cape San Bias, Florida Big Bend, and Sarasota The contour interval is 1 C. Figure 4.14 Same as Figure 4.13 except sampled on April 15. Figure 4.15 Same as Figure 4.13 e x cept sampled on May 15. Figure 4.16. Modeled near-surface Lagrangian drifter trajectories for the period April1, 1999 to May 31, 1999. Drifters were started along nine different transects with initial positions given by solid dots. Figure 4.17. Same as Figure 4.16 except for near-bottom Lagrangian drifter trajectories. xii 106 107 108 110 111 112 115 116

PAGE 19

Figure 4.18a The relative contributions to the depth-averaged temperature balance by ocean circulation and diffusion at station A. Three time series are shown: the advection, the diffusion, and their sum (which equals the local rate of change of depth-averaged temperatuere ). Accompanying each time series are their seasonal means and standard deviations in units of C day-1 as measures of the seasonal and synoptic scale variability. 119 Figure 4.18b Same as Figure 4.18a except for station B. 120 Figure 4.18c Same as Figure 4.18a except for station C. 121 Figure 4.18d Same as Figure 4.18a except for station D. 122 Figure 4.19. Time series of the depth profiles of the individual terms that comprise the temperature balance at station A. The left hand panels show the horizontal and vertical components of the ocean advection and their sum, and the right hand panels show the diffusion, the diffusion plus the advection, and the temperature. To the right of each panel is the seasonal mean profile. The contour interval for each of the budget terms is 0.05C day-1 and the contour interval for temperature is 0.5C. Shading indicates warming and clear indicates cooling. Figure 4.20. Same as Figure 4.19 except for station B. Figure 4.21. Same as Figure 4.19 except for station C. Figure 4.22. Same as Figure 4.19 except for station D. xiii 125 126 127 128

PAGE 20

Figure 4.23. The contributions made to the seasonal mean rate of change of SST by each of the advection and diffusion terms. The left panels show the advection terms and their sum, and the right panels show the horizontal and vertical diffusion terms, the sum of all terms. The contour interval is 0.02C day1 Bold lines indicate warming and thin lines indicate cooling. Figure 5.1. The West Florida Shelf and its associated local and offshore forcing influences. Figure 5.2. The locations of the hydrographic and in-situ current measurements. Circles denote the Texas A&M NEGOM ship survey stations The thin lines denote the USF ECOHAB hydrographic transects TA (Tampa) and SA (Sarasota). The thick line denotes the Mote Marine Lab hydrographic transect. The double cross denotes the NOAA meteorological buoy 42036. Stars denote the moored ADCP sites PC, AS 1/CM2, and LB3/EC4 at the 30 m, 50 m, and 20 m isobaths. Figure 5.3a Time series of NOAA buoy winds, AS 1/CM2 mid-depth currents (50m isobath), and LB3/EC4 mid-depth currents (20m isobath) from January to September 1998, sampled daily after low-pass filtering to remove oscillations at time scales shorter than 36 hrs. Figure 5.3b Time series of NOAA buoy winds, AS1/CM2 mid-depth currents (50m isobath), and LB3/EC4 mid-depth currents (20m isobath) from January to September 1999, sampled daily after low-pass filtering to remove oscillations at time scales shorter than 36 hrs xiv 130 138 142 143 144

PAGE 21

Figure 5.4 Figure 5.5 Figure 5.6 Figure 5.7 Figure 5.8 Figure 5.9 Time series of NOAA buoy winds and P A mid-depth currents (30m isobath) from March to August 1998, sampled daily after low-pass filtering to remove oscillations at time scales shorter than 36 hrs. Temperature sections from the Mote Marine Lab (March 30th and May 18th) and the USF [June lOth (SA); July 7th (SA), August 5th (SA), September 9th (TA), and November 9th (SA)] hydrographic transects. Spring season averaged wind fields for 1998, 1999, and climatology. West Florida Shelf model domain. Modeled currents, sea surface temperature and sea surface salinity fields under local forcing only, compared for the spring seasons of 1998 (left panels) and 1999 (right panels). From top to bottom in both panels are the seasonal and depth averaged currents, and the modeled sea surface temperature and sea surface salinity fields sampled on May 31st. Comparisons between the observed and the modeled currents at the 50m isobath (AS l/CM2) sampled near-surface, mid-depth, and near-bottom, along with the NOAA buoy 42036 winds for the case of local forcing only. The vector time series are sampled daily after low-pass filtering to remove oscillations at time scales shorter than 36 hrs. Figure 5.10 Comparisons between the observed and the modeled currents at the 20m isobath (LB3/EC4) sampled near-surface, mid-depth, XV 146 148 149 152 158 156

PAGE 22

and near-bottom, along with the NOAA buoy 42036 winds for the case of local forcing only. The vector time series are sampled daily after low-pass filtering to remove oscillations at time scales shorter than 36 hrs. Figure 5.11 Comparisons between the observed and the modeled currents at the 30m isobath (PC) sampled near-surface, mid-depth, and near-bottom, along with the NOAA buoy 42036 winds for the case of local forcing only The vector time series are sampled daily after low-pass filtering to remove oscillations at time scales shorter than 36 hrs. Figure 5 12 Comparison between the NEGOM measured (May 6-15) and the modeled (sampled on May 10) bottom temperature fields for the case of local forcing only. Figure 5.13 Comparison between the NEGOM measured (May 6-15) and the modeled (sampled on May 10) subsurface velocity fields at 14m depth for the case of local forcing only. Figure 5.14 Comparison between the NEGOM measured (May 6-15) and the modeled (sampled on May 10) subsurface velocity fields at 50 m depth for the case of local forcing only. Figure 5.15a Across-shelf transects of the modeled temperature, across-shelf (u) and along-shelf (v) velocity components sampl e d from a movi e animation at D e Soto Canyon on March lOth, 12th, 14th, 157 158 160 161 162 and 19th for the case of local forcing only. 164 XVI

PAGE 23

Figure 5.15b Across-shelf transects of the modeled temperature, across-shelf (u) and along-shelf (v) velocity components sampled from a movie animation at the Big Bend on March lOth, 12th, 14th, and 19th for the case of local forcing only. Figure 5.16 Modeled temperatures along the Sarasota transect sampled on March 30th and May 18th for the case of local forcing only. Figure 5.17 TOPEX/POSEIDON sea surface height anomalies sampled along track 26 from November 1992 to November 2001. Shading denotes positive anomalies. Figure 5.18 Modeled currents, sea surface temperature, and sea surface salinity fields for the spring season of 1998. The left hand panels are for the case of local forcing only; the right panels are for the case of local plus Loop Current forcing. From top to bottom in both panels are the seasonal and depth averaged currents, and the modeled sea surface temperature and sea surface salinity fields sampled on May 31st. Figure 5.19 Comparisons between the observed and the modeled currents at the 50m isobath (AS 1/CM2) sampled near-surface, mid-depth, and near-bottom, along with the NOAA buoy 42036 winds for the case of local plus Loop Current forcing. The vector time series are sampled daily after low-pass filtering to remove oscillations at time scales shorter than 36 hrs. Figure 5.20 Comparisons between the observed and the modeled currents at the 20m isobath (LB3/EC4) sampled near-surface, mid-depth, and near-bottom, along with the NOAA buoy 42036 winds for xvii 165 167 169 172 173

PAGE 24

the case of local plus Loop Current forcing. The vector time series are sampled daily after low-pass filtering to remove oscillations at time scales shorter than 36 hrs. Figure 5.21 Comparisons between the observed and the modeled currents at the 30m isobath (PC) sampled near-surface, mid-depth, and near-bottom, along with the NOAA buoy 42036 winds for the case of local plus Loop Current forcing. The vector time series are sampled daily after low-pass filtering to remove oscillations at time scales shorter than 36 hrs. Figure 5.22 Modeled temperatures along the Big Bend and Sarasota sections sampled on May 15th for the case of local plus Loop Current forcing. Figure 5.23 Modeled bottom velocity and temperature fields sampled on May 15th for the case of local plus Loop Current forcing. Figure 5.24 Modeled temperature sampled at the Sarasota transect on Aug. 174 176 178 179 5th for the case of local plus Loop Current forcing. 180 Figure 5.25 Temperature/Salinity, Temperature/ Phosphate, Temperature/Silicate, and Temperature/(Nitrate+Nitrite) relationships based on WOCE hydrographic section A22 data sampled between the latitude range of 11 N to 25N. Figure 5.26 Progressive Vector Diagram (PVD) comparisons between observed (thin lines) and modeled (thick lines) currents at the 50 m isobath (AS 1/CM2) for the six month period March to August 1998. The left hand panels are for the case of local xviii 184

PAGE 25

forcing only; the right hand panels are for the case of local plus Loop Current forcing. In each case from top to bottom are the near-surface, mid-depth, and near bottom PVDs, respectively. 186 Figure 5.27 Progressive Vector Diagram (PVD) comparisons between observed (thin lines) and modeled (thick lines) currents at the 20m isobath (LB3/EC4) for the six month period March to August 1998, irrespective of the data gap (Figures 10 or 20). The left hand panels are for the case of local forcing only; the right hand panels are for the case of local plus Loop Current forcing. In each case from top to bottom are the near-surface, mid-depth, and near bottom PVDs, respectively. 187 Figure 5.28 Modeled trajectories for neutrally buoyant Lagrangian drifters released near the bottom along the DeSoto Canyon, Cape San Bias, Big Bend, and Sarasota transects on either March 151 (left hand panels) or June 151 (right hand panels) and tracked for the respective three-month spring or summer seasons of 1998, under the influence of both local and Loop Current forcing Figure 6.1. The regional model grid (upper panels) and bathymetry and station locations (lower panels). Sea level comparisons are with Florida tide gauges at Pensacola, Apalachicola, St. Petersburg, and Naples. Velocity comparisons are with acoustic Doppler current profiles from instruments moored at 25, 20 and lOrn isobath (1-3). Temperature budget is diagnosed at station A, B. Seasonal mean temperature budget is diagnosed along a transect off Sarasota, FL. XIX 189 201

PAGE 26

Figure 6.2. Comparisons between modeled (bold) and observed (thin) sea level at Pensacola, Apalachicola, St. Petersburg, and Naples as quantified by squared correlation long with NCEP wind velocity sampled at station A. 205 Figure 6.3. Comparisons between modeled and observed currents at the 25 m isobath (mooring NA2) sampled at surface, mid-depth and near the bottom, along with the NCEP wind sampled at Station A. Each vector current time series is accompanied by its seasonal mean east and north velocity components (left-hand couplet), and each model/data comparison is quantified by its squared complex correlation coefficient, phase angle (or angular deviation of the model vectors from the data vector measured counterclockwise), and regression coefficient (righthand triplet) 207 Figure 6.4. Comparisons between modeled (Case I) and observed currents at the 20 m isobath (mooring EC4) sampled at surface, middepth and near bottom along with the NCEP wind velocity sampled at Station A. Quantitative comparisons are as in Figure 6.3. 208 Figure 6.5. Comparisons between modeled (Case I) and observed currents at the 10m isobath (mooring EC5) sampled at surface, middepth and near bottom along with the NCEP wind velocity sampled at Station A. Quantitative comparisons are as in Figure 6.3. 209 Figure 6.6. Comparisons between modeled currents of Case I and Case II at the 25 m isobath (mooring NA2) sampled at surface, middepth and near bottom along with the NCEP wind velocity XX

PAGE 27

sampled at Station A. Figure 6.7. Comparisons between modeled (bold) and observed (thin) seasonal mean velocity vectors and hodograph ellipses at middepth for all three mooring locations on the WFS between 25 to 10 m isobath. Upper (lower) panel is for Case I (Case II). Figure 6.8. Modeled (Case I) seasonal mean velocity vectors for the depth averaged and near surface, mid-depth, and near bottom sigma levels, k=2, 10, 20, respectively. Figure 6.9. Modeled (Case II) seasonal mean velocity vectors and their variance ellipses for near surface, mid-depth, and near bottom sigma levels, k=2, 10, 20, respectively. Figure 6.10. Monthly(September, October and November) and seasonal mean surface wind fields. Figure 6.11. Evolution of the monthly mean, depth averaged velocity vectors for September, October and November along with the fall 1998 seasonal mean for Case I. Figure 6.12. Evolution of the monthly mean, depth averaged velocity vectors for September, October and November along with the 211 212 213 214 216 219 fall 1998 seasonal mean for Case II. 220 Figure 6.13. Modeled Lagrangian trajectories tracked over three month of the drifters released along 25m, 50m, and lOOm isobath near the bottom. Left and right panels are for Case I and Case II, respectively. XXI 222

PAGE 28

Figure 6.14. Modeled (Case I) sea surface temperature and sea surface salinity fields at the end of fall 1998 model simulation (November 30, 1998) 224 Figure 6.15. Modeled (Case II) sea surface temperature and sea surface salinity fields at the end of fall 1998 model simulation (November 30, 1998) 225 Figure 6.16. (a) The relative contributions to the depth-averaged temperature balance by ocean circulation and diffusion at station A. Three time series are shown: the advection, the diffusion and their sum (which is nearly exactly equals the local change of depth-average temperature). Accompanying each time series are their seasonal means and standard deviations in units of C dai1 as measures of the seasonal and synoptic scale variability. (b) Same as (a), but for station B. 228 Figure 6.17. (a) Time series of the depth profiles of the individual terms that compose the temperature balance at Station A. The left-hand panels show the horizontal and vertical components of the ocean advection and their sum, and the right-hand panels show the diffusion, the diffusion plus the advection, and the temperature. To the right of each panel is the seasonal mean profile. The contour interval for each of the budget term is 0.1 C dai1 and the contour interval for temperature is 1 C. Shading indicates wanning and clear indicants cooling. (b) Same as (a), but for station B. 231 Figure 6.18. Across-shelf transect (off Sarasota, FL) profiles of seasonal means of ocean advection, diffusion terms and their sums for Fall 98 Case I (upper panels), Fall98 Case II (middle panels) xxii

PAGE 29

and Spring 99 (lower panels). The contour interval is indicated in each panel. The shading and clear contour denote warming and cooling effects, respectively. xxiii 233

PAGE 30

THE RESPONSES OF THE WEST FLORIDA SHELF TO FORCING AT TIDAL, SYNOPTIC, SEASONAL AND INTER-ANNUAL TJME SCALES by RUOYINGHE An Abstract of a dissertation submitted in partial fulfillment of the requirements for the degree of Doctor of Philosophy College of Marine Science University of South Florida Date of Approval: July 1, 2001 Major Professor : Robert H. Weisberg, Ph.D.

PAGE 31

Observations and numerical models are used to study the responses of the west Florida shelf to forcing at tidal, synoptic, seasonal and inter-annual time scales. The observations include coastal sea level and moored current meter data, and hydrographic and current data taken from ship surveys. The model is a regional adaptation of three dimensional, primitive equations, Princeton Ocean Model (POM) that uses an orthogonal curvilinear coordinate system in the horizontal and a sigma coordinate system in the vertical. The west Florida shelf tides, especially the semi-diurnal constituents, are found to be primarily barotropic. Tides on the shelf are generally weak and minor contributors to the shelf water mixing and material property transports when compared with water motions occurring at synoptic and seasonal time scales. The west Florida shelf circulations at synoptic and seasonal scales respond primarily to the shelf wide local forcing by winds, surface heat fluxes, and river runoffs. Shelf wide winds and surface heat fluxes are most important and they act together to modulate the shelf circulation and produce some unique, yearly-occurring physical and biological shelf water features (e.g., the spring "cold tongue" and the "green river"). Interactions between the west Florida shelf water and the deep ocean Loop Current are constrained by the bottom topography. There are two types of interactions between the Loop Current and the shelf: (1) Loop Current impacts in the north that have relatively small effect on the shelf circulation and (2) Loop Current impacts at the southwest corner (west of the Florida Keys) that can set the shelf water in motion. Comparisons of the shelf circulations in 1998 and 1999 indicate that the inter annual variability between those two years is due to anomalies in both shelf wide local XXV

PAGE 32

forcing and interactions with the Loop Current. In either case, the bottom Ekman layer provides the conduit for transporting cold, nutrient rich deep water onshore, thereby contributing to near-shore ecological variations. Temperature budget analyses are performed for both the spring and fall seasonal transitions. Shelf water temperature balances in both seasonal transitions are found to be fully three-dimensional. Counter-intuitively, upwelling in the fall can provide a warming effect, thereby lengthening the fall transition. Abstract Approved:-----------------Major Professor: Robert H. Weisberg, Ph.D. Professor, College of Marine Science Date Approved: SJ z.A1) '1.. XXVI

PAGE 33

CHAPTER 1 INTRODUCTION I seem to have been only like a boy playing on the seashore, and diverting myself in now and then finding a smoother pebble or a prettier shell than ordinary, whilst the great ocean of truth lay all undiscovered before me." -Isaac Newton The continental shelf occupies only about 10% of the global ocean area and even a much smaller portion of its volume, but it dominates the economic importance of the sea. Fluxes of sediments, biologically active materials, and sometimes toxic compounds from the land to the continental shelf are rapidly increasing. If we are to comprehend the impact of these fluxes to the continental shelf environment and ecosystem, we must conquer the complexity of coastal circulation and its transport processes. Coastal circulation and its transport processes are determined by the physical forcing of both local and deep ocean origins. One of the most important questions on coastal circulation and its transport processes then concerns the relative importance of local and deep-ocean forcing. We define local forcing as the shelf-wide inputs of momentum (wind) and buoyancy (surface heat and the riverine fresh water). Similarly, deep-ocean forcing is defined as momentum and buoyancy input seaward of, or at, the shelf break. The west Florida shelf (WFS) is chosen to study coastal circulation and continental shelf processes as it is, in many ways, an excellent natural laboratory (Figure 1.1). It is influenced (remotely) from the deep-ocean by the momentum and buoyancy 1

PAGE 34

2 inputs of the Gulf of Mexico Loop Current that enters through the Yucatan Channel and exits through the Rorida Strait. When the highly baroclinic Loop Current and its eddies impinge on the WFS, they can set up a potential for driving material fluxes onto the shelf. Throughout the WFS, local forcing includes the momentum input of the winds, and the buoyancy inputs of both the surface heat flux and the river runoffs. Driven both locally and remotely, the WFS circulation and transport processes are further complicated by variable and complex shelf topography. Partially closed by the Rorida Key at its southern end, the WFS moving northward changes from being wide and gently sloping to being narrow and steeply sloping. The shelf width reduces to a minimum west of Cape San Bias at the Desoto Canyon. It then widens again in the Mississippi Bight. Understanding the circulation and transport processes on the WFS is also important for meeting the increasing need for environmental protection. With ever growing population and increasing industrial development along the coastal zone, the near shore waters of the WFS are subjected to increasing environmental stresses from numerous sources. Discharges of municipal and industrial wastes, together with potential spills of toxic substances from coastal commerce, contribute to the water quality problem. Naturally occurring, episodic, seasonal blooms of toxic dinoflagellates (red tides) thought to originate at mid-shelf (Steidinger, 1983; Vargo et al., 1987) cause huge disruption to both commercial and recreational fisheries almost every year. The understanding of these topics relies on the understanding of the coastal circulation and its transport processes. This chapter will first review the present state of knowledge on the WFS shelf circulation and then give the motivation and objectives of this dissertation work.

PAGE 35

1.1. PRESENT STATE OF KNOWLEDGE ON THE WEST FLORIDA SHELF CIRCULATION A starting point for the recent studies of the WFS circulation is the Lagrangian drifter measurements for biological purposes reported on by Chew (1955), Tolbert and Salsman (1964), and Williams (1977). Multipleyears of drifter tracks suggest both a seasonality to the surface currents and some large-scale eddy systems. Direct current measurements on the WFS using in-situ moorings began during the period 1973-1975. Observational studies thereafter (e.g., Niller, 1976; Price, 1976; Mitchum and Sturges, 1982; Cragg et al., 1983; Marmorino, 1983; Weatherly and Thistle, 1997) show that the shelf circulation variability at synoptic scales is highly coherent with the local wind forcing, especially with the passage of the synoptic scale weather systems. In addition, observed velocity vector turning with depth provides evidence of frictional boundary layer formation. (e.g., Weatherly and Martin, 1978). More recently, Weisberg et al. (1996) hypothesize that the seasonally varying shelf circulation is controlled in part by a baroclinic structure imposed by seasonal heating and cooling. 3 Numerical model studies of the WFS circulation (e.g., Marmorino, 1982; Hsueh et al., 1982;Cooper, 1987;Yang and Weisberg, 1999; Yang et al., 1999; Li and Weisberg, 1999a, b; Weisberg et al., 2000; Weisberg, et al., 2001), in addition to confirming the long-term findings of the previous observational studies, provide valuable information on how the circulation responds to idealized forcing functions (e.g., climatological or spatially uniform wind stresses). However, most of these modeling studies suffer from incomplete and I or unrealistic forcing functions, which prohibit quantitative comparisons

PAGE 36

between model results and data. The Weisberg et al. (2000, 2001) studies begin to break away from these limitations by using more realistic initial and forcing fields. 4 Both previous observational and numerical studies have identified the shelf-wide local wind stress to be an important factor in driving the WFS circulation. Other effects of other factors including tides, the Loop Current and local buoyancy, however, are not as well understood. The tides, predominantly of mixed semi-diurnal and diurnal species over the WFS, have typical sea level amplitudes ranging from about 0.1 m at the shelf break to about 0.4 m along the coast (Koblinsky, 1981). Along with these sea level variations, the tidal currents through advection and bottom stress may also contribute to the WFS circulation. Previous studies on the Loop Current were generally focused on the activities of the current and its eddies in deep water (e.g., Behringer et al., 1977; Paluszkiewicz. et al., 1983; Vukovich, 1988; Sturges, 1994; Sturges and Leben, 2000). They show that the extent to which the Loop Current penetrates into the Gulf of Mexico and the frequency at which it sheds a large anticyclonic eddy are difficult to predict. The Loop Current northward intrusion events may cause circulation and sea level fluctuations over the WFS by altering the hydrography and imposing a momentum flux in the shelf break region (Huh et al., 1978). However, there is scant evidence showing any persistent forcing by the Loop Current within the middle to inner shelf. For instance, Marmorino (1982, 1983a) found that the coastal sea level fluctuations are not related to the Loop Current. Hetland et al. (1999), on the basis of analytical and numerical model studies, plus TOPEX-Poseidon altimeter and drifter data, suggest that the impingement of the Loop

PAGE 37

Current on the southernmost comer of the WFS can generate a broad southward current along the WFS north of the impact site. Direct current observations of one-year duration, however, are not conclusive on this hypothesis (Meyers et al., 2001) 5 Local buoyancy impacts including surface heat flux, evaporation minus precipitation, and coastal freshwater fluxes, also contribute to the shelf circulation and its hydrographic variability. Net surface heat flux can affect water stratification through heating and cooling. This effect, being depth-dependent, is more evident in shallow regions. Similarly, coastal freshwater fluxes contribute to the shelf circulation and hydrography. During the spring and summer seasons, the discharges of the Mississippi River even affects the hydrographic structure along the shelf break to the middle shelf (e.g., Dowgiallo, 1994). 1.2. MOTIVATION AND OBJECTIVES Many issues regarding the WFS circulation therefore remain in question. First, the study of tides on the WFS has been limited to a few sea level and current measurements. The spatial distribution of tides and their contributions to the shelf water mixing and material properties transport remains unresolved. Second, while the shelf circulation is generally coherent with wind (momentum) forcing, the relative importance between momentum and buoyancy fluxes in driving the shelf circulation at synoptic and seasonal scales is unclear. Third, seasonal coverage over the shelf has been lacking in past studies. While the previous limited current measurements hint at a seasonally varying circulation pattern (Weisberg et al., 1996), there are no quantitative measures and descriptions of this. Such seasonality is important biologically, as it may have consequences for shelf

PAGE 38

nutrient fluxes and plankton dynamics. Fourth, in driving the shelf circulation, what are the roles of the local momentum and buoyancy fluxes relative to momentum and buoyancy fluxes of offshore origin? Understanding of this is of practical importance for understanding all of the biological, chemical, and geological properties on the WFS. Likewise, there is a poor understanding of how and where water enters and exits the 6 shelf, i.e., how the shelf communicates with other parts of shelf and with the deep Gulf of Mexico. Given the above facts, the goals of this dissertation are to: (1) describe the west Florida shelf circulation at tidal, synoptic, seasonal, and inter-annual time scales; (2) for forcing occurring locally on the shelf, determine the relative contributions of the momentum and buoyancy fluxes; (3) determine the relative importance between such local forcing and forcing of deep-ocean origin, and (4) describe the along-shelf and across-shelf material transports associated with the circulation and the related biological (primary productivity), chemical (nutrient distribution), and geological (sediment resuspension) processes. Both observational and numerical approaches are taken to study these questions. The observations include coastal sea level data, in-situ moored current measurements using acoustic Doppler current profilers (ADCP), and current and hydrographic data taken from ship surveys. The model is a regional adaptation of the three dimensional, primitive equation, sigma coordinate Princeton Ocean Model (POM) which uses an orthogonal curvilinear coordinate system in the horizontal and a sigma coordinate system in the vertical. The model domain extends from the Mississippi River in the northwest to the Florida Keys in the southeast, with one open boundary that arcs

PAGE 39

7 between these two locations. The model horizontal resolution varies from less than 2 km near the coast to about 6 km near the open boundary. Results and conclusions of this work are compiled into five research papers (He and Weisberg, 2002a; He and Weisberg, 2002b; He and Weisberg, 2002c; Weisberg and He, 2002; He and Weisberg, 2002d), which are presented in Chapters 2, 3, 4, 5, and 6, respectively, in this dissertation. Chapter 2, focusing on the barotropic tides, considers the WFS motion at tidal time scale and the contribution of tides to the turbulent mixing and material property transports. Chapter 3 presents a case study of a Loop Current intrusion on the WFS in June 2000, where in-situ data and idealized numerical simulations are used to depict and determine the relative importance of the local and deep water forcing in affecting the shelf circulation during that Loop Current intrusion event. Chapter 4 describes the WFS response to local forcing only and the associated seasonal circulation and shelf water temperature budget in spring transition 1999, a year when the effects of the deep water Loop Current, as evidenced by the shelf hydrography, is minimal. Chapter 5, by comparing two spring seasons of 1998 and 1999, presents a study on the anomalous stratification and coastal upwelling observed in spring to summer of 1998. Inter-annual variability in both local and deep water forcing and their relative importance are identified and these account for the anomalous circulation and water properties observed during that period of time. In contrast to the spring transition, the fall transition happens when surface heat fluxes change from warming to cooling and the water changes from stratified to vertically well-mixed. How the shelf circulation and temperature budget vary in the fall transition 1998 are discussed in Chapter 6, where differences in shelf

PAGE 40

8 circulation for cases with and without the Loop Current are also described with twin experiments. Finally, Chapter 7 provides a summary of this dissertation. Figure 1.1. The west Florida shelf and its associated local and deep-ocean forcing.

PAGE 41

REFERENCES Behringer, D.W., R.L. Molinari, and J.F. Festa (1977), The variability of anticyclone current patterns in the Gulf of Mexico, J. Geophys. Res, 82, 5479-5488 9 Chew, F. (1955), The summer circulation of the Florida west coast offshore water as deduced from the pattern of thermocline depths and a non-geostrophic equation of motion. Univ. Miami Mar. Lab., Technical Report, 55-12, 6 pp Cooper, C.( 1987), A numerical modeling study of low-frequency circulation on the west Florida shelf, Coastal Engineering, 11, 29-56 Cragg, J., G. Mitchum, and W. Sturges (1983), Wind-induced sea-surface slopes on the West Florida shelf, J. Phys. Oceanogr., 13, 2201-2212 Dowgiallo, M.J., ed.(1994), Coastal oceanographic effects of the summer 1993 Mississippi River flooding, Special NOAA Report, 77pp. He, R. and R. H. Weisberg (2002a), Tides on the West Florida Shelf, J of Phys. Oceanogr., in press, 2002a He, R. and R. H. Weisberg (2002b), A Loop Current intrusion case study on the West Florida shelf. J. of Phys. Oceanogr., revised and resubmitted He, R. and R. H. Weisberg (2002c), West Florida Shelf circulation and temperature budget for the 1999 spring transition. Cont. Shelf Research, 22, 5, pp 719-748 He, R. and R. H. Weisberg (2002d), West Florida shelf circulation and temperature budget for the 1998 fall transition, Cont. Shelf Res. submitted Hetland, R. Ya Hsueh, Leben R. Niller, P. (1999), A loop current induced jet along the edge of the west Florida shelf. Geophy. Res. Letter, 26, 2239-2242 Hsueh, Y., G. 0. Mannarino, and Linda L. Vansant (1982), Numerical model studies of the winter-storm response of the West Florida shelf, J. Phys. Oceanogr., 12, 1037-1050 Huh, 0. S., W. J. Wiseman, and L. J. Rouse (1981), Intrusion of Loop Current waters onto the West Florida shelf, J. Geophys. Res., 86, 4184-4192

PAGE 42

Koblinsky, C. J. (1981), The M2 tide on the West Aorida shelf. Deep-Sea Research, 28A, 12,1517-1532 10 Li, Z., and R. H. Weisberg ( 1999a), West Aorida continental shelf response to upwelling favorable wind forcing, part I: kinematic description, J. Geophys. Res., 104, 13507-13527 Li, Z., and R. H. Weisberg (1999b ), West Aorida continental shelf response to upwelling favorable wind forcing, part II: dynamical analyses, J. Geophys. Res., 104, 23427-23442 Marmorino, G. 0. (1982), Wind-forced sea level variability along the West Aorida shelf (winter, 1978), J. Phys. Oceanogr., 12, 389-405 Marmorino, G. 0 (1983), Variability of current, temperature, and bottom pressure across the West Aorida continental shelf, winter, 1981-1982, J. Geophys. Res., 88, c7, 4439-4457 Meyers, S.D., E.M. Siegel, R.H. Weisberg (2001), Observation of currents on the West Aorida shelfbreak. Geophys. Res. Lett. 28,2037-2040 Mitchum, T. G., and W. Sturges (1982), Wind-driven currents on the West Aorida shelf, J. Phys. Oceanogr., 12, 1310-1317 Niiler, P. P. (1976), Observations of low-frequency currents on the West Aorida continental shelf, Memoires Societe Royale des Sciences de Liege, 6, X, 331-358, 1976. Paluszkiewicz, T., L. Atkinson, E. S. Parmentier, and C. R. McClain (1986), Observations of a Loop Current frontal eddy intrusion onto the West Aorida shelf, J. Geophys. Res., 88, 9639-9651 Price, J. F. (1976), Several aspects of the response of shelf waters to a cold front passage, Memoires Societe Royale des Sciences de Liege, 6, X, 201-208 Sturges, W.(1994), The Frequency of ring separations from the Loop Current. J. Phys. Oceanogr,24, 1647-1651 Sturges, W., and R. Leben (2002), Frequency of ring separation from the Loop Current in the Gulf of Mexico: A revised estimate. J. Phys. Oceanogr., 30, 1814-1819 Tolbert, W. H., and G. G. Salsman (1964), Surface circulation of the eastern Gulf of Mexico as determined by drift bottle studies, J. Geophys. Res., 69, 223-230 Vargo, G.A., K. L. Carder, W. Gregg, E. Shanly and C. Heil (1987), The potential contribution of primary production by red tides to the West Aorida Shelf

PAGE 43

ecosystem, Limnol. And Oceanogr., 32, 762-767 Vukovich, F. M. (1988), Loop current boundary variations, J. Geophys. Res., 93, C12, 15585-15591 Weatherly, G L., and P J. Martin (1978), On the structure and dynamics of the oceanic bottom boundary layer, J. Phys. Oceanogr., 8, 557-570 Weatherly G. L., and D. Thistle (1997), On the wintertime currents in the Aorida Big Bend region, Cont. Shelf Res. 17, 1297-1319 Williams, J., W. F. Grey E. B. Murphy, and J. J Crane (1977), Drift bottle analyses of eastern Gulf of Mexico surface circulation, Memoirs of the Hourglass cruises, Marine Research Laboratory Aorida Department of Natural Resources, St. Petersburg, Aorida 11 Weisberg, R. H., B. D. Black, H. Yang (1996), Seasonal modulation of the West Aorida continental shelf circulation, Geophys Res Lett ., 23 2247-2250 Weisberg, R. H., B. D. Black, Z. Li (2000), A upwelling case study on the Aorida's west coast. J. Geophys Res. 105, 11459-11469 Weisberg, R. H., Z. Li and F MullerKarger (2001), West Aorida Shelf response to local wind forcing: April1998. J. Geophys. Res. 106 (C12), 31239-31262 Weisberg, R. H. and R. He (2002). Anomalous circulation on the west Aorida shelf: local and deep-ocean forcing contribution, J Geophys. Res. submitted Yang, H., R. H. Weisberg (1999), Response of the west-Aorida continental shelf to climatological monthly mean wind forcing, J. Geophy. Res., 104, 5301-5320 Yang, H, R. H. Weisberg, P P. Niiler W. Sturges, W Johnson (1999), Lagrangian circulation and forbidden zone on the West Aorida Shelf. Cont. Shelf Res. 19. 1221-1245

PAGE 44

12 CHAPTER2 TIDES ON THE WEST FLORIDA SHELF 2.1. ABSTRACT The principal semidiurnal (M2 and S 2 ) and diurnal (K1 and 01 ) tidal constituents are described on the west Florida continental shelf (WFS) using a combination of in-situ measurements and a three-dimensional, primitive equation numerical model. The measurements are of sea level and currents along the coastline and across the shelf width, respectively. The model extends from west of the Mississippi River to the Florida Keys with an open boundary arcing between. It is along this open boundary where the regional model is forced by a global tide model. Standard barotropic tidal analyses are performed for both the data and the model, and quantifiable metrics are provided for comparison. Based on these comparisons we present co-amplitude and co-phase charts for sea level and velocity hodographs for currents. The semi-diurnal constituents show marked spatial variability, whereas the diurnal constituents are spatially more uniform. Apalachicola Bay is a demarcation point for the semi-diurnal tides that are well developed to the southeast along the WFS, but minimal to the west. The largest semi-diurnal tides are in the Florida Big Bend and Florida Bay regions with a relative minimum in between just to the south of Tampa Bay. These spatial distributions may be explained on the basis of local geometry. A Lagrangian Stokes drift, coherently directed toward the northwest, is identified, but of relatively small magnitude compared with the potential for particle

PAGE 45

transport by seasonal and synoptic scale forcing. Bottom stress-induced tidal mixing is examined and estimates are made of the bottom logarithmic layer height by the M2 tidal currents. 2.2. INTRODUCTION 13 The Gulf of Mexico is a semi-enclosed basin connected to the Atlantic Ocean by the Florida Straits and to the Caribbean Sea by the Yucatan Channel. A notable feature of the Gulf of Mexico tides compared with other places around the world is a dominance of the diurnal over the semi-diurnal constituents [Reid and Whitaker, 1981]. This is in contrast with the east coast of the United States, where the tides are predominantly semi diurnal [Zetler and Hansen, 1971]. Previous Gulf of Mexico tides studies concluded that the diurnal tide is primarily co-oscillating, entering the Gulf of Mexico through the Florida Straits and existing through the Yucatan Channel. [Grace, 1932; Zetler and Hansen, 1971]. Located on the eastern side of the Gulf of Mexico, the west Florida continental shelf (WFS) is one of North America's broadest continental shelves. Apparently different from most of the Gulf of Mexico basin, semi-diurnal tides are appreciable here. Although this tidal structure was discussed in previous studies [Clarke, 1991; Reid and Whitaker, 1981], descriptions ofWFS tides are limited to coastal sea level and a few regional current measurements. Koblinsky (1981) described the M2 tide over the southern portion of the WFS using velocity measurements from five moorings over a 2-year duration. The surface wave crests of the M2 tide, considered to be stationary, linear, and barotropic, were found to parallel the isobaths. Internal tides were not found and the

PAGE 46

14 temperature distribution was only slightly distorted by the surface wave. Marmorino (1983) added analyses of velocity data from four moorings deployed in the Florida Big Bend region. Semi-diurnal tidal constituent (M2 and S 2 ) energy was found to decrease in the offshore direction, whereas the diurnal tidal constituent (K 1 and 0 1 ) energy was more uniform across the shelf. As a result, the characterization of the tidal fluctuations changes from predominantly diurnal in deep water to semi-diurnal near the coast. Such spatial inhomogeneity exists throughout the WFS. Weisberg et al. (1996) using velocity data from the 47 m isobath, showed that particle displacements in the semi-diurnal and diurnal bands are typically about 1 km. However, inertial oscillations, during months when the water column is stratified, can cause an increase in the diurnal band particle excursions to about 5 km. A deficiency of moored current meter data alone is that realistic arrays are insufficient to map the tides over the entire WFS. To do this we need a combination of in-situ data and a high resolution, area-encompassing model. The present paper combines recent in-situ measurements with a three-dimensional, primitive equation numerical model to describe and map the four major tidal constituents (M2, S2, K1, and 01 ) on the shelf from the Florida Keys to the Mississippi River. We concentrate on the barotropic tides since the baroclinic tides are seasonally modulated and an order of magnitude smaller. Exceptions to the baroclinic motions being relatively small are seasonally modulated inertial oscillations whose frequency overlaps with the diurnal tides. Baroclinic tides and inertial oscillations, whose effects on the deterministic barotropic tides are negligible [Clarke, 1991], will be the subject of a future correspondence.

PAGE 47

15 We begin in section 2.3 with the observations of coastal sea level and shelf-wide currents, analyzing the barotropic tides by standard methods. Section 2.4 then introduces the regional model and discusses how it is driven at the open boundary by deep-ocean barotropic tides. The modeled tides and their comparisons with the in-situ data for the M2 S2 and 01 constituents are described in section 2.5. Based on these comparisons we provide maps of the principal constituent tidal ellipses over the entire domain using the model. We also use the model to estimate the Lagrangian transports induced by the barotropic tides and their attendant non-linear interactions. Since the amplitude of the tides determines their contribution to turbulent mixing, this is a topic of discussion in section 2.6. The results are summarized in Section 2.7. 2.3. OBSERVATIONS The in-situ observations are taken from nine tide gauges operated by NOAA NOS and twelve current meter moorings deployed by the University of South Florida (Fig.2.1). The nine tide gauges, referred to as South Pass, Waveland, Pensacola, Panama City, Apalachicola, Cedar Key, St. Petersburg, Naples, and Key West span the northeast Gulf of Mexico from the Mississippi River to the Florida Keys. The velocity data are from acoustic Doppler current profilers deployed across the WFS between the 300 m to 10 m isobaths. 2.3.1 TIDAL HEIGHT The sea level data are analyzed using the least squares error method of Forman and Henry (1979). The analysis includes as many as 146 possible tidal constituents, 45

PAGE 48

Q) 28 'tJ :::s ; as ...1 27 26 25 -90 -89 -88 -87 -86 -85 -84 -83 -82 -81 -80 Longitude Fig. 2.1. Model domain and observational locations. The nine tide gauges (denoted by stars) are: South Pass, Waveland, Pensacola, Panama City, Apalachicola, Cedar Key, St. Petersburg, Naples, and Key West. The 12 moorings for velocity data (denoted by triangles) are: ASl (47m), TSl (31m), TS2 (47m), TS3 (46m), TS4 (63m), TS5 (142m), TS6 (296m), EC2 (50m), EC3 (30m), EC4 (20m), EC5 (10m), and EC6 (10m). 16

PAGE 49

17 of these being astronomical in origin and the remaining 101 being shallow water constituents [e.g., Godin, 1972] that arise by distortions of the astronomical tidal constituents due to the non-linear effects of shoaling depth. Year-long time series of sea level are used for each of the nine sea level station analyses inclusive of the years 19961999. Data gaps preclude using the same year for all stations. Table 2.1 lists the amplitudes and phases (relative to the Greenwich meridian) for the M2 S2 K1 and 01 constituents, which in sum contain over 90% of the tidal variance. We note that the semi diurnal tides to the south of are larger than those west of Apalachicola Bay. In contrast with the diurnal tides that have similar amplitudes at all nine stations, the semi-diurnal tides show large spatial variations. All of the four major constituents have amplitudes that peak near Cedar Key. To quantify the contribution that the tides make to the total sea level variance, a predicated tidal height time series, P is constructed by adding the four major tidal constituents. With being the observed sea surface height time series after removing the mean, the root mean square tidal residual (i.e., the sub-tidal component), a r and the normalized residual, arIa 0 where a 0 is the standard deviation of are calculated over the time series record length, T, as (2.1) The ratios of the sub-tidal variability to the total sea level variability, arIa 0 are given as percentages in Table 2.2. We note that west of Apalachicola Bay, where the tides are generally small, the sub-tidal variability accounts for more than 70% of total variability.

PAGE 50

18 Table 2.1. Comparison of observed and computed tidal elevation at reference sites. Station Amplitude H (mxl0-2 ) Llli Phase


PAGE 51

19 Table 2.2. Normalized rms tidal residuals showing the contributions of non-tidal fluctuations to the sea level variability at the nine tide gauges spanning the study domain. Gauge South WavePensaco) Panam Apalachi Cedar St. Pete Naples Key pass -land -a -aCity -cola Key West CT, Oft 70.88 74.74 70.76 71.03 70.64 50.71 57.24 47.14 56.41 -o CTo South of Apalachicola Bay, where the tides are larger, the tidal and the sub-tidal variabilities contribute about equally to the total sea level variability. 2.3.2. TIDAL CURRENTS In parallel with sea level the tidal analyses for the moored velocity data are performed using the least squares error method of Forman (1978). Since the 12 mooring locations were generally not co-deployed the analysis intervals (between 1996 and 1999) differ in record length and time origin. We generally used the entire record length available at each site, and these varied from around three months to one year. Parameters describing the tidal hodographs are calculated at each depth bin (0.5 m to 10 m depending on water depth) for each mooring. The amplitudes of the ellipse semimajor axis and the ellipse orientations (measured anti-clockwise from east) are shown as a function of depth in Figs. 2.2 and 2.3 for the four major constituents: Mz, S2 01 and K1 For a purely barotropic tide the amplitude and orientation would be independent of depth and appear as vertical lines. Depth profile deviations from straight lines are related to bottom friction, baroclinicity, contamination by inertial motions, or combinations of these. As shown in Fig. 2.2 for relatively shallow depths and in Fig. 2.3. for relatively deeper depths, the semi-diurnal tides indeed appear to be barotropic, which agrees with

PAGE 52

EC2(Umaj) -40 -45 -50 10 15 EC4(Umaj) -2 -4 -6 E' -6 ::' a -1a Q) 0 -12 -14 -16 -18 I II 'I EC2(0rientation) -5 -10 -15 -20 -25 -30 -35 -40 -45 50 100 150 EC4(0rientation) -2 -4 -6 -8 -10 -12 -14 -16 -18 \ \ -20 -20 0 10 15 0 50 100 150 EC6(Umaj) EC6(0rientation) -1 -2 -3 E' -4 =a -5 Q) 0 -6 -7 -8 -9 i r I I I -1 -2 'I' -3 -4 -5 -6 -7 -8 -9 -10 -10 0 10 15 0 50 100 150 [em] [dgree] EC3(Umaj) -5 -10 -15 -20 II II \J v -25 /I EC3(0rientation) -5 -10 -15 -20 -25 II II t ( I ; / / -30 -30 0 10 15 0 50 100 150 ECS(Umaj) -1 -2 I! ; : : /! -5 -6 I/ .! -7 UJ ui -8 -9 ECS(Orientation) -1 -2 -3 -4 -5 -6 -7 -8 -9 \ / I \ \ \ '( ; ,, I I \ I I -10 -10 0 10 15 0 50 100 150 [em] [dgree] Fig. 2.2. Vertical profiles of the velocity hodograph ellipses semi-major axis amplitudes and orientations (measured anti-clockwise from east) for the EC designated moorings. The thick solid and dashed lines denote M2 and S2 respectively. The thin solid and dashed lines denote 01 and K1 respectively. 20

PAGE 53

TS1(Umaj) -5 -10 'E E. -15 (J) 0 -20 -25 ol i / -30 0 10 15 TS3(Umaj) -5 -10 -15 E -2o ..r::: 0 -30 -35 -40 I. I; I II 'I ,. I \ -45 0 10 15 TSS(Umaj) -20 -40 'E -eo :c c. l (J) -80 l 0 ,, I : \ -100 -120 -140 0 10 [em] 15 TS1 (Orientation) -5 I -10 I I -15 -20 -25 -30 50 100 150 TS3(0rientation) -5 -10 -15 -20 -25 -30 -35 -40 I } \ \ 46 0 50 100 150 TS5(0rientation) -20 -40 -eo -eo -100 -120 -140 1--------------_r 50 100 150 [degree] -5 -10 -15 -20 -25 -30 I :, f I} f1 r ( i I I TS2(Umaj) -35 -40 0 -10 -20 -30 -40 -50 -60 0 -100 -150 -200 -250 0 10 TS4(Umaj) I I I I I i \ I i I f I I I \ I \ I ( \ I; 10 TS6(Umaj) 10 [em] 15 15 15 TS2(0rientation) -5 -10 -15 -20 -25 I'( I; I! \ 1 I \ J -40 -45 50 100 150 TS4(0rientation) -10 -20 -30 -40 -50 -60 0 50 100 150 TS6(0rientatio n) \ -so \ ----100 -150 I \ \ \ '-. -200 \ -w-, ---' -250 0 50 100 150 [degree] Fig. 2.3. Vertical profiles of the velocity hodograph ellipses semi-major axis amplitudes and orientations (measured anti-clockwise from east) for the TS designated moorings. The thick solid and dashed lines denote M2 and S2 respectively. The thin solid and dashed lines denote 01 and K" respectively. 21

PAGE 54

22 the finding of Koblinsky (1981) for the WFS M2 tide. In contrast, the diurnal tides, especially in deeper water, tend to show vertical structure. This is most likely due to contamination by inertial motions since in the central WFS the inertial period is about 25.90 hrs which is close to the 01 and K1 periods of 25.82 hrs and 23.93 hrs, respectively. WFS inertial oscillations are modulated in time along with the seasonally varying stratification [Weisberg et al., 1996], and they are evident in velocity component plots as baroclinic modes consistent with the orientation reversals seen for the deeper records in Fig. 2.3. Overall, the observed tidal current amplitudes on the WFS are weak (on the order of a few em s-1 ). Of the principal constituents, the M2 current is the strongest; its amplitude tends to increase shoreward, and its ellipse orientation tends to align perpendicular to the isobaths While weaker, the diurnal constituent amplitudes are more uniformly distributed across the shelf, and these findings agree with those of Mannorino (1983) for the Florida Big Bend region. 2.3.3. PHASES OF TIDAL HEIGHT AND TIDAL CURRENTS One question of interest (particularly for recreational fishermen) is the relative phase between the tidal height at the coast and the tidal currents on the WFS. We address this question relative to the St. Petersburg tide gauge, the reference gauge for much of the WFS, in two different ways Table 2.3a shows the phases (and corresponding time lags) between high tide at St. Petersburg and maximum tidal currents (directed toward the coast) at the 12 ADCP mooring sites for each of the major tidal constituents. For the semi-diurnal tides, the times of maximum tidal current generally lead th e high tide at St.

PAGE 55

23 Petersburg by 2-9 hours. On the other hand, the maximum tidal currents of the diurnal tide, lag the high tide time at St. Petersburg by 10-20 hours. Table 2.3a. Greenwich phases of the major tidal constituents at the St. Petersburg tide gauge and at the 12 ADCP stations (left hand columns), plus the relative times between high water at St. Petersburg and maximum (shoreward directed) semi-major axis tidal currents at the 12 ADCP stations (right hand columns). Phase differences are converted to time differences using speeds of M 2 S 2 01 and K1 of 29.98 deg/hr, 30.00 deg/hr, 13.94 deg/hr and 15.04 deg/hr, respectively, where"+" and"-" indicate time lags and leads, respectively. Station M2 s2 01 K1 (degree) (hour) (degree) (hour) (degree) (hour) (degree)/(hour) St. Petersburg 198.10 0.00 216.10 0 .00 38 70 0.00 50.50 0.00 AS1 148.80 -1. 70 123 .81 -3.07 23.26 -1.10 311.92 17.38 TS1 51.39 5.06 70.41 -4.86 269.34 16.54 262.49 14.09 TS2 21.39 -6 10 33.37 -6 09 250.58 15.19 237.69 12.44 TS3 63.57 -4.64 84.84 -4.37 246.39 14.89 280.66 15.30 TS4 52 08 -5.04 69.01 -4.90 239.42 14.39 250.05 13.26 TS5 54. 16 -4.97 68 94 -4.90 261.06 15.95 241.93 12.72 TS6 53.56 -4 98 69.73 -4.87 322.91 20.38 208.68 10.51 EC2 45 .17 -5.2 7 58.13 5.26 260.83 15.93 286.86 15.71 EC3 47.09 -5.21 56.72 -5.31 256.08 15.59 284.58 15.56 EC4 20.35 -6 13 40.23 -5.86 226.26 13.45 270.08 14.59 EC5 -9.64 -6.93 33.71 -6.07 221.70 13.12 275.70 14.97 EC6 -93.69 -9.76 -106.10 -10 74 159.10 8 63 139.98 5 94 Table 2.3b. The time lags between high tide at St. Petersburg and the maximum (shoreward directed) semi-major axis tidal currents at the 12 ADCP stations determined by the maximum lag-correlation coefficient between the composites of the M2, S2, Oh and K1 constituents Station St. Pete AS1 TS1 TS2 TS3 TS4 TS5 Time Lag 0.00 -2.31 -5.53 -6.45 -5.25 -5.18 -5.65 (hour) Station St. Pete TS6 EC2 EC3 EC4 EC5 EC6 Time Lag 0.00 -5.68 -5.77 -5.86 6.36 -7.37 -8.16 (hour)

PAGE 56

24 A second approach is to construct time series of the tidal height at St. Petersburg and the tidal currents at the 12 ADCP sites by combining the four major tidal constituents, and then producing a set of lags for these composite time series. Table 2.3b gives these maximum correlation time lags between the water level TJ and the major component of tidal current u defined as (2.2) where R(-r) = E[T}(t)u(t + -r)], crTJand cru are the standard deviation for TJ and u, respectively and -ris the time lag The time lags determined for the composite are similar to the time lags found for M2 constituent alone (Table 2.3a), reiterating the dominance of the M2 constituent on the WFS. 2 4. HYDRODYNAMIC MODEL Our work is preceded by several numerical model studies of Gulf of Mexico tides. Reid and Whitaker (1981) applied a finite difference version of the linearized Laplace tidal equations in a two-dimensional15' by 15' (28 km) horizontal grid to portray the barotropic response of the Gulf of Mexico to tidal forcing. While describing the deepwater tides well, their coarse grid size was insufficient to describe the tidal structures on the shelf. Westerink et al. ( 1993) applied a finite-element based hydrodynamic model (ADCIRC-2DDI), also depth-integrated and two-dimensional, to the Western North Atlantic, the Gulf of Mexico, and the Caribbean Sea, to develop a tidal constituent database. C omputations were pre sent e d for varying horizontal resolutions ranging from

PAGE 57

25 coarse (1.6Xl.6) to fine (6'x 6'), and discussion was given to the importance of model resolution. Their model comparisons with sea level data show good agreements in the Atlantic Ocean, but persistent errors occurred in the Gulf of Mexico and the Caribbean Sea. These authors suggested that insufficient resolution over the continental shelf, particularly in the shallowest regions, might be responsible for the relatively poor numerical convergence in those regions. More recently, Kantha et al. (http://www.ssc.erc.msstate.eduffides2D/) developed a data-assimilative, barotropic tidal model for the Gulf of Mexico with similar grid resolution as Reid and Whitaker (1981 ). Improved results follow from the assimilation of selected coastal tide gauge data. Nevertheless, this model has insufficient resolution to describe the structure of the tidal variations on the WFS. Our approach is therefore a regional one. The hydrodynamic model used is the Princeton Ocean Model (POM), a three-dimensional, nonlinear, primitive equation model with Boussinesq and hydrostatic approximations [Blumberg and Mellor, 1987]. The model uses an orthogonal curvilinear coordinate system in the horizontal and a sigma (a = z TJ ) coordinate system in the vertical, where z is the conventional vertical TJ+H coordinate (positive upward from zero at the mean water level), His the local water depth, and TJ is the tidal variation about the mean water depth. Our model domain (Fig. 2.4) extends from the Mississippi River in the northwest to the Florida Keys in the southeast, with one open boundary that arcs between these two locations. Horizontal resolution varies from less than 2 km near the coast to about 6 km near the open boundary, and the minimum water depth is set at 2m. This grid allows us to resolve the

PAGE 58

30 29 Q) 28 "0 2 ....J 27 26 25 Model Grid -91 -90 -89 -88 -87 -86 -85 -84 -83 -82 -81 -80 Longitude Fig. 2.4 The model grid used for the regional tidal simulation. 26

PAGE 59

27 complicated features of the coastline and the isobath geometries in order to explore the associated structures of the WFS tides. We use 21 sigma layers in the vertical with higher resolution near the bottom to better resolve the frictional boundary dynamics. Bottom stress is obtained by a quadratic drag law in which a non-dimensional drag coefficient is calculated on the basis of a specified (0.01 m) bottom roughness length. The model has a total of 121x81x21 grid points, and the time step for the external mode is 12 seconds. Water density is assumed homogeneous in this three-dimensional barotropic tide study. The three-dimensional distribution of the vertical eddy viscosity is computed using the Mellor and Yamada (1982) level 2.5 turbulence closure scheme, and the horizontal eddy viscosity is calculated using the shear-dependent Smagorinsky formulation [Smagorinsky, 1963] with a coefficient of 0.2. Dissipation by vertical friction is found to largely exceed that by horizontal friction. The bathymetry adopted in the model begins with the ETOP05 5'x5' global bathymetry data set. Since this data set is inaccurate in the southern portion of WFS we corrected it by incorporating data from NOAA navigation charts and NGDC bathymetry contours. Our modified bathymetry has been used in other WFS modeling studies with satisfactory results (e.g., He and Weisberg, 2002; Weisberg et al., 2001; Yang et al., 1999). Tidal forcing of the model is exclusively at the open boundary. There, the tidal elevations are specified by a linear interpolation of the output from the TOPEX/Poseidon data assimilated global tidal model of Tierney et al. (2000). This barotropic model, with 15' resolution, provides eight tidal constituents (M2, S2, N2, K2, K1o 01o P1 and Q1) over a

PAGE 60

28 nearly global domain that extends from 80S to 66N. The model assimilates both coastal tide gauge data and ocean tides derived from four years of TOPEX/Poseidon satellite altimetry data. Since it uses procedures that tend to preserve the spatial tidal structures in shallow water, it is considered to be a suitable product for forcing higher resolution regional tidal models. Figure 2.5 shows the distributions of tidal amplitudes and phases for M2, S2, K1 and 01 constituents along the open boundary where we have a total of 121 model grids. 2.5. MODELED ELEVATIONS AND CURRENTS 2.5.1. GENERAL FEATURES The model is forced at the open boundary by specifying the composite M 2 S 2 and 01 constituent elevations there. Elevations, currents, and turbulence quantities are computed over the model interior by integrating forward in time beginning from a state of rest. An initial spin-up period of five inertial cycles (about five days) is used to suppress transients. The subsequent 30 days are then analyzed using the same techniques as for the in-situ data to retrieve individual constituent amplitude and phase distributions over the model domain. Table 2.1 provides amplitude and phase comparisons between the modeled and observed elevations of M 2 S2, K1 and 01 at the nine coastal tide gauge stations spanning the analysis domain. Amplitude differences are generally less than 2 em; the singular exception being at Cedar Key where it is 2.5 em for M2. Phase differences are generally less than 10; with Naples at 10.1 o for 01 being the largest. Both the amplitude and phase differences show equally likely plus or minus values

PAGE 61

25 E' Q) -g 15 .:!::: c( 5 ----.. -----------------'G) 300 Cl Q) "C 'a;' 200 Ul ca .r::: D. 1-100 :!: ----!--:----.-_ 1 20 40 60 80 100 121 Boundary Grid Index Fig. 2.5. The distributions of tidal amplitudes {upper panel) and phases (lower panel) along the model open boundary for the M2 S2 K1 and 01 constituents (see legend pro vided). The abscissa is the grid index that includes both points over land and water. The land points are masked so only the points over water are shown. 29

PAGE 62

30 demonstrating that the modeled tides are not biased. Root mean square (rms) values of the phase differences, when converted to time, amount to less than 10 min (20 min) for the semi-diurnal (diurnal species). Considering the model tide gauge sampling relative to the actual tide gauge positions and the near-shore masking (2 m being the shallowest model depth), these phase agreements are both good and as good as can be expected. Therefore, no model tuning was considered. The along-coast variations of the amplitudes and phases of the four major tidal constituents are graphically shown in Fig. 2.6, where circles denote the observations and crosses indicate the model results. Allowing for scale changes in the abscissa we see a relative spatial homogeneity in the diurnal species compared with a more spatial inhomogeneous distribution for the semi-diurnal species. We also note several bulls-eyes in the model-observation comparison. The agreements of Table 2.1 and Fig. 2.6. justify using the model to produce co amplitude and co-phase charts for the M 2 S 2 Oh and K1 constituents (Fig. 2.7). In general agreement with the findings of the previously cited Gulf of Mexico tide studies, these maps provide further details of the tidal structures on the WFS. The semi-diurnal and diurnal species are distinctly different. The phase of the M2 tide advances toward the northwest. Apalachicola Bay separates the M 2 tidal regime into two parts. To the west the M 2 tide is weak. To the east it is strong. Both the Florida Big Bend and the Florida Bay regions show relative maxima with a relative minimum in between just to the south of Tampa Bay. The S 2 constituent shows features similar to the M2 constituent, but with a much smaller amplitude. The phase patterns for the semi-diurnal species appear to result by diffraction around the Florida Keys and onto Cape San Bias.

PAGE 63

31 Southpass ()( M2 South pass ox M2 Waveland ox Waveland :() Pensacola Pensacola l() Panama Panama l() Apalachl o x Apalachl Cedar Key 0 Cedar Key St. Pete (): St. Pete Naples >() Naples I Key West >0 Key West 0 10 20 30 40 50 100 150 200 250 300 South pass S2 South pass S2 Waveland I Waveland >() Pensacola Pensacola Panama Panama >0 Apalachl o x Apalachl Cedar Key 0 X Cedar Key ()( St. Pete 0 X St. Pete 0< Naples X 0 Naples ()( Key West X 0 Key West 0 2 4 6 8 10 12 50 100 150 200 250 300 Southpass X 0 01 Southpass 01 Waveland l() Waveland >C Pensacola () Pensacola x o Panama X 0 Panama X 0 Apalachl 0 X Apalachl l() Cedar Key xo Cedar Key () St. Pete 0< St. Pete ox Naples 0 X Naples 0 X Key West o x Key West ox 8 10 12 14 16 18 -10 0 10 20 30 40 50 South pass xo K1 South pass 0 X K1 Waveland ()( Waveland l() Pensacola I Pensacola x o Panama >() Panama X 0 Apalachl ox Apalachl xo Cedar Key X 0 Cedar Key ox St. Pete ) St. Pete 0: Naples :() Naples 0 X Key West (0 Key West >0 5 10 15 20 25 -20 0 20 40 60 80 Amplitude [em] GMT Phase [degree] Figure 2.6 Comparisons between modeled and observed amplitudes (left panels) and Greenwich phases (right panels) for S2 K., and 01 constituents at the nine coastal tide gauge locations. Crosses (circles) denote modeled (observed) values Bulls eyes are where they overlay.

PAGE 64

32 Amplification of the semi-diurnal tides in the Aorida Big Bend has been discussed by previous investigators. Reid and Whitaker (1981) considered a resonance deriving from the counterclockwise speed of propagation matching the group speed for a gravest mode edge wave ( gS If where Sis the bottom slope-Kajiura, 1958). With these speeds in close alignment and with the M2 tide wrapping around the basin approximately once per tidal cycle, near resonance may be achieved (these authors estimated an average M2 propagation speed of 98 m s1 and an edge wave speed of 120m s-1). While this type of resonance is possible, we note that bottom topography and edge wave speeds vary around the shelf as does the M2 tide amplitude and phase gradient. Local geometry may also induce amplification. The isobath (Fig. 2.1) and the coamplitude plots for the semi-diurnal tides (Fig. 2.7) show similarity. Where the shelf is narrow the amplitudes are small and conversely The two regions of widest shelf are the Aorida Big Bend and the Aorida Bay with a relative minimum in between, as reflected in the co-amplitude distributions. Effects of local geometry are considered in the analytical treatments of Battisti and Clarke (1982a,b) and Clarke (1991). For a continental shelf with an along-shore scale much greater than the shelf width these authors developed a one dimensional tide model to explain some of the observed barotropic tidal characteristics. Using a linearized bottom stress (nuh), the linear Laplace tidal equations are and [.i. + !..]v + fu = -g 7] dt h y (2.3) For tidal motions over a continental shelf with constant slope a and width a, i.e.,

PAGE 65

M2 52 31 31 29 29 / 2 8 I 28 \ 27 27 }
PAGE 66

34 h(x) = a x, 05 x 5 a, they expressed sea level in terms of a zero order Bessel function as, --= J 0 [2(.ua) ] where f.Jil::::: a 'f}(a) 112 (m2 12) rJ(O) ga (2.4) For the semi-diurnal tide ( m 2 -f 2 > 0) on a wide shelf (a! a is large), f.Jil ---t 1 and 'f}(a) I l](O) ---t 0, implying amplification of semi-diurnal tides across a wide shelf. To account for along-shore variations in the tides, Lentz et al. (2001) developed an analytical, flat-bottom, two-dimensional model with varying shelf width. These analytical models are both consistent with the observed semi-diurnal tide behaviors on the WFS. They suggest that the co-amplitude distributions may be the result of local geometry, as contrasted with the previous basin-scale resonance arguments. Unlike the semi-diurnal tides, the diurnal tides show little co-amplitude variations across the analysis domain. Both the amplitudes and phases for the 01 and K1 constituents are nearly spatially uniform. This is consistent with the Zetler and Hansen ( 1971) finding that the diurnal tide is essentially in phase throughout the Gulf of Mexico. It is also consistent with equation 4 (Clarke, 1991), since form::::: f on the WFS, f.Jil ---t 0 so the diurnal tides do not amplify toward the coast. The 01 and K1 constituents also have comparable amplitudes with mean ratio of 0.95. Basin-scale resonance for these constituents is not expected because their propagation speeds are about half those of the semi-diurnal constituents so they do not circumscribe the basin within a tidal period as do the semi-diurnal species [Reid, personal communication].

PAGE 67

The ratio between the amplitudes of the principal semi-diurnal and diurnal tide constituents, CM2+S2)/(Kt+Ot), is shown in Fig. 2.8. While the WFS tidal regime is generally characterized as mixed and mainly of diurnal type, this is not true of the inner half of the WFS where the ratio exceeds 0.5 and especially of the Florida Big Bend and Florida Bay regions where the semi-diurnal tides dominate. It is only over the outer portion of the WFS and to the west of Apalachicola Bay where the diurnal species dominate. 35 Given the model sea level descriptions and the model and in-situ data agreements for sea level at the coast, we now examine the tidal currents in a similar manner. Fig. 2.9 compares the depth-averaged tidal current hodographs (calculated for all of the in-situ data locations of Fig. 2.1) with the depth-averaged tidal current hodographs calculated from the model sampled at all of these in-situ data sites. Qualitatively, the comparisons are very good for the semi-diurnal constituents, with ellipse semi-major amplitudes, eccentricities, and orientations all in agreement (see Table 2.4 for listings of the hodograph ellipse semi-major axes, semi-minor axes, and orientations). Particularly notable is how the ellipse orientations diverge away from the region of relative minimum semi-diurnal tide elevation just south of Tampa Bay. Eccentricities at the stations close to the coast show the largest departures between the data and the model, which may be a consequence of the model turbulence parameterization. Deviations from an inviscid fluid semi-minor to semi-major axis ratio of wf -I are due to the friction, which is consistent with the largest departures being in the shallowest water. Nevertheless, the agreements are still very good. The overall agreements between the observed and modeled diurnal

PAGE 68

30 29 28 27 26 25 (M2+S2)1(01 +1<1) -90 -89 -88 -87 88 -85 -84 .... 0 I I \ -83 -82 -81 Figure 2.8. The amplitude ratio between the principle semi-diurnal and diurnal tides. 36

PAGE 69

constituents, while also good, are degraded somewhat from those of their semi-diurnal brethren, and there are additional reasons for this behavior. 37 As pointed out by Reid and Whitaker (1981), modeled currents are more sensitive to factors such as bottom friction, horizontal eddy diffusivity, and topography than are modeled water levels. A particular problem in the Gulf of Mexico for diurnal constituents is the closeness in frequency between these and the local inertial period. Large inertial oscillations exist in the WFS observations when the water column is stratified. While not a subject of this paper, we did investigate several different techniques in attempting to minimize inertial oscillation interference in the data. For instance, we broke the data records into month-long pieces to distinguish barotropic tides during months without stratification versus mixed barotropic tides and inertial oscillations during months with stratification. Non-uniform modulation made this too subjective, however. Recognizing that the inertial oscillations when present are predominantly of first baroclinic mode structure we opted to perform the data analyses on depth-averaged currents, effectively averaging out the inertial oscillations. Nevertheless, some non-tidal inertial oscillatory behavior may still be contaminating the diurnal constituent hodographs of Fig. 2.9. Similar problems do not exist for the semi-diurnal constituents since they are removed in frequency from the inertial oscillations. Some amplification is observed in the data during months when the WFS is stratified, but this is small relative to the barotropic tides. Whether such amplification is the result of internal tide generation or a reduction in the effects of bottom friction by the stratification remains to be determined.

PAGE 70

Observation 29. 5 29 28. 5 28 27.5 27 26.5 -85 -84 -83 -82 -81 29.5 29 28.5 28 27. 5 27 0 26.5 -85 -84 -83 -82 -81 29. 5 29 28.5 28 27. 5 27 26.5 0 Do 0 tJ c::> \\ oo 0 -85 -84 -83 -82 -81 29. 5 29 28. 5 28 27. 5 27 26.5 0 OJ


PAGE 71

39 Table 2.4. A comparison of computed and observed tidal ellipses parameters Semi-major (X10-2m/s) Semi-minor (X10-2m/s) Orientation(degree) ----------------------------------------------------------------Observed Modeled Observed Modeled Observed Modeled AS1 9.38 9.52 -2.01 -2.22 29.05 31.34 TS1 10.21 8.78 -1.30 -0.97 49.52 42.98 TS2 8.12 7.38 -2.14 -1.34 31.94 33.26 TS3 9.95 8.92 -1.84 -1.79 32.99 34.29 M2 TS4 7.21 6.19 -1.70 -1.57 23.95 26.37 TS5 3.83 3.49 -1.52 -0.89 24.05 25.06 TS6 1. 79 1. 96 -0.28 -0.49 25.64 29.58 EC2 5.82 5.84 -2.48 -1.12 23.05 18.93 EC3 5.24 5.07 -2.68 -0.70 22.13 19.71 EC4 4.40 4.15 -2.71 -0.49 30.58 22.73 EC5 2.75 2.22 -1.92 -0.49 79.44 59.75 EC6 14.35 12.01 -2.03 -1.03 151.08 155.76 -----------------------------------------AS1 4.65 2.55 -0.71 -0.54 35.88 29.47 TS1 2.81 2.63 -0.09 -0.05 50.48 41.37 TS2 3.08 2.16 -0.41 -0.26 33.04 30.87 TS3 2.82 2.57 -0.58 -0.36 26.17 32.13 S2 TS4 3.83 1. 80 -0.70 -0.38 25.43 24.02 TS5 1. 60 1. 01 -0.66 -0.24 29.46 22.93 TS6 0.70 0.56 -0.21 -0.15 38.61 28.17 EC2 1. 98 1. 86 -0.93 -0.37 21.30 15.94 EC3 1. 69 1. 63 -1.12 -0.25 18.88 14.96 EC4 1. 52 1. 34 -0.75 -0.20 22.38 16.47 EC5 0.86 0.68 -0.44 -0.17 26.32 35.76 EC6 5.91 4.23 -1.03 -0.52 150.65 154.71 ------------------------------------------------------------AS1 5.44 4.27 -4.08 -3.18 29.44 20.81 TS1 2.68 3.17 -1.40 -2.49 64.88 65.88 TS2 2.25 3.08 -0.69 -2.51 49.19 58.98 TS3 3.71 3.72 -2.69 -3.08 39.13 45.84 01 TS4 2.06 2.93 -1.24 -2.50 40.48 46.56 TS5 2.95 1. 95 -1.09 -1.53 40.78 51.05 TS6 1. 47 1.24 -0.66 -0.88 65.52 44.82 EC2 2.43 2.10 -1.87 -1.95 70.87 86.65 EC3 2. 38 2.01 -1.50 -1.64 105.01 117.00 EC4 2.27 1. 79 -0.51 -1.32 103.83 129.66 EC5 2.63 1. 30 -0.43 -0.74 106.65 137.52 EC6 3.50 2.53 -1.36 -,0.81 134.03 132.82 ------------------------------------------------------------AS1 4. 34 4.93 -3.49 -3.32 19.12 24.44 TS1 4. 38 3.90 -3.53 -2.70 50.56 57.05 TS2 4.34 3.65 -3.09 -2.68 40.14 49.93 TS3 6.11 4.41 -4.61 -3.24 32.49 41.59 Kl TS4 3.80 3.40 -2.64 -2.64 38.58 39.43 TS5 3.87 2.19 -3.41 -1.60 55.58 44.20 TS6 3.00 1. 38 -2.27 -0.91 23.79 41.98 EC2 2.77 2.47 -1.41 -2.18 38.07 54.62 EC3 3.53 2.26 -2.97 -1.95 112.20 100.32 EC4 2.75 2.03 -1.94 -1.60 105.56 119.29 EC5 1. 74 1. 56 -0.23 -0.90 141.99 129.42 EC6 3.44 2.83 -0.69 -0.33 134.64 130.94

PAGE 72

40 The agreements between the observed and modeled tidal current ellipses are sufficiently good to warrant the mapping of these ellipses over the entire model domain. These are shown in Fig. 2.10 for the Mz, S2 and K1 constituents. The general patterns seen in the co-amplitude contours are reflected here. M2 currents amplify from very small values in deep water to largest values (approaching 20 em s-1 ) near-shore in the Florida Big Bend and Florida Bay. The relative minimum in between to the south of Tampa Bay is evident with the hodograph orientations diverging away from this point. To the west of Apalachicola Bay we see very little M2 tidal current except just to the east of the Mississippi River where the shelf again widens away from the DeSoto Canyon. The S2 tidal currents behave very similarly to the M2 tidal current except that they are weaker. The and K1 ellipses appear very similar to one another, being weaker than the M2 and stronger than the S2 ellipses over the WFS, but stronger than either of these to the west of Apalachicola Bay. Eccentricities in general agree with the tendencies from inviscid theory that the tidal ellipses should be more circular (ffit1 ) for the diurnal species than for the semi-diurnal species, and the polarizations are clockwise. 2.5.2 TIDAL RESIDUAL CURRENT AND THE LAGRANGIAN TRANSPORT The model, being fully non-linear, is capable of generating responses in addition to the four linear tidal harmonics with which it is forced. Here we investigate these tidal residual currents (Ures, Vres) defined as the difference between the model responses (U, V) and the four constituent tidal harmonic currents analyzed from the model output. Thus, the Eulerian residual current is calculated as:

PAGE 73

u res = uL,ui cos(a;tai) V, ., = V-L, v; cos(a;t/3; ) (2.5) 41 where (u;, v;) and ( /3;) are the analyzed harmonic constants for the M2 S2 K1 and 01 tidal constituents. Figure 2.11 presents the residual currents in both Eulerian and Lagrangian frameworks. The Eulerian representation (upper panel) is the residual surface current field after subtracting from each grid point the linear least squares fit of the four major tidal constituents. The Eulerian residual field is weak (generally less than 0.01m s 1 ) and generally directed toward the north or northwest consistent with the propagation direction for the semi-diurnal tides. Vectors appear largest in the Florida Big Bend where they also form an anticyclonic gyre. The Lagrangian representation (lower panel) is calculated by tracking surface drifters originating at every fourth grid point over a 30-day period, i.e ., they are Lagrangian trajectories for dynamically passive particles released at the surface on day 6 (after the five day spin-up period) and tracked through day 35. The patterns appear very similar to the Eulerian representation. Particle displacements are small (order 10 km per month) suggesting that barotropic tidal current rectification is not a mechanism of major importance in transporting materials on the WFS 2.6. BOTTOM STRESS AND TURBULENCE MIXING Turbulence mixing in this POM based barotropic tide model derives from bottom stress calculated us ing a quadratic drag law. Here we examine the bottom stress distribution, its mi x in g ramifications for the water column, and the implic a tions of these findin g s for simpler models that employ a lin e ar drag l a w For simplicity only the M z current is considered since it is the dominant contributor to the WFS tidal currents.

PAGE 74

M2 31 { 30 .... { ' ? \ 29 I'. l?c?/f/_ "'o -..,\. 28 .. V -., ::'<\ , 27 .... ......... 26 , ... 25 -90 -89 88 -87 86 -85 -84 -83 82 -81 30 _,........._ 0 0 29 .. 0 0 0 :' . . 00 28 0 0 0 d.,., 0000\ 27 26 25 0 0 ooJ.,_ Ooo-; 0 0 0, ... :.--:, ....... -90 -89 88 -87 86 -85 -84 -83 82 -81 28 27 26 25 ., ., /.?/ ,,.,t?i:. .... -90 -89 -88 -87 86 -85 -84 -83 -82 -81 28 27 e's 26 25 -90 89 -88 87 86 85 84 83 -82 -81 Fig. 2.10. The distributions over the entire model domain of the modeled M2 S2 01 and K1 tidal current hodograph ellipses. 42

PAGE 75

30 29 28 27 26 25 -88 -88 Surface Tidal Current Residual .. \ ... \ r ,.-/..,. \ 1 cm/s J'* ', -87 86 \ . '.. \ .. . ,, ,, \\ ,,, ,....._ _;'//1 } -86 -84 . \' \ \\' -,-I ... ,,, ,,,,__.. ,,,, r ''---1 -85 I I \ I 1 . I I t I t \ I I f f f ';;r-1' I I f f f 7U. I I I t I ; ''It,,\;. -84 I I I \ I I I \'' r I \ ', -83 I ' ' -82 Surface Drifters -82 -88 -86 -84 -81 -82 Fig. 2 .11. Eulerian (upper panel) and Lagrangian (lower panel) representations of the tidal residual surface circula t ion. 43

PAGE 76

44 Bottom stress, for either depth-dependent or depth-averaged models, may be defined in several different ways, such as: (2.6) where pis the water density and for the first two (depth-dependent) cases u. is the friction velocity, Co is a bottom drag coefficient, and u is the near-bottom current. For the third (depth-averaged) case u is the depth-averaged current and r is a resistance coefficient. For our application, we consider the rms value of Tb calculated over 10M2 tidal cycles from which the friction velocity u. is obtained as (Tb I p)112 A map of this u. for the WFS M2 tide is given in the upper panel of Fig. 2.12. The spatial variations of u. or the magnitude of turbulence mixing reflects the spatial variations in the tidal current magnitude. Regions of the largest tidal mixing are associated with regions of the strongest tidal currents found in relatively shallow water. In particular the maximum values of u. (greater than 0.36 em s-1 ) are found in the Florida Big Bend and in Florida Bay. Since bottom stress is calculated by a quadratic drag law, it is of interest to check the sensitivity of our results to the bottom drag coefficient C0 The adjustable parameter in Cd = max {0.0025, (2.7) is the bottom roughness length zo. The default value for zo is 0.01 m. For sensitivity studies we ran a series of model experiments in which we varied zo between 0.002 m and 0.05 m and found no significant differences in model performance. We then replaced

PAGE 77

45 equation 7 with a smaller spatially uniform drag coefficient (Cd = 0.001) This substantial change in drag coefficient, especially in shallow water, resulted in a disproportionately small change in the model response. In analogy to a damped harmonic oscillator, where proportionately small (large) increases in response occur away from (near) resonance, our sensitivity findings suggest that tidal amplifications in the Big Bend and Florida Bay regions are non-resonant, supporting the earlier assertion that geometry is the cause of the observed amplification. For the case of linear, depth-averaged models bottom stress is set by the resistance coefficient r As examples Clarke (1991) used r=2x104 ms1 whereas Lentz et al (2001) found r=5x104 ms1 to be appropriate for the North Carolina shelf Using nns bottom stress and depth-averaged current distributions we arrive at r estimations here (equation 6) that range across the shelf from 0.6-6x10-4 m s1 (middle panel of Fig. 2. 12). Thus the use of constant values for r while appealing for linear model calculations, may either overestimate or underestimate the bottom stress on continental shelves. Value judgements about quadratic versus linear drag laws, aside, we simply emphasize that bottom friction parameteriz a tion for models r e mains a compl e x, empirical issue. Boundary layer theory identifies different layers [e g Soulsby 1987] each with different behaviors. The "bed lay e r" immediately adjacent to the bottom is where mol e cular vi sc o s ity v control s th e dynamic s Within this lay e r (of a few centim e ters thickness) we have dU 2 pv-= T = -r(O) = pu dz u u.z (2.8) or -=-u. v

PAGE 78

27 26 25 27 26 25 Friction Velocity [cm/s] (': v -90 -89 -88 -87 -86 -85 -84 -83 -82 -81 -90 -89 -88 -87 -66 -85 -84 -83 -82 -81 Fig. 2.12. The spatial distribution of mixing properties inferred from tidal currents. From the top to bottom are the friction velocity (u*), the resistance coefficient (r), and the log-layer thickness. Units are provided in each panel, as are bathymetric contours (light lines) for the 20m, 50 m, 100m, 200m, 1000 m and 2000 m isobaths. 46

PAGE 79

47 Above the bed layer is the "logarithmic layer" within which neither the details of the bed nor the free-stream flow affect the local dynamics, and for which the velocity profile follows the universal form, u 1 z -=-ln-u. k z0 where k == 0.4 is the von Karman constant. Above the logarithmic layer is the "outer layer" for which the velocity and turbulence profiles depend on the nature of flow and hence obviates universality. Defining an eddy viscosity A [e.g., Wimbush and Munk, 1970] as, dU T=p(A+v)-, dz (2.9) (2.10) noting that A>> v where the constant stress and logarithmic regions overlap, and using equations (8) and (10), we arrive at A= cu.z (2.11) By similarity we can use this specification for A to estimate the Ekman layer thickness, 8, for steady state flow regimes, i.e., a au u fU =-(A-)=cu.az az 8 s: u. u=c-f (2.12) where f is the local Corio lis parameter and c=0.1 -0.4 [e.g., Loder and Greenberg, 1986, Weatherly and Martin, 1978]. Oscillatory boundary layers require further analysis if the local acceleration becomes of equal or greater importance as the Coriolis acceleration. Introducing the complex current, W = u + iv, into the momentum equations:

PAGE 80

48 au a av and -+ fu =-(A-) at az az (2.13) and assuming W oc e;ar where a is the frequency of oscillation, we arrive at (in analogy to equation 12): a aw w z(a + f)W =-(A) oc cu.az az 8 => 181= cu. (a+ f) (2.14) For the case of the M2 tide-induced bottom boundary layer on the WFS, with aM2 :::: 2f, the bottom boundary layer thickness may be estimated as 8 = cu. Using a mid-3/ range value of c=0 2 the bottom boundary thickness induced by the M2 tidal current over most of the WFS is estimated at about 2 5 m (except near Cedar Key and in Florida Bay where it is comparable to the water depth) Therefore, tidal mixing alone is insufficient to mix the water column on the WFS The height of the logarithmic layer is taken as some fraction of the Ekman depth. Soulsby (1987), for instance, defines the logarithmic layer height as 0.18 and with this definition the lower panel of Fig. 2.12 shows the M2 tidal current induced logarithmic layer height for the WFS. Values are generally less than 1 m. Empirical determinations of the log layer based on the in-situ bottom boundary layer measurement on the WFS are presently in progress and these will be reported on separately. 2.7. SUMMARY Using in situ data and a numeri cal circulation model we examine the structure of the WFS tides and the e ffects of the tides in di s tributing wat e r prop e rti e s and in mi x in g

PAGE 81

49 the water column vertically. Attention is limited to barotropic tides and to the four major tidal constituents (M2, S2, 01 and K1 ) that account for the bulk of the tidal variance (-90%) of the region. The data are from coastal sea level stations ranging from the Mississippi River delta to the Aorida Keys, and from current meter moorings deployed across the shelf between the 10 m to 300 m isobaths. The model is a regional adaptation of the POM extending from west of the Mississippi River to the Aorida Keys with a single open boundary arcing between these land termini. Regional tides are produced by forcing the model at its open boundary using the composite M2, S2 01 and K1 sea level variability from the global (TOPEX/Poseidon assimilated) tide model of Tierney et al. (2000). With standard tidal analysis tools, we produce co-amplitude and co-phase maps for each constituent and compare both sea level and currents against actual observations. Based on the fidelity of comparison we then use the model products to discuss the WFS barotropic tides. Apalachicola Bay, in the Aorida Panhandle, is a dividing point between appreciable semi-diurnal tides to the east over the WFS and minimal semi-diurnal tides from there to the Mississippi River delta. In contrast with the semi-diurnal tides, the diurnal tides are spatially more uniform. With respect to overall rms sea level variability, tides account for about 30% west of, versus about 50% southeast of Apalachicola Bay, respectively. The balance of the rms variability is due to synoptic scale weather-induced sea level change. Co-amplitude maps show largest tides in the regions where the shelf is widest, i.e., in the Aorida Big Bend and in Aorida Bay. The region just south of Tampa Bay shows a relative minimum. Both the amplification of the semi-diurnal constituents and the spatial variations with shelf width are consistent with linear theory [e.g. Clarke,

PAGE 82

50 1991 and Lentz, 2001]. Such theory also accounts for the relative spatial uniformity of the diurnal constituents. We conclude that the spatial distributions of the tides on the WFS are the result of local geometry, as contrasted with previous ideas about basin-wide tidal resonance. The tidal currents on the WFS, especially the semi-diurnal constituents, are primarily barotropic. Very good agreements are obtained between the in-situ data derived and the modeled tidal hodograph ellipses. Semi-major axes are generally less than 0.1 m s1 and the ellipses are generally oriented normal to the shoreline. Ellipse orientation, however, does reflect the co-amplitude distributions with tidal currents tending to diverge (converge) on regions of relative minima (maxima). A relatively small seasonal modulation is observed in the semi-diurnal tidal currents due to either internal tides or reductions in friction by stratification. Larger seasonal modulation is observed in the diurnal constituents due to the generation of inertial oscillations under stratified conditions. Inertial oscillations, when present, are generally organized as first baroclinic modes so they tend to cancel out when performing vertical averages. This helps to facilitate the barotropic tidal analyses. Baroclinic tides and inertial oscillations will be reported on separately. The non-linearity of the model allows it to generate mean currents despite its linear harmonic forcing. We present these residual currents in both Eulerian and Lagrangian frameworks to assess their potential effects on tidal residual material transports. While a coherent Stokes-drift pattern toward the northwest emerges, the magnitudes are small (one to two orders of magnitude less than the semi-major axis speeds) amounting to net particle trajectories of generally less than 10 krn per month.

PAGE 83

51 Albeit persistent, this is smaller than the potential particle trajectories for either seasonal or synoptic scale forcing [e.g., Weisberg et al., 1996]. With the M2 tide having the largest tidal current magnitude we use this constituent to examine the spatial distribution of bottom stress by tides on the WFS and the role that this may play in water column mixing. Except for the shallow regions of the Florida Big Bend and Florida Bay, the potential for mixing by tides is weak. Attempting to parameterize friction in a linear model using a resistance coefficient is limited by the spatial variability in bottom stress; nevertheless, the values estimated are within the range of values used in other studies. Estimates of logarithmic layer thickness by the M2 tide are generally less than 1 m and empirical studies are under way to quantify this better. In closing, with quantifiable metrics we show that regional tides on the WFS may be adequately described by forcing a regional, primitive equation, three-dimensional model with a global tide model at the open boundary, and that the properties of the tidal variations result primarily from the local geometry.

PAGE 84

52 REFERENCES Battisti, D.S. and A. J. Clarke, (1982a), A simple method for estimating barotropic tidal currents on continental margins with specific application to the M2 tide off the Atlantic and Pacific coasts of the United States. J. Phys. Oceanogr. 12, 8-16 Battisti, D. S. and A. J. Clarke, (1982b), Estimation of nearshore tidal currents on nonsmooth continental shelves. J. Geophys. Res., 87, 7873-7878. Blumberg, A. F., and G. L. Mellor (1987), A description of a three-dimensional coastal ocean circulation model, Three-Dimensional Coastal Ocean Models, Vol. 4, N. Heaps (ed.), 208-233, AGU, Washington, D. C. Clarke, A. J. (1991), The dynamics of barotropic tides over the continental shelf and slope (Review). In: Tidal hydrogynamics. B. Parker,editor, Johm Wiley & Sons, Inc Foreman, M.G.G., (1978), Manual for tidal currents analysis and prediction. Pacific Marine Science Report 78-6, Institute of Ocean Science, Patricia Bay, Sidney, BC, 70 pp Forman, M.G.G., Henry, R. F. (1979), Tidal analysis based on high and low water observations. Pacific Marine Science Report 79-15, Institute of Ocean Science, Patricia Bay, Sidney, BC, 39 pp Godin, G., (1972), The analysis of tides, University ofToronto Press, Toronto, Ont. Canada, 264,pp Grace, S.F. (1932), The principle diurnal constituent of tidal motion in the Gulf of Mexico. Mon. Not. R. Astr. Soc. Geophys. Suppl., 3(2):70-83 He, R. and R. H. Weisberg (2002), West Florida shelf circulation and temperature budget for the spring transition, 1999. Cont. Shelf Res., 22, 5, 719-748 Kajiura, Kinjiro. (1958), Effect of Coriolis force on edge waves (II) specific examples of free and forced waves. Journal ofMarine Research, 16(2), 145-157 Koblinsky, C. J. (1981), The M2 tide on the West Florida Shelf. Deep Sea Res. Vol. 28A, 1517-1532

PAGE 85

Lentz, S.M. Carr, and T.H.C Herbers (2001), Barotropic tides on the north Caroline shelf. J. Phys. Oceanogr. 31 (7), 1843-1859 53 Loder, J. W. and D. A. Greenberg (1986), Predicted positions of tidal fronts in the Gulf of Maine region. Continental Shelf Res., 6, 394-414, 1986 Mannarino, G. 0. (1983), Variability of current, temperature, and bottom pressure across the west Florida continental shelf, witner 1981-1982, J. Geophys. Res. Vol 88, C7, 4439-4457. Mellor, G. L., and T. Yamada (1982), Development of a turbulence closure model for geophysical fluid problems, Rev. Geophys., 20, 851-875 Reid. R. 0. and R.E. Whitaker (1981), Numerical model for astronomical tides in the Gulf of Mexico. Texas A&M report for U.S. Army Engineers Waterway Experiment Station, pp 115 Smagorinsky,J. (1963), General circulation experiments with primitive equations. I. The basic Experiment. Mon. Weather Rev., 91, 99-164, 1963 Soulsby, R. L (1983). The bottom boundary layer of shelf seas. In Physical Oceanography of Coastal and shelf seas. Oceanogr. Ser. Vol., 35, edited by B. Johns, pp 189-266, Elsevier Sci., New York, 1983 Tierney, C.C, L.H. Kantha and G.H. Born (2000), Shallow and deep water global ocean tides from altimetry and numerical modeling. J. Geophys. Res. 105, 11259-11277 Weatherly, G. L. and P. Martin (1978), On the structure and dynamics of the oceanic bottom boundary layer. J. Phys. Oceanogr. 8, 557-570 Weisberg, R.H. B. D. Black, H. Yang (1996), Seasonal modulation of the West Florida continental shelf circulation. Geophys. Res. Lett., 23, 2247-2250 Weisberg, R. H. and Z. Li and F. Muller-Karger (2001), West Florida shelf response to local wind forcing, April1998. J.Geophy. Res., 106 (C12), 31239-31262 Westerink, J, J, R. A. Luettich, N. Scheffner (1993), ADCIRC: an advanced three dimensional circulation model for shelves, coast, and estuaries. Report 3: Development of a tidal constituent database for the western North Atlantic and Gulf of Mexico. Technical report DRP-92-6, U.S. Army Corps of Engineers. Wimbush, M and W. H. Munk (1970), The benthic boundary layer. The Seas, 4, 731-758

PAGE 86

54 Yang, H, R. H. Weisberg, P. P. Niiler, W. Sturges, W. Johnson (1999), Lagrangian circulation and forbidden zone on the West Florida Shelf. Continental shelf Res. 19. 1221-1245 Zetlzer, B. D. and D. V. Hansen (1971), Tides in the Gulf of Mexico. In: Contributions on the physical oceanography of the Gulf of Mexico. Vol. 2, L. R. A. Capurro and J. L. Reid, editors, Gulf Publishing Company, pp 265-275

PAGE 87

55 CHAPTER3 A LOOP CURRENT INTRUSION CASE STUDY ON THE WEST FLORIDA SHELF 3 1.ABSTRACT The Gulf of Mexico Loop Current intruded upon the West Florida Continental Shelf in June 2000. In-situ currents and hydrography along with satellite temperature and altimetry measurements are used to describe this event and its effects on the shelf. A strong southward current is observed to flow along the shelf slope seaward of the intruded water boundary. This current transported cold, nutrient rich water from the north, thereby producing anomalous hydrographic features near the shelf break (80m isobath). An array of moored velocity profilers reveals that the currents landward of the intruded water are independent of the Loop Current and primarily driven by local winds. These observational findings are discussed relative to a series of idealized numerical model simulations inclusive of forcing by both the Loop Current and local winds. 3.2. INTRODUCTION The West Florida Continental Shelf (WFS) is broad and gently sloping with its lOOm isobath situated some 150-200 km offshore. The WFS circulation, driven by tides, winds, and buoyancy fluxes, is also influenced by the Gulf of Mexico Loop Current (LC) that enters through the Yucatan Strait and exits as the Gulf Stream though the Straits of

PAGE 88

56 Florida. Tidal currents are relatively weak (e g. Koblinsky, 1981; He and Weisberg 2002a), and sub-tidal sea level and current variations are correlated with synoptic scale wind variations (e.g., Niller, 1976 ; Mitchum and Sturges, 1986; Cragg et al., 1983; Marrnorino, 1983 ; Weisberg et al. 2001). What remains unclear is how deep-ocean and shelf waters interact and how such interactions influence the shelf circulation and water properties. With the important issue of LC effects on the WFS unresolved, the present paper provides a case study of a LC impact event for which in-situ currents and hydrographic data exist across the entire shelf We describe the impact event using satellite sea surface temperature (SST) and sea surface height (SSH) information together with in-situ data. We then perform numerical model experiments [using the primitive equation, Princeton Ocean Model (POM) of Blumberg and Mellor 1987] under idealized forcing by the LC and by local winds to reconcile the observed features. Section 2 describes the relevant data of June 2000. Section 3 presents the idealized LC model and the experimental results, with and without wind forcing. Section 4 then summarizes and discusses the implications of our findings. 3.3.DATA Intrusions of the L C into the eastern Gulf of Mexico and onto the WFS are th e topic of several papers, but each with limited data sets [e.g., Leipper 1970; Niller 1976; Molinari e t al 1977 ; Behringer et al 1977, Huh e t a l 1981 Paluszkiewicz et al1983]. As part o f an Ecology of Harmful Algal Blooms (EC OHAB) region a l field study we encountered a LC intrusion onto the central portion of the WFS in June

PAGE 89

2000. This section describes the intrusion using shipboard hydrographic data, satellite remotely sensed SST and SSH products, and water column velocity data from moored ADCPs. 3.3.1. HYDROGRAPHIC DATA 57 The multi-disciplinary Florida ECOHAB project conducted monthly hydrographic surveys on the central WFS over the approximate three-year period beginning in summer 1998. The cruise track included transects offshore of Sarasota Florida that extended across the shelf between the 10m and the 200m isobaths (Figure 3.1). CTD casts along with biological and chemical measurements were taken every 5 nm. Two such hydrographic cruises were performed in June 2000 (June 6-8 and June 2830), with a total 57 CTD casts for each cruise; 22 along the Sarasota transect. Figure 3.2 shows the across-shelf structures of the temperature, salinity, density, and chlorophyll fluorescence fields measured along the Sarasota transect on June 6th and June 28th. Steeply sloping isotherms are observed seaward of the shelf-break suggestive of a strong southward baroclinic current. Straddling the shelf-break at the bottom is a mass of relatively cold and fresh water upwelled from deeper depths. Water of this temperature and salinity occupies the monotonic, positive slope portion of the TIS curve found between Antarctic Intermediate Water of southern and 18 degree water of northern hemisphere origins (e.g., Schmitz, 1991), and is characteristic of mid-depth waters found within the LC. While a consequence of upwelling, this water did not upwell locally, as evidenced in the lack of connectivity along the Sarasota transect between water properties at the shelf-break and those farther offshore. We surmise that upwelling occurred farther

PAGE 90

27. 5 27 26. 5 26 25. 5 -85 \ \ \ \ \ \ \ ._ l \ \ 8 \ I r-' .. _.ROrtda Keys CM2CMP2 0 0 ... CM4 c\. . 4 se EC:f : .a.. e.' 8 0 -l It) U' C\1 >:-84.5 -84 -83. 5 CM3 0 "' q -83 Longitude . 1 '0 \ ,:f!'? . '\1 '; : 1 .. o" .\ -82. 5 -82 -81 5 Figure 3 .1. West Florida shelf geometry, locations of the hydrographic casts (denoted by dotes), and the moored ADCP current measurements(triangles denote upward looking and circles denote downward looking ADCPs ). 58

PAGE 91

59 north and that the upwelled water was advected by the shelf-break currents. How this may have occurred will be discussed in section 4. The appearance of such anomalously cold water at the shelf-break in June 2000 is also unique to the monthly ECOHAB cruise data. All other Sarasota hydrographic sections from June 1998 to September 2001 show that waters of 16 degrees or colder were below 170m, whereas here it is as shallow as 70m. Chlorophyll fluorescence is also interesting. Qualitatively, we see high fluorescence originating at the shelfbreak in the vicinity of the upwelled water and extending inshore along the bottom. Chlorophyll fluorescence requires two ingredients: nutrients and light. The upwelled water provides the nutrients, and the shallow depths provide for the light. Nutrient concentrations may also be elevated near-shore due to land drainage through the Tampa Bay and Charlotte Harbor estuaries. These two sources of nutrients (shelf-break and near-shore), both with available light, are connected through the bottom Ekman layer. Thus, and especially under stratified conditions (e.g., Weisberg et al., 2001), the bottom Ekman layer provides an effective across-shelf conduit for the delivery of biologically important materials. As will be shown later, the inner-shelf circulation in June 2000 was primarily of a wind-induced downwelling type. This is reflected in the changes of the temperature, density, and fluorescence isolines between the June 6th and June 28th transects, attesting to the bottom Ekman layer playing an essential role in the WFS biological productivity. Using the thermal wind relationship, and assuming a reference level of zero baroclinic geostrophic current at the bottom, we compute the along-shelf velocity distributions for the June 6th and June 28th transects (Figure 3.3). Corresponding to the

PAGE 92

0 -20 -40 I -60 .c -80 a -100 -120 Temperature fC) -140 -160 June 6, 2000 -20 -40 -60 I -80 .c a -100 -120 -140 Salinity (PSU) -160 June 6, 2000 0 0 50 100 -20 -40 -60 I -80 c. 100 -120 SigmaTheta(kg/m3 ) -140 -160 June 6, 2000 -20 -40 -60 I -80 .c a -100 -120 140 Fluorescence (ug/1) -160 June 6, 2000 0 50 100 Offshore Distanc e [km] -40 -60 -80 -100 -120 -140 -160 -40 -60 -80 -100 -120 -140 -160 -20 -40 -60 -80 -100 -120 -140 -160 -20 -40 -60 -80 -100 -120 140 160 150 0 Temperature fC) June 28, 2000 Salinity (PSU) June 28,2000 50 SigmaTheta(kglm3 ) June 28,2000 100 Fluorescence (ug/1) June 28,2000 50 100 Offshor e Distance [km] 150 150 Fig. 3.2. Across-shelf (Sarasota transect) distributions of temperatur e, salinity, density and chlorophyll fluorescenc e sampled on June 6th (the left panels) and June 28th (the right panels). 60

PAGE 93

0 -20 -40 -60 E .I:: -80 a. 8 -100 -120 -140 -160 0 -20 -40 -60 E .I:: -80 a. 8 100 120 -140 -160 0 0 Geostrophic Current (cm/s) June 6, 2000 50 100 150 Geostrophic Current (cm/s) June 28, 2000 50 100 150 Offshore Distance [km) Figure 3.3. Calculated across-shelf (along the Sarasota transect) geostrophic currents (assuming zero at bottom) on June 6'h (the upper panel) and June 28th (the lower panel). Southward currents, being negative, are denoted by solid lines. The Contour interval is 5 cm/s. 61

PAGE 94

62 largest across-shelf density gradients, the strongest southward flows are observed near the shelf-break with magnitudes of about 0.70 ms-1 on June 6th and 0.45 ms-1 on June 28th. As anticipated from the Taylor-Proudman theorem (e.g., Brink, 1998), these LC-related southward currents diminish in magnitude to zero within 10-20 km of the shelf-break, and consistent with the baroclinic Rossby radius of deformation (NH/f) calculated at the current core, the scale width of these primarily baroclinic currents are about 30-35 km, as observed. As a result, the geostrophically inferred currents on the shelf are comparatively weak, with speeds generally less than 0.05 ms-1 3.3 2. SATELLITE DATA Complimenting the shipboard hydrographic measurements are satellite-derived images of SST (by A VHRR) and SSH (by TOPEX/ERS-2 altimetry). These provide the larger scale picture of the environmental conditions sampled along the ship track. Our analysis period, chosen to bracket the ship surveys, is June 151 to July 6th. The composite SST images were obtained from the Applied Physics Laboratory, Johns Hopkins University; and the SSH images were obtained from the Astrodynamic Research Group, University of Colorado. These are shown in Figures 3.4, 3.5, and 3.6. Each panel includes the 75 m and 200m isobaths to indicate the position of the LC relative to the WFS Figure 3.4 is the June 6th SST image. A well-defined frontal feature is observed south of 28N as a band of relatively cold water that loops around and strikes the WFS between the 200m and 75 m isobaths. As evidenced in the shipboard hydrography, this is where the thermocline intersects the surface at the outer edge of the LC.

PAGE 95

Fig. 3.4 June 6th 2000 AVHRR SST obtained (with permission) from the Applied Phys ics Laboratory at Johns Hopkins University. The 75 m and 200 m isobath contours are overlaid to show the LC front penetration onto the WFS. 63

PAGE 96

. : ;J .... i i 2 :i .. .. i i Fig. 3.5. Time series of AVHRR SST maps from June 1st to July 61 h 2000 obtained (with permission) from the Applied Physics Laboratory at Johns Hopkins University. The 75 m and 200 m isobath contours are overlaid to show the LC front penetration onto the WFS. 64

PAGE 97

65 How these features evolve over the case study period June 151 to July 61h is shown for SST and SSH in Figures 3.5 and3. 6, respectively. The cold, outer edge of the LC penetrating the shelf to between the 200m and 75 m isobaths is clearly evident as an anticyclonically wrapping filament from June 151 to June 8th, after which the feature becomes more amorphous as SST homogenizes with summertime surface heating. The ECOHAB hydrography shows this filament to be located at the shelf-break (approximately at the 80m isobath on this transect) on June 6th (Figure 3.2) where the 28"C isotherm broaches the surface. The SSH gradient features support these SST descriptions. A gradient in SSH is evident on the shelf through around June 8th, after which the leading edge of the LC retreats offshore and to the south. On around June 22nd the leading edge of the LC again sidles toward the northeast and impacts the shelf-break. By early July there is evidence of an across-shelf directed component to the LC flow at the continental slope, looping around anticyclonically to become an along-shelf flow in the vicinity of the shelf-break. 3.3.3. CURRENT DATA An array of ADCP moorings on the WFS (Figure 3.1) provides in-situ current measurements for the case study period. The array spanned the entire shelf with moorings at the 150m (EC1) 75 m (CM4), 50 m (CM2), 30m (EC3), 25 m (NA2), 20m (EC4), and 10m (EC5 and EC6) isobaths. The outermost moorings (150m and 75 m isobaths) were deployed on June 25th so these coincide only with the later part of the case study. After low-pass filtering to remove tidal and inertial oscillations and other fluctuations at times scales shorter than 36 hrs, daily velocity vector averages are shown

PAGE 98

l:' : <'!. y, ,?:f .. . r [\1 -<:.-,! .... :till-_ :... ---" JJ -> :7'1 .. _._ -I!'M S/.t;o; .. 11':11 H .. . li-!1.1 U o:
PAGE 99

67 for the near-surface and near-bottom bins in Figures 3.7 and 8, respectively. The location of these bins varies with the instrument depth and mooring type (buoyed by surface or subsurface floatation or mounted at the bottom), and at each location we show velocity data from the (uncontaminated) bins closest to either the surface or the bottom. Also shown are coastal wind velocity vectors from Venice, FL. similarly low-pass filtered and then daily averaged. Note the different scales used for each of the vector presentations. In particular the bold vector used to distinguish the 150m isobath currents from the others is largely reduced in dimension since these currents are much stronger. Thus, after July 4th, when the vectors at the 75 m isobath appear larger than those at the 150m isobath, the velocity disparity is actually about three times less than it looks. Similarly, the near-bottom currents are accentuated by a factor of two relative to the near-surface currents since the near-bottom currents tend to be relatively smaller. Markedly different behaviors are seen between the currents observed on the inner-shelf versus those observed near the shelf-break. Over the inner-shelf (e.g., see Lentz, 1994 and Weisberg et al., 2001 for related but different definitions), taken here to be roughly inshore of the 50 m isobath, the current variations are determined primarily by the local winds. This is not the case at the 150m and 75 m isobaths where the LC intrusion is most influential. On June 26th we see a substantial along-shelf directed current at the 150m isobath in isolation from any of the locations farther inshore, including the 75 m isobath. Currents at the 75 m isobath begin to flow along-shelf with those at the 150m isobath around July 3rd, and they peak in magnitude on July 7th, after which the extended record shows that they diminish again. The largest near-surface and near-bottom currents during this time interval occur on July 3rd at the 150m isobath at about 1.0 ms-1 and 0.4 ms-1

PAGE 100

\ \ \ \ .. ,_ '... '1. \ \ \ .. ... \ ....\ ... I \ \ .,. \ ' i '. \ \ .. \ ' : \ I Figure 3.7. \ \ .. t \ ; 1 i 100c1: )Ia : \ ., , .. .\. .,\ ,.. i '-, \ ; ... \-\ Time series of near subsurface currents observed on the WFS from June 1st and July 6th 2000. Thick darkk arrows denote currents at 150 m isobath; thin dark arrows denote currents at all other isobaths; and gray arrows denote the coastal winds at Venice, FL. The depth contours correspond to the 20m, 50 m, 75 m and 150m isobaths.75 m. 0\ 00

PAGE 101

\ \ "" ., ; Figure 3.8. \ 50cm/)., 15crnls > \ \ l, \, \ ,., , \ '1 \ \ '\ Time series of near bottom currents observed on the WFS from June 1st and July 6th 2000. Thick darkk arrows denote currents at 150 m isobath; thin dark arrows denote currents at all other isobaths ; and gray arro ws d enote the coastal winds at Venice, FL. The depth contours correspond to the 20m, 50 m, 75 m and 150m isobaths.75 m 0'1 \0

PAGE 102

70 respectively. On July 7th the near-surface and near-bottom currents at the 75 m isobath are about 0.3 ms-1 and 0.25 ms-1 respectively. Current measurements at both of these locations extend through September 2001. Typical speeds are about 0.1-0.2 ms-1 Thus, the strong current event reported here is a unique for this 15-month-long record. From these observations and those previously reported by Meyers et al. (2001), it appears that LC impact events are infrequent and when they occur they affect the currents near the shelf-break, leaving the currents over the inner-shelf largely driven by local forcing (winds and buoyancy fluxes). Compatible with the geostrophic assumption made for Figure 3.3 we may also analyze the ADCP velocity profile to estimate the barotropic component to the total flow. Thus, we set the Corio lis force equal to the sum of the barotropic and the baroclinic contributions to the pressure gradient force, v = vbarorropic + vbaroclinic = vbarorropic +fig aa f p(z)dz Po X (3.1) with the vertical shear of the observed along-shelf current equaling the vertical shear of the baroclinic current calculated by thermal wind. Figure 3.9 compares the ADCPobserved and the thermal wind-calculated along-shelf velocity component profiles for both the 75 m and the 150m isobaths on June 28th. The vertical shears tend to agree over the ranges for which data are available. The implied mean barotropic portions of the flows at the 150m and the 75 m isobaths are about 0.25 ms-1 and 0.1 ms-1 respectively. The barotropic portions of the flows at both locations are directed southward, whereas the baroclinic portions are reversed (southward at the 150 m isobath and northward at 75m

PAGE 103

. ...... .... ... ' ........ . . . ObserVed : :calculated : Difference -10 bHr:ererice : observed : c!lli:u!3teil . -30 .... ... . ....... ... -50 . : \ ... : : -20 .... : . . -30 :[ -70 .s:; ii Q) 0 -40 -90 -50 -110 -60 .... = ..... : ..... : .. -130 ... -: ... :... : . CM4: -70 ..... : ..... : ..... : ..... : ..... : .... ..... -150 -80 -70 -60 -50 -40 -30 -20 -10 0 -15 -10 -5 0 5 10 15 20 [an's) [an's) Fig. 3.9. The June 281 h depth profiles of the ADCP observed currents (thick light lines), the thermal wind calculated baroclinic geostrophic currents (thick dark lines), and their difference (thin dark lines) at the 150 rn (left hand panel) and the 75 rn (right hand panel) isobaths. 71

PAGE 104

72 isobath). So while the barotropic portion of the shelf response to this LC intrusion extends across the shelf-break, a node is crossed for the baroclinic portion due to the doming of the isopycnals at the shelf-break. This counteracting effect of the barotropic and baroclinic responses further limits the penetration of the LC intrusion onto the shelf. If we further assume a landward exponential decay (Chapman and Brink, 1987) for the barotropic pressure gradient (or geostrophic current) this implies a reduction of the LC induced barotropic current to 0.05 ms-1 by the 50 m isobath, the magnitude of which is easily exceeded by the local wind-driven circulation. Thus, neither the baroclinic, nor the barotropic parts of the LC-induced circulation should be major factors on the inner-shelf, consistent with Figures 3.7 and 3.8. 3.4. NUMERICAL SIMULATIONS The observations suggest that the inner-shelf and shelf-break regions act independently of one another as the LC impacts the region of the shelf break. The LC controls the flow field in the vicinity of the shelf-break when it is nearby, whereas local forcing controls the flow field over the inner-shelf. We attempt to further understand these observations by performing numerical model experiments under realistic geometry, but with idealized forcing. The model is a regional adaptation of the primitive equation, POM (Blumberg and Mellor, 1987), the details of which are given in He and Weisberg (2002b ). Our WFS adaptation extends from west of the Mississippi River in the northwest to the Florida Keys in the southeast with a single open boundary arching between these end points. We explore the role of the LC on the WFS by controlling the flows into and out of the open boundary while allowing the model LC to freely evolve

PAGE 105

within the model domain. Three types of experiments are performed, all with constant density to emphasize the barotropic response since previous studies (e.g., Chapman and Brink, 1987) show that it penetrates farther inshore than the baroclinic response. The three experiments are : 1) forcing by the LC only, 2) forcing by the LC, plus upwelling favorable winds, and 3) forcing by the LC, plus downwelling favorable winds. We generate a LC by diagnostically imposing a sea level distribution along the open boundary: (3.2) 73 where AO is the amplitude of the sea level perturbation, x is grid index along the open boundary, and xo is the grid index corresponding to the location of maximum sea level. Based on a geostrophic balance, the open boundary sea level distribution corresponds to regions of inflow in the north and outflow in the south. From the satellite altimetry analysis we set A0=0.4 m which corresponds to a geostrophic current magnitude of about 0.5 ms-1 Such current is a little smaller than observed at the shelf break but considerably larger than the barotropic portion in Figure 3.9. These open boundary sea level and geostrophic current distributions are given in Figure 3.10. When wind stress is added we use a uniform value of 0.1 Nm-2 directed along-shore (with respect to the west Aorida coastline) either to the southeast (upwelling) or to the northwest (downwelling). Each experiment consists of a 25-day model run in which a quasi-steady state is reached within about 5 days. Model results in the form of depth-averaged current fields are given in Figure 3 .11. The thick line along the open boundary corresponds to the region of sea level control, and the resulting model-determined LC inflows and outflows are clearly

PAGE 106

74 depicted. Panel A is with the LC only. Penetration of the LC into the model domain is greatly inhibited by the sloping topography. Currents intrude onto the shelf up to about the 200m to 100m isobaths, after which the vectors are imperceptibly small. Because of the cyclonic propagation of topographically trapped waves the region along the shelf break to the north of the LC impact is also set into motion as a shelf-break jet that extends to the northwest comer of the model domain. Where the shelf narrows at the DeSoto Canyon in the north, this jet excites currents in shallower water consistent with the findings of Chapman and Brink (1987). Regardless of shelf width, however, wind forcing dominates the inner-shelf responses. Panels B and C show the responses to the LC, plus either upwelling or downwelling favorable winds, respectively. A more detailed analysis of the wind-driven responses alone under a constant density setting is given by Li and Weisberg (1999a,b). The main point here is that given the mixture of the two forcing functions (LC and winds), the LC controls the shelf slope and break regions, whereas the winds control the inner-shelf region, as confirmed by our observations in Figures 3.7 and 3.8. 3.5. SUMMARY AND DISCUSSION Using in-situ currents and hydrographic data, along with satellite SST and SSH analyses, we describe a LC intrusion event that occurred on the WFS in June 2000, and we explored the relative roles of LC-induced and local wind-induced currents in accounting for the observations through idealized numerical model experiments. Consistent with the Taylor-Proudman theorem, the planetary vorticity constraint by the sloping bottom limits the barotropic portion of the LC penetration onto the shelf. Since

PAGE 107

0 4 0 3 0 2 i 0 1 -0 .c g -0. 1 gj -0. 2 -0. 3 -0. 4 "'
PAGE 108

30 29 28 27 26 25 A _,... -90 -88 -86 -84 -82 30 29 28 ll 27 () 26 25 8 24 -90 -88 -86 -84 -82 30 29 28 27 26 25 c 24 -90 -88 -86 -84 -82 Fig 3.11. Depth-averaged current maps from the idealized LC and wind-forced model experiments Panel A is for the LC only case. Panel B is for the LC plus upwelling favorable (0.1 Nm2 ) winds case. Panel C is for the LC plus downwelling favorable (0 1 Nm-2 ) winds case. Thick lines along the open boundary indicate the sea surface height perturbation region. The depth contours correspond to the 2 0 m, 50 m 100 m, 200 m, 1000 m, and 2000 m isobaths 76

PAGE 109

77 the LC is largely baroclinic, the across-shelf scale of the resulting current is compatible with the baroclinic Ross by radius of deformation evaluated near the shelf break. Penetration onto the shelf is further inhibited by the counteracting affects of the baroclinic and barotropic portions of the flow field. The barotropic portion, while of larger across-shelf scale, is counteracted by the baroclinic portion that reverses sign due to doming of isopycnals at the shelf break. Thus, the shelf-break currents are largely LC controlled, whereas the inner-shelf currents are largely controlled by the local winds. Numerical experiments, using a regional adaptation of the POM with realistic topography and idealized barotropic, constant density forcing, supports the conclusion on the near independence between the LC-driven shelf-break and the wind-driven inner shelf for this mid-point LC-shelf interaction. A barotropic LC impact on the shelf slope is confined seaward of the shelf-break setting up a shelf slope current to the north (in the cyclonic direction of continental shelf wave propagation). Adding either upwelling or downwelling favorable winds demonstrates that the wind-driven responses greatly exceed the LC responses over the inner-shelf. LC intrusions onto the WFS in reality have both barotropic and baroclinic parts. The isopycnals associated with the shelf slope current to the north of the impact region therefore tilt upward toward the shelf causing relatively deep-water to upwell at the shelf break. If the shelf-break either shoals, or the shelf narrows, in the cyclonic direction, then water upwelled in the shelf slope current may reach isobaths that are shallower than those that can be attained locally where the LC impacts the shelf. The WFS geometry provides both of these factors. The shelf break at the impact region is about 80 m deep, whereas in the DeSoto Canyon it is about 40 m deep. The shelf also narrows appreciably

PAGE 110

78 between Cape San Blas and DeSoto Canyon (Figure 3.1). The addition of wind-driven upwelling over the narrower portions of the shelf can further facilitate upwelling onto the shelf. Note that for the period June 3-8, at the beginning of this LC intrusion event, the winds were upwelling favorable. Thus, the ingredients (a LC-induced shelf slope jet with currents of magnitude 0.4 ms 1 upstream topography changes, and upwelling favorable wind forcing) existed to account for the observed deep-water properties being upwelled and transported from the north.

PAGE 111

REFERENCES Behringer, D.W., R.L. Molinari, and J.F. Festa (1997), The variability of anticyclone current patterns in the Gulf of Mexico, J. Geophys. Res, 82, 5479-5488 Blumberg A. F. and G. L. Mellor (1987), A description of a three-dimensional coastal ocean circulation model, Three-dimensional Coastal Ocean Model, Vol. 4, N. Heaps (ed.), 208-233, AGU, Washington, D. C. Brink, K. H., Wind-driven currents over the continental shelf (1998), The Sea, Vol. 10, Chapter 1 Chapman, D. C. and K. H. Brink (1987), Shelf and slope circulation induced by fluctuating offshore forcing, J. Geophys. Res., 92, 11741-11759 Cragg, J., G. Mitchum and W. Sturges (1983), Wind-induced sea-surface slopes on the West Florida Shelf, J. Phys. Oceanogr., 13, 2201-2212 79 He, R., R. H. Weisberg (2002a), Tides on the West Florida Shelf, J. of Phys. Oceanogr., in press He, R., R. H. Weisberg (2002b), West Florida shelf circulation and temperature budget for the 1999 spring transition. Continental Shelf Research, 22,5,719-748 Huh, 0, W .J. Wiseman, and Lawrence Rouse (1981 ), Intrusion of loop current onto the west Florida continental shelf, J. Geophys. Res, 86,4186-4192 Koblinsky, C. J. (1981), The M2 tide on the West Florida Shelf. Deep Sea Research, 28A, 12, 1517-1532 Leipper, D.F. (1970), A sequence of current patterns in the Gulf of Mexico, J. Goephys. Res, 75, 637-657 Lentz, S. J. (1994 ), Current dynamics over the northern California inner shelf, J. Phys. Oceanogr., 24, 2461-2478 Li, Z., and R.H. Weisberg (1999a), West Florida continental shelf response to upwelling favorable wind forcing, part 1: kinematic description. J.Geophys. Res., 104, 1350713527

PAGE 112

80 Li, Z., and R.H. Weisberg (1999b), West Florida continental shelf response to upwelling favorable wind forcing, part II: dynamical analyses, J. Geophys. Res., 104, 2342723442 Marrnorino, G. 0. (1983), Wind-forced sea level variability along the West Florida Shelf, winter, 1978. J. Phys. Oceanogr. 12, 389-405 Meyers, S.D., E.M. Siegel, R.H. Weisberg (2001), Observation of currents on the West Florida shelf break. Geophys. Res. Lett. 28, 2037-2040 Mitchum G. and W. Sturges (1982), Wind-driven currents on the West Florida Shelf, J. Phys. Oceanogr. 12, 1210-1217 Molinari, R.L., S. Baig, D.W. Behringer, G.A. Maul, and R. Legeckis (1977), Winter intrusions of the loop current, Science, 198, 505-506 Niller, P. P. (1976), Observations of low-frequency currents on the West Florida continental shelf, Memoires Societe Royale des Sciences de Liege, 6, X, 331-358 Paluszkiewicz. K, L.P. Atkinson, E.S. Posmentier and C.R. McClain (1983), Observations of a loop current frontal eddy intrusion onto the west Florida shelf J. Goephys. Res. 88, 9639-9651 Schmitz, Jr. W. J., Richardson, P. L. (1991), On the source of the Florida Current. Deep Sea Research, 38 (Suppl. 1), S389-S409 Vukovitch, P.M., B.W. Crissman, M.Bushnell and W.J. King (1979), Some aspects of the oceanography of the Gulf of Mexico via satellite and in situ data, J.Geophy, Res, 84, 7749-7768 Weisberg, R H., Z. Li, F. MullerKarger (2001), West Florida shelf response to local wind forcing, April1998. J.Geophy. Res., 106 (C12) 31,239-31,262

PAGE 113

81 CHAPTER4 WEST FLORIDA SHELF CIRCULATION AND TEMPERATURE BUDGET FOR THE 1999 SPRING TRANSITION 4.1. ABSTRACT Mid-latitude continental shelves undergo a spring transition as the net surface heat flux changes from cooling to warming. Using in-situ data and a numerical circulation model we investigate the circulation and temperature budget on the West Aorida Continental Shelf (WFS) for the spring transition of 1999. The model is a regional adaptation of the primitive equation, Princeton Ocean Model forced by NCEP reanalysis wind and heat flux fields and by river inflows. Based on agreements between the modeled and observed fields we use the model to draw inferences on how the surface momentum and heat fluxes affect the seasonal and synoptic scale variability. We account for a strong southeastward current at mid-shelf by the baroclinic response to combined wind and buoyancy forcing, and we show how this local forcing leads to annually occurring cold and low salinity tongues. Through term-by-term analyses of the temperature budget we describe the WFS temperature evolution in spring. Heat flux largely controls the seasonal transition, whereas ocean circulation largely controls the synoptic scale variability. These two processes, however, are closely linked. Bottom topography and coastline geometry are important in generating regions of convergence

PAGE 114

and divergence. Rivers contribute to the local hydrography and are important ecologically. Along with upwelling, river inflows facilitate frontal aggregation of nutrients and the spring formation of a high concentration chlorophyll plume near the shelf break (the so-called 'Green River') coinciding with the cold, low salinity tongues. These features originate by local, shelf-wide forcing; the Loop Current is not an essential ingredient. 4.2. INTRODUCTION The west Florida continental shelf (WFS) is one of the broadest continental shelves in North America. Between the Florida Keys and the Florida Big Bend, the WFS isobaths vary smoothly, and generally parallel the coastline. This geometry changes along the Florida Panhandle in the north where the coastline undergoes a right angle bend, and the shelf width decreases to a minimum at the DeSoto Canyon. The WFS circulation, forced by tides, winds, buoyancy, and possible interactions with the Gulf of Mexico Loop Current, varies on time scales from semi-diurnal to inter annual (Weisberg et al., 1996). Monthly mean currents at mid-shelf suggest a seasonal cycle with along-shore flows to the southeast in spring, and to the northwest in late summer to early autumn. Weisberg et al. (1996) hypothesized that these seasonal currents are baroclinic based on an observed thermal wind shear and the seasonal reversal of the across-shelf density gradient. As a consequence of the spring transition in surface heat flux from cooling to warming, they argued that spatial differences in heating (from the coast to offshore by increasing depth and from the south to north by solar declination) form a mid-shelf cold tongue and a seasonally maximum across-shelf density gradient,

PAGE 115

83 that supports a southeastward current. Here we examine this locally forced, seasonal circulation hypothesis by focusing on the spring transition for 1999, a year when the Loop Current, as evidenced in relatively flat isopycnal topography at the shelf break, did not have a strong direct influence the WFS. Our objective is to describe the circulation and temperature budget for spring 1999 with respect to the shelf-wide winds, surface heat fluxes, and river inflows. An argument used to describe the transition from wintertime horizontal stratification to summertime vertical stratification for mid-latitude shelves is that decreased winds and increased solar heating conspire to form a thermocline. The details of this process are not well understood, however. Chapman and Gawarkiewicz (1993) reason that nonlinearity in the equation of state can account for the elimination of horizontal stratification by spatially uniform heating, but Morey (1999) pointed out that this argument is valid only for certain salinity and temperature ranges. Other processes must also be important. Morey (1999), using a two-dimensional model with a sloping bottom, argue that the surface heat flux divided by the water depth is the critical factor in the seasonal transition, essentially the differential heating argument advanced earlier. The degree to which his argument is valid in the fully three-dimensional sense and the regional partition between ocean dynamical and local heating affects are also topics of our paper. The observational record [e.g. Niiler, 1976; Mitchum and Sturges, 1982; Cragg et al, 1983; Marmorino, 1983; Mitchum and Clarke, 1986a] shows that the WFS circulation and sea level variations are highly correlated with the synoptic scale wind stress variations. The passage of cold fronts also affects the local temperature balance (e.g.,

PAGE 116

84 Price, 1976). Along with these local synoptic scale variations are baroclinic effects that originate with the Loop Current at the shelfbreak [e.g., Paluszkiewicz et al, 1983]. What remains unclear is the relative importance between the momentum and buoyancy that are input either locally, or at the shelf break. Such questions are of multi-disciplinary interest since, despite its oligotrophic description, the WFS supports productive ecosystems. These include episodic toxic dinoflagellate blooms (red tides) near the coast (Steidinger, 1981; Vargo et al, 1985), a seasonal chlorophyll plume near the shelfbreak (Gilbes et al, 1996), and important commercial and recreational fisheries throughout the WFS. An improved understanding of the circulation and how it affects seasonally varying water properties and influences organism growth and distribution are necessary. This paper focuses on local wind and buoyancy forcing during the spring transition of 1999, independent of the Loop Current. We use the primitive equation, Princeton Ocean Model (POM) described by Blumberg and Mellor (1987) forced by National Center for Environmental Predication (NCEP) reanalysis winds and net surface heat flux and by river inflows. The only role of the adjacent Gulf of Mexico in this study is to set the vertical distribution of temperature and salinity for initializing the model density field. Once begun, the integration proceeds solely on the basis of local forcing. By running twin model experiments, one with heat flux and the other without, we explore the relative importance of wind and buoyancy in affecting the seasonal and synoptic scale variability. Section 4.3 describes the model and forcing fields. Section 4.4. compares model results with in-situ observations. Based upon these comparisons the model is used in

PAGE 117

85 section 4.4 to describe the seasonal mean circulation on the WFS for spring 1999 and the evolution of the corresponding temperature and salinity fields. Section 4.5 presents a term-by-term analysis of the three-dimensional temperature budget. The results are summarized and discussed in section 4.6. 4.3. MODEL AND FORCING FIELD 4.3.1. MODEL We use the Princeton Ocean Model (POM) for the following reasons. First, it has an embedded turbulence closure sub-model (Mellor and Yamada, 1974, 1982; Galperin et al., 1988) for parameterizing vertical turbulence mixing. Second, it employs a sigma coordinate in the vertical, which is well suited to study the nonlinear dynamics over a shallow, gently sloping continental shelf. Third, its orthogonal curvilinear coordinates in the horizontal are convenient for resolving the near-shore regions. Previous WFS POM applications include Yang et al. (1999a,b), Li and Weisberg (1999a,b), Weisberg et al. (2000), and Weisberg et al. (2001). Yang et al. (1999a) studied the WFS response to climatological monthly mean wind forcing. Qualitative agreements were found between the model and observations to some extent, but monthly mean wind stress alone could not account for the southeastward current observed at mid shelf in spring (Weisberg et al., 1996). Li and Weisberg (1999a,b) focused on synoptic scale winds, respectively describing the kinematics and dynamics of WFS responses to idealized upwelling favorable wind forcing under constant density. The inner-shelf length scale was found to be a frictional one, consistent with the analytical work of Mitchum and Clarke (1986b). The same model was also applied to a specific upwelling

PAGE 118

86 case study with both constant density and stratified conditions (Weisberg et al. 2000). A comparison of in-situ data with model results confirmed a simple Ekman-geostrophic route to spin-up and identified regional upwelling centers promoted by coastline and isobath geometries. The utility of this model in replicating the longer-term synoptic scale variability and the sensitivity of the WFS response to its background density state was then shown by Weisberg et al. (2001). They found an asymmetry in the inner-shelf responses to upwelling and downwelling favorable winds was found that helps to clarify the scale of the inner-shelf response. Consistent with Ekman dynamics, the inner shelf is the region with divergent bottom boundary layer. Since thermal wind effects can either enhance or decrease the bottom boundary layer development, the inner-shelf under stratified conditions can respond asymmetrically to upwelling and downwelling favorable winds. Many WFS circulation questions remain. For instance, the relative effects of buoyancy and wind forcing have not been considered in the papers just cited. While previous drifter and current meter measurements hint at seasonally varying circulation patterns (Tolbert and Salsman,1964; Williams et al., 1967; Weisberg et al., 1996; Yang et al., 1999b), there are no definitive measures of these. The communication of water between the deep ocean and the shelf and the communication of water between different regions of the shelf remain poorly understood To address these questions it is necessary to explore a broader region and consider more complete forcing functions than in previous studies. The model domain (Fig 4.1) extends from the Florida Keys in the southeast to west of the Mississippi River in the northwest, and it has one open boundary arcing

PAGE 119

87 between these two locations for which a radiation boundary condition (Orlanski, 1976) is used. The model domain includes the major rivers that impact the WFS and the Desoto Canyon region where the shelf is narrowest. Its orthogonal curvilinear grid has horizontal resolution that varies from less than 2km near the coast to 6 km near the open boundary. Vertically, the sigma coordinate has 21layers with higher resolution near the surface and bottom to better resolve the frictional boundary dynamics. In total, the model has 121x81x21 grid points. Horizontal diffusivities are parameterized using the Smagorinsky (1963) formulation with a coefficient of 0.2. Bottom stress, 'Z"b, is calculated by a quadratic law with variable drag coefficient having a minimum value of 0.0025. A mode splitting technique is used for computational efficiency (Blumberg and Mellor, 1987). Here we use external and internal time steps of 6 seconds and 360 seconds, respectively. The model is initialized at rest with horizontally uniform stratification. Stratification above 200m is based on temperature and salinity observations taken during a March 1999 trans-shelf hydrographic survey [from the Ecology of Harmful Algal Blooms (ECOHAB) Program]. Stratification below 200m is based on the climatological temperature and salinity profiles. From this initial zero-baroclinicity state, the model spins up rapidly, generating baroclinicity in balance with the wind and buoyancy forcing. An alternative is to begin with a baroclinic field and allow the model currents to come into balance diagnostically with this field before proceeding with the spring simulation. The hydrographic data are not sufficient for this, however, and spurious currents due to incorrect density would corrupt the

PAGE 120

30 28 :J ...J 27 -!COo 26 \:--'A 1-CM2 \ \ .. 2-NA2 j \\ I \ i 3 -EC:i/' \ 4-EC4 \,\.I 5-ECS \,; 6-EC6 25 -90 -88 -86 -84 -82 -80 Longitude Fig. 4.1. The regional model grid (upper panel) and bathymetry and station locations (lower panel). Sea level comparisons are with Florida tide gauges at Pensacola, Apalachicola, St. Petersburg, and Naples Velocity comparisons are with acoustic Dop pler current profiles from instruments moored at the 50m, 30m, 25m, 20m, and 1Om isobaths (1-6). Temperature is described along transects I-IV and the temperature budget is diagnosed at Stations A, B, C, and D 88

PAGE 121

89 experiment. Consistent with our objective of determining the WFS responses to local, shelf-wide forcing only, our initial baroclinicity-free state is a sensible choice. Tidal forcing is excluded in the present application since we are not considering high frequency variability. It is recognized that tidal mixing can affect the synoptic and seasonal scales when the tidal currents are large, but here the tidal currents are only a few em s1 Modeled and observed tidal current analyses will be reported on separately. 4.3.2. ATMOSPHERIC FORCING Different from previous WFS model studies that considered wind forcing only, here we include both wind and thermohaline forcing. The wind and heat flux fields are from the NCEP daily reanalysis product for the period February 28, 1999 to June 1, 1999. These values, with a grid resolution of 2.5X2.5, are interpolated onto the model grid. The NCEP winds agree well with in-situ buoy winds for the spring 1999 season Unlike the winds, however, coarse resolution renders the NCEP heat flux unrealistic because of smaller scale WFS temperature structures. We correct for this using a relaxation method (e.g., Ezer et al, 1992; Chu et al, 1999). Thus, the surface heat flux forcing is given by (4.1) where QH is the net heat flux, Bobs is an interpolation of the monthly-mean satellite observed sea surface temperature, and Cp is the specific heat. The salinity flux in this study is set to be zero. The relaxation coefficient, C, or the reciprocal of the restoring time per unit area, is set at 1m/day. Such relaxation prevents deviations from observed

PAGE 122

monthly-mean SST in an attempt to force realistic baroclinic flow structures. These structures are facilitated by turbulence mixing through the coefficients KM and KH computed with the Mellor and Yamada (1982) 2.5level turbulence closure sub-model. 4.3.3. LATERAL BOUNDARY FORCING 90 Gulf of Mexico Loop Current forcing is excluded in this study with a pure radiation condition along the open boundary for two reasons. First, previous observations and model studies concluded that persistent forcing of the middle and inner-shelf by the Loop Current is minimal (Marmorino, 1982, 1983 a,b). Second, modeling the effects on the WFS of an aperiodically varying Loop Current and its associated eddies (e.g., Sturges and Leben, 2000) remains a great challenge (Marmorino, 1982; Cooper, 1987), presupposing that the Loop Current itself is being described properly. To better assess the role of the Loop Current as a WFS boundary condition it will be necessary to nest a regional model with a larger Gulf of Mexico/Caribbean/ Atlantic Ocean model. This is beyond the scope of the present paper that focuses on local, shelf-wide forcing only. We find, however, that local forcing is capable of driving much of the observed synoptic and seasonal scale variability. Seven major rivers are introduced into the model domain for land derived buoyancy forcing. These are the Mississippi, Mobile, Apalachicola, Suwannee, Hillsborough, Peace and Shark rivers. We use the technique of Kourafalou et al (1996)(also see Pullen 2000), whereby interpolated monthly mean mass flux data for these rivers are input to the top sigma level at the grid cells closest to the rivers' locations.

PAGE 123

We define the spring season here as March 1 to May 31, and we focus on this period for 1999. As an initial value problem we begin from a state of rest on February 91 28. With no initial baroclinicity the spin-up phase proceeds rapidly over the course of a few pendulum days, consistent with the barotropic response arguments for a gently sloping shelf of Clarke and Brink (1985). Under the conditions of surface cooling that occurs prior to the spring warming transition in mid-March, convective mixing very efficiently adjusts the initial density field on the shallow shelf. In other words, the "memory" of initial density field for this spring transition experiment is short (a couple of days), and sensitivity experiments that we performed using longer spin-up times showed very little difference from the present model results. 4.4. MODEL AND DATA COMPARISION 4.4.1 SEA LEVEL Since the model is forced without tides, all of the model and data comparisons are shown after low-pass filtering to exclude tidal and inertial period oscillations Sea surface height comparisons are given in Fig. 4.2. at four different tide-gauge stations from Pensacola in the northwest to Naples in the southeast. Agreement is good at all of these with squared correlation coefficients exceeding 0.80. We conclude that coastal sea level for this three-month period responds primarily to local, shelf-wide forcing. 4.4. 2 CURRENTS Comparisons are made between the modeled and observed velocity vector time series at the 50m, 30m, 25m, 20m, and lOrn isobaths (moorings CM2, EC3, NA2, EC4,

PAGE 124

5 I o -5 0.5 Pensacola f =0.84 I o -0.5 0.5 Apalachicola f =0 89 I o -0.5 0 .5 St. Petersburg f =0.82 I o -0. 5 0 5 Naples f = 0.80 I o -0. 5 6 11 16 21 26 31 5 10 15 20 25 30 5 10 15 20 25 30 Mar Apr May Fig. 4.2. Comparisons between modeled (bold) and observed (thin) sea level at Pensacola, Apalachicola, St. Petersburg, and Naples as quantified by a squared correlation coefficient, along with the NCEP wind velocity sampled at station A. 92

PAGE 125

93 EC5, and EC6 in Fig. 4.1). As examples, we show the modeled and observed vector time series at the 50m, 25m, and lOrn locations in Figs. 4.3-4.5, respectively. The observations are from acoustic Doppler current profilers (ADCP), and for each location we show comparisons at three different depths: near-surface, mid-water column, and near-bottom. These comparisons are quantified by a complex correlation analysis (e.g., Kundu 1976). Defining the modeled and observed velocity vectors in the Argand plane as w1 = u, + iv1 and w2 = u2 + iv2 respectively, the complex squared correlation coefficient is 2 p(w,,w2)= [w,(t)w2(t)] [ w 1 (t)w1 (t) w; (t)w2 (t)] (4.2) where the overbar denotes a time average. The complex correlation has an amplitude and a phase, the amplitude being the correlation coefficient and the phase being the angle (measured counterclockwise) between the modeled and the observed currents. Like sea level, the modeled and observed currents also compare well as measured by the two sets of numbers provided with each plot. The left-hand sets are the seasonal mean east and north velocity components. The right-hand sets are the squared correlation coefficient, phase angle, and regression coefficient. At all stations and depths the squared correlation coefficients range between 0.62 and 0.82, and the orientations agree to within -10 to +20 degrees. More importantly, as seen directly from the time series, the model gets the sense of the velocity rotation correct in both the surface and bottom Ekman layers. A deficiency in the modeled currents, however, lies in the amplitudes. The regression coefficients show that the model underestimates the observed velocity fluctuations by between 20 to 50 percent.

PAGE 126

10 Model 25m [0.57,-1 10] [0.63,-5. 30,0.51] 0 1 I 111\ ('--w 'I' 11\ )\\'''''' 'II' \\11 I 1'11\\lr "'1' '11\111 """ '\\\"'""'""111 '"""' -10 T 10 Model 40m [0.17, 0.85] [0 .681 15.0010 54] 0 '" \(11 '\\\' '""' 1(\\11 ''"'''" ''""'''"'111111' "' -10 6 11 16 21 26 31 5 10 15 2 0 25 30 5 10 15 20 25 30 Mar Apr M a y F ig 4.3. Comparisons between modeled and observed currents at the 50m isobath (moor ing CM2) sampled at depths of 3m, 25m and 40m along with the NCEP wind velocity sampled at station A. Each vector current time series is a ccompanied by its seasonal mean east and north velocity components (left hand c ouple t), and each model/data com parison is qu an t ifie d by its squar e d c omplex corr e lation coeffici e n t, phase angle (or angu lar devi a tion of the model v ec tor fr om the d ata v e ctor measur e d co unt e rclockwis e) and regression coefficient (right hand triplet). 94

PAGE 127

10 0 -10 10 -10 10 -10 10 u; l -10 10 -10 10 Model 20m [0.99,-0.72] [0.74,21 .91 ,0.59] -10 6 11 16 21 26 31 5 10 15 20 25 30 5 10 15 20 25 30 Mar Apr May Fig. 4.4. Comparisons between modeled and observed currents at the 25m isobath (moor ing NA2) sampled at depths of 3m, 12m and 20m, along with the NCEP wind velocity sampled at station A. Quantitative comparisons are as in Fig 4.3. 95

PAGE 128

10 i 0 -10 10 i 0 -10 10 i 0 -10 10 10 i -10 10 Model 08m (0.62, 0 25] [0.80, 12.00,0.58] I __ ,\\, -r \'0"""':;'' p I I "' (\\< "\\ 10 T 6 11 16 21 26 31 5 10 15 20 25 30 5 10 15 20 25 30 Mar Apr May Fig. 4.5. Comparisons between modeled and observed currents at the lOrn isobath (moor ing EC5) sampled at depths of 2m, 5m and 8m, along with the NCEP wind velocity sampled at station A. Quantitative comparisons are as in Fig 4.3. 96

PAGE 129

97 Notwithstanding the amplitude disparity, the model reproduces the patterns of current variability reasonably well. The systematic underestimate of the currents may be the result of the low resolution NCEP forcing fields, and hence too much smoothing when interpolating these fields onto the model grid. Model performance also degrades between the shallowest and deepest comparison sites, i.e., the lOrn isobath currents agree better than the 50m isobath currents. This is expected based on the frictional scale of the inner-shelf response to wind forcing (e.g., Weisberg et al, 2001). The 50m isobath is at the outer half of the inner-shelf so we anticipate a decreased correlation there. The modeled and observed velocity comparisons are summarized in Fig. 4.6 where we show the mid-depth seasonal mean vectors and the variance hodograph ellipses (principal axes) at all of the mooring locations. The mean vectors compare reasonably well, and while the ellipse semi-major axes are off by between 20-50%, the orientations and eccentricities tend to agree. On the basis of these agreements we now use the model to discuss the WFS circulation in spring 1999. 4.5. MEAN CIRCULATION 4.5.1 FLOW FIELDS The seasonal mean circulation, obtained by averaging the model flow fields from March 1 to May 31, is presented in Fig 4.7. Three-dimensional flow features arise because of the WFS geometry that includes partial blocking by the Florida Keys, the coastline changes of the Big Bend, and the intrusion of Desoto Canyon. These features are depicted in horizontal velocity field maps given for sigma layers 2 (near-surface), 10 (mid-level), 20 (near-bottom), and for the depth-average.

PAGE 130

27.8 27.6 27.4 27.2 Cl) "'C ::I =:: 27 10 ..J 26.8 26.6 26.4 26.2 \ \ \ \ \ Modeled \ \ I \ \ \ \ \ I \ -84 -83.5 -83 Longitude -82.5 -82 Fig. 4.6. Comparisons between modeled (bold) and observed (thin) seasonal mean veloc ity vectors and hodograph ellipses (variance principal compnent axes) at mid-depth for all six mooring locations on the WFS between the 50m and lOrn isobaths. 98

PAGE 131

Depth Averaged Current Near Surface Current Mid-depth Current Near Bottom Current Fig. 4. 7. Modeled seasonal mean (spring 1999) velocity vectors for the depth averaged and near-surface, mid-water column, and near-bottom sigma levels, k=2, 10 and 20, respectively. 99

PAGE 132

100 The mid-level and depth-averaged fields are similar (and with the same scale), and they show the general nature of the 1999 spring season currents. A jet exists with axis situated between the mid-shelf and the shelf break. This jet originates along the northern coast east of the Mississippi River, and it flows along the relatively narrow Florida Panhandle shelf as a closely confined coastal feature. The coastal jet bifurcates at Cape San Bias into a mid-shelf part that heads along-isobath toward the southeast and a coastal part that hugs the Big Bend coastline. The mid-shelf part is consistent with the spring season southeastward current described by Weisberg et al. (1996). This mid-shelf current again bifurcates upon approaching the Florida Keys with a portion turning toward Florida Bay and another portion turning farther offshore. Why this occurs and how it affects the Florida Bay/Florida Keys region are evident in the near-bottom and near surface flow fields (note the scale changes). Near the bottom we see a convergence of vectors on Florida Bay and the Florida Keys. In contrast to this, near the surface we see flow paralleling the Florida Keys. The near bottom flow upwells, feeding the near surface flow, and the second bifurcation is due to the recirculation associated with the upwelling. Li and Weisberg (1999a) discussed a similar recirculation due to the Florida Keys for the case of upwelling winds under a constant density setting. From the surface current map we can appreciate why surface drifters originating in the north do not penetrate the southeast portion of the WFS; the so-called Forbidden Zone' described by Yang et al. (1999b). We also note that the currents within the Big Bend and those that flow southward near-shore between Cedar Key and Sarasota are much weaker than the currents at mid-shelf. As will be shown in the next section, this is a consequence of a

PAGE 133

101 surface heat flux-induced cyclonic circulation that adds destructively to the wind driven flow near shore and constructively to the wind driven flow offshore. Three-dimensionality in the flow field is further evident in the vertical component of velocity shown for near-bottom, middle, and near-surface sigma levels in Fig. 4.8. Upwelling is prominent north of the Florida Keys with a maximum near the Dry Tortugas, as is often evident in satellite SST imagery (e.g., Weisberg et al., 2000). Upwelling is also prominent west of DeSoto Canyon, around Cape San Bias, and along the west Florida coastline. Other regions show downwelling. This vertical circulation is a function of both the bottom boundary layer and the interior. The near bottom w can be reconciled based on the near bottom horizontal velocity field (Fig. 4.7) and the bottom kinematic boundary condition. Mid-water column divergence then accounts for the transition of w from its near-bottom to near-surface values. Upwelling occurs along the entire near-shore domain near the bottom, whereas upwelling is more localized near the surface. 4.5.2 TEMPERATURE AND SALINITY FIELDS The modeled surface temperature and salinity fields sampled at the end of the model run on May 31 are presented in Fig. 4.9. Since the initial model temperature and salinity fields are horizontally uniform, this figure shows the combined effects of the momentum and buoyancy fluxes in changing the surface temperatur e and salinity. The two principal features are the mid-shelf cold tongue that extends southeastward from Cape San Bias and the low salinity tongue that also extends southeastward but displaced seaward of the cold tongue. Both of these featur es occur annually on the WFS. The c old

PAGE 134

30 28 27 26 Cl = 0.05m/day K=6 30 '\,; ' -. 28 27 .... '26 25 30 28 27 Cl = 0.05m/day K =12 26 Cl = 0.05m/day K=16 25 -90 -88 -86 -84 -82 -80 Fig. 4 8. Modeled seasonal mean vertical velocity component fields (converted to the z plane) sampled at the near-surface, mid-water colwnn, and near-bottom sigma layers, k=6, 12, and 16, respectively. Bold lines indicate upwelling Thin lines indicate downwelling. 102

PAGE 135

29 28 \__ 27 26 Temperature 5/31/1999 25 30 29 28 27 26 Salinity 5/31/1999 25 -90 -88 86 84 -82 80 F ig. 4.9. S e a surface temperature and sea surface s a linity fields at th e end of spring 1999 model simulation (May 31, 1999). 103

PAGE 136

tongue is imposed to some extent through the surface heat flux relaxation, whereas the low salinity tongue is a fully prognostic result of the model. 104 The low salinity tongue derives as a river plume accumulating fresh water primarily from the Mississippi River with additions from the Mobile and Apalachicola Rivers. Modulated inter-annually, the low salinity tongue extends southward each year, and in some years (1993, for instance-see Dowgiallo et al., 1995) it can be traced around the Florida peninsular to the Carolinas. Using Coastal Zone Color Scanner data between 1979 and 1986, Gilbes et al. (1995) reported a spring chlorophyll plume at mid shelf (termed the 'Green River') that also extends southeast from Cape San Bias. Their explanations for the plume included: i nutrient fluxes from the Apalachicola River; ii. nutrient fluxes from the Mississippi and Mobile Rivers; iii. seasonal changes in steric height between the shelf and deeper Gulf of Mexico; and iv. circulation of water from the Loop Current. Our results help to clarify these speculations. Consistent with Weisberg et al. (1996) the 'Green River' is associated with both the cold tongue and the low salinity tongue. The low salinity tongue, at least in 1999, appears to originate at the Mississippi River It is advected eastward, on average, by the spring coastal j e t and then southeastward where the jet bifurcates at Cape San Bias This bifurcation, in part, is related baroclinically to the cold tongue; hence the offset between the low salinity and cold tongues. Since the Loop Current is not included in our model, and since it did not extend to the northeast Gulf of Mexico in spring 1999, we can rule out the Loop Current as a primary conveyance of these salinity and temperature features What remains unclear is the relative importance of land-derived v e rsus upwelling d e riv e d nutrients in f ueling the chlorophyll plum e. Aggregation of nutrie nts and phytoplankton by the front a l

PAGE 137

105 regions of the cold tongue/low salinity tongue complex can nevertheless account for the 'Green River'. To explore the roles of momentum and heat flux in the formation of these surface temperature and salinity features we ran a model twin experiment forced by NCEP winds only. Fig 4.10 shows the seasonal mean circulation with wind forcing only for comparison with Fig.4.7 with both wind and heat flux forcing. Differences occur in all fields. The most obvious is the location of the jet, as seen in either the depth-averaged or the middle level fields. With wind forcing only the jet has a stronger coastal expression over the entire WFS, as contrasted with a stronger mid-shelf jet under wind and heat flux forcing. Baroclinicity explains the difference. With wind and heat flux forcing a dynamical feedback occurs with the cold tongue causing a cyclonic baroclinic circulation that adds constructively (destructively) with the wind forced circulation at mid-shelf (near-shore). A baroclinic circulation also arises without heat flux, but this adds constructively with the wind forced circulation near-shore. This latter effect for wind forcing only is seen in the surface temperature field of Fig. 4.11. Surface temperature is colder everywhere since the ocean circulation merely rearranges it. By virtue of mean upwelling the near-shore region is colder than the offshore region which causes a southward baroclinic circulation that adds constructively with the southward wind-driven circulation. Given the importance of surface heat flux on the seasonal circulation it is instructive to see how the baroclinic contribution evolves. This is shown in Fig. 4.12 for the March, April, and May monthly mean, depth-averaged flows relative to the seasonal mean. March shows an anticyclonic circulation with southeastward flow near-shore.

PAGE 138

Depth Averaged Current Near Surface Current \ Mid-depth Current Near Bottom Current ' Fig. 4.1 0. Modeled seasonal mean (spring 1999) velocity vectors for the depth averaged and near-surface, mid-water column, and near-bottom sigma levels, k=2, 10, and 20, respectively, from the model twin experiment forced with wind stress only. 106

PAGE 139

28 -27 26 Temperature 5/31/1999 25 -90 -88 -86 -84 -82 -80 Fig. 4.11. Sea surface temperature at the end of spring 1999 simulation (May 31, 1999) for the model twin experiment forced by wind stress only. 107

PAGE 140

April Mean Current May Mean Current Seasonal Mean Current Fig. 4.12. Evolution of the monthly mean, depth averaged velocity vectors for March, April, and May relative to the spring 1999 seasonal mean. 108

PAGE 141

109 During April, once the effects of positive heat flux set in, we see the formation of the mid-shelf jet and a cessation or reversal of the near-shore flow. This further develops into a strong mid-shelf/shelf-break current in May as the cyclonic baroclinic flow becomes fully developed. This is consistent with the Weisberg et al (1996) hypothesis on the origin of the springtime southeastward current. Temperature cross-sections at various positions along the WFS provide further information on the cold tongue evolution. These are shown for wind and heat flux forcing and for wind forcing only on March 15, April15 and May 15, in Figs. 4.13, 4.14, and 4.15, respectively, at four transects offshore of DeSoto Canyon, Cape San Blas, the Big Bend, and Sarasota. Since the heat flux is initially out of the ocean the March 15 transects, with or without heat flux, are similar at depth. They differ on the shelf where surface cooling, coupled with efficient convective mixing, produces typical wintertime horizontal stratification. By April15, with a reversal in the sign of the heat flux, the two cases depart everywhere. The Big Bend transect shows a dome of cold water at mid-shelf in the heat flux case. This originates primarily by the burial (by water being heated from above) of the cold water previously formed in the Big Bend. Water is warmer in the shallows near-shore for the same reason that they were colder in the previous season (a similar heat flux over a shallower layer will result in a larger internal energy change per unit area). There are also secondary contributions from upwelling at the shelf break. Here the shelf break upwelling is from the Big Bend as opposed to the DeSoto Canyon, and this is in part a consequence of bottom topography. The shelf break in the Big Bend occurs about 20m deeper than in the DeSoto Canyon. Hence, less upwelling is required for water to rise onto the shelf in the Big Bend than in the DeSoto Canyon. This finding

PAGE 142

0 Wind+HeatFiux Wind O n l y '" -50 I -100 a. ., 0 DeSoto Cany n -150 Transect I 3115 -200 -180 -140 -120 -100 -80 -80 -40 -20 -180 -140 -120 100 -80 -60 0 -50 I -100 a. ., 0 Cape San B l a -150 Transec t II 3/15 -200 200 -150 -100 -50 0 -200 -150 100 -50 0 0 -50 I = -100 a. ., Big Bend 0 Transect Ill Transect Ill -150 3/15 3/15 200 300 250 -200 -150 -100 50 0-300 250 -200 150 -100 -50 0 0 -20 -40 -60 I -80 i 100 120 Sarasota 140 Transect IV Transect IV -160 3/15 3/15 -1110 200 150 100 -50 200 150 100 -50 0 Offshore Distance (km] Offshore Distance [km) Fig. 4.13. Modeled temperature sections sampled on March 15 across transects originat ing at DeSoto Canyon, Cape San Blas, Florida Big Bend, and Sarasota. The contour interval is 1 C. 110

PAGE 143

-150 -200 -160 -140 -120 0 -50 :[ = -100 Q. -150 -200 -200 0 -50 :[ :6 -100 Q. ., 0 -150 -200 -300 -250 0 -20 -40 -60 :[ -80 Q. -100 ., 0 -120 -140 -160 180 -200 -150 Wind+HeatFiux -100 -80 -60 -150 100 Transect I 4/15 -40 -20 -50 -180 -140 0 200 Transect Ill 4/15 -200 -150 -100 -50 0-300 Transect IV 4/15 -100 -50 -200 Offshore Distance [km) -120 -250 -150 -100 -150 200 DeSoto Cany n Transect I 4/15 -80 -60 -40 -20 0 Cape San Bla Transect II 4/15 100 -50 0 Big Bend Transect Ill 4/15 -150 -100 -50 0 Sarasota Transect IV 4/15 100 -50 0 Offshore Distance (km) Fig. 4.14. Same as Fig. 4.13 except sampled onApril15. 111

PAGE 144

-150 -50 I -100 Q. ., 0 -150 -200 -200 0 -50 I :6 -100 g. 0 150 -200 -300 0 -20 --40 -60 I -80 -100 ., 0 -120 140 160 180 -200 Wind+HeatFiux 150 -100 -50 Transect Ill 5/15 0 -200 -150 -100 DeSoto Cany n Transect I 5/15 Cape San Bla Transect II 5/15 -50 Big Bend Transect Ill 5/15 0 -250 -200 -150 -100 50 0-300 -250 -200 -150 1 00 50 0 -150 Transect IV 5/15 -100 -50 Offshore Distance [km] -200 150 Sarasota Transect IV 5115 -100 -50 Offshore Distance [km] 0 Fig. 4 .15. Same as Fig. 4.13 except sampled on May 15. 112

PAGE 145

113 is clear in continuous animations where we also see that the advective time scale between the DeSoto Canyon and the Big Bend is longer than the synoptic weather time scale. For spring 1999, at least, we do not see a direct connection between the DeSoto Canyon and Big Bend regions as would be evidenced by continuous isotherms. Note that while 21 C waters occur on the shelf offshore of both the DeSoto Canyon and Big Bend regions these waters are not seen on the shelf at the intervening Cape San Bias cross section. Local upwelling in the Big Bend region (due to a relatively deep shelf break) is therefore important, and it may help to explain the relatively high productivity of the Florida Middle Grounds. A similar subsurface cold-water core is seen farther south off Sarasota and for the same reasons. It is smaller in magnitude because of differential cooling and heating from north to south, i.e., when the heat flux reversed from cooling to warming the temperatures off Sarasota were already warmer than those in the Big Bend. Dynamically, the doming of cold water at mid shelf (and hence the cold tongue) induces the cyclonic baroclinic circulation. This is further facilitated by the no heat flux bottom boundary condition that bends the isotherms perpendicular to the bottom. These combined effects of differential heating (near-shore to offshore by bottom slope and southward to northward by solar declination) and dynamical feedbacks (baroclinic pressure gradients associated with the temperature gradients) conspire to cause the spring season temperature transition from well mixed to stratified with a cold tongue at mid shelf. The May 15 transects show the continued evolution of these processes. By now the mid-shelf isotherm doming has relaxed due to continued heating and so has the mid shelf circulation due to decreased winds and baroclinicity. The location of the jet is now

PAGE 146

farther offshore at the shelf break, and the modeled temperature agrees with observed ECOHAB Program hydrography. 4.5.3. LAGRANGIAN DRIFTERS 114 Model-simulated Lagrangian drifter tracks provide additional information on the evolution of the cold and low salinity tongues and the regions for possible communication between the deep ocean and the shelf. Dynamically passive particles released in the model on April 1 stare tracked through May 31st. These are shown in Figs. 4.16 and 4.17 for particles originating either near the surface or the bottom, respectively. Most evident for the particles released near the surface is advection by the mid-shelf jet and the bifurcations at Cape San Bias and north of the Florida Keys. The particle trajectories agree with the Eulerian maps previously shown and with the cold and low salinity tongue features. With the exception of those in the mid-shelf jet the particle displacements are not very large. The near-surface drifters also show a propensity to flow either along-isobath or offshore. The near-bottom released drifters show subtle, but important differences, particularly in the Florida Big Bend/Middle Ground region. There we see onshore flow between the shelf break and the mid-shelf. This is consistent with the changes in hydrography discussed in the previous section, supporting our speculation that the Big Bend may be an important region for deep ocean/shelf interactions.

PAGE 147

Transect 1 -89 -88 -87 -86 -85 Transect 4 Transect 2 29.5 29 28.5 28 -90 -89 -88 -87 -86 -85 Transect 5 Transect 3 -90 -88 -86 -84 -82 Transect 6 29.5 28.5 r-------------r---t 29.5 29 28.5 28 27.5 27 ;\ 28 27.5 27 26.5 26 26.5 -87 -86 -83 -82 26 -87 -85 -84 -86 -85 -84 -83 -82 25.5 Transect 7 28.5 r--------r---. 28 27.5 27 26.5 26 25.5 25 24.5 j...._ _ .....:..... ___ _,_.-'!1 -86 -85 -84 -83 -82 26. 5 26 25.5 25 Transect 8 -85 -84 -83 -82 -81 -86 -85 -84 -83 Transect 9 -85 -84 -83 -82 -81 Fig. 4.16. Modeled near-surface Lagrangian drifter trajectories for the period April1, 1999 to May 31, 1999. Drifters were started along nine different transects with initial positions given by solid dots. -82 115

PAGE 148

116 Transect 1 Transect 2 Transect 3 30 5 30.5 30 30 30 29.5 29.5 29 29 29 28 28 5 28 5 28 28 27 -90 -89 -88 -87 -86 -85 -90 -89 -88 -87 -86 -85 -90 -88 -86 -84 -82 Transect 4 Transect 5 Transect 6 30 29 5 28 5 29 5 29 28 29 28 5 27 5 28 5 28 27 28 27.5 27. 5 27 26 5 27 26. 5 26 26 5 26 25 5 -87 -86 -85 -84 -83 -82 -87 -86 -85 -84 -83 -82 -86 -85 -84 -83 -82 Transect 7 28 5 Transect 8 28 27 Transect 9 27 5 26 26 5 27 26 25 5 26 5 25 25 5 26 24 5 25 .5 25 24 25 2 4 5 -85 -84 -83 -82 -81 -85 -84 -83 -82 -81 24 5 -86 -85 -84 -83 -82 Fig. 4 17. Same as Fig. 4.16 except for near-bottom Lagrangian drifter trajectories.

PAGE 149

117 4.6. TEMPERATURBUDGET 4.6.1 THE TEMPERATURE EQATION To analyze the temperature budget, the temperature equation is recast from its modeled flux divergence, sigma level form to an advective, z-level form. Thus, we diagnose a b c d vf hfx h/y which equates the local rate of change of temperature (a) to a combination of the flow field advective rate of change (b+c+d), and the rates of change by vertical diffusion (vf), and horizontal diffusion (hfx+h/y). The temperature balance is explored using: i. time series of vertical averages, ii. time series of vertical profiles at four different locations, and iii. term by term spatial maps of the sea surface temperature rates of change. The four analysis locations (see Fig 4.1) are chosen with respect to the cold tongue and the upwelling region north of the Florida Keys. Point A is on the 50m isobath at the seaward side of the Big Bend shelf subsurface cold dome. Point B is on the 40m isobath west of Tampa Bay where the cold tongue begins to taper off. Point C is on the 15m isobath offshore of Sarasota on the inshore side of the cold tongue. Point Dis north of the Florida Keys. A term-by-term analysis of (3) quantifies the contributions by each physical process in changing the temperature. 4.6.2. Depth-averaged balances

PAGE 150

118 A depth-averaged temperature equation is obtained by vertically integrating equation (4.3). Since horizontal temperature diffusion is generally at least an order of magnitude less than the other terms, the depth-averaged diffusion term is essentially the depth-averaged vertical diffusion, Q l(pC pH), where Q is the net surface heat flux, p and Cp are the seawater density and specific heat, and His the water depth. The temperature variations depend on both ocean advection and diffusion. With a twodimensional model, Morey (1999) proposed that temperature, in a depth-averaged sense, could be well represented without advection, i.e., dT oT Q(t) -=-=--(4.4) The validity of this assertion for various locations on the WFS can be evaluated by a fully three-dimensional analysis. Time series of the depth averaged ocean advection and diffusion terms and their sum for locations A-D are shown in Figs. 4.18a-4.18d, respectively. While not shown in these figures, we note that the sum of these two terms is nearly exactly equal to the negative of the local rate of temperature change, which is essentially a check on our budget analysis. At station A, the depth-averaged rates of change are relatively small, and a clear distinction exists between the balances occurring on seasonal and synoptic time scales. The seasonal change (as given by the three-month mean values) is primarily by surface heat flux, whereas the synoptic variability (as given by the standard deviations) is primarily by ocean advection. Although a relatively small contribution to the seasonal change, the ocean advection does provide a cooling influence that offsets the surface heating by about 4% at this location. Thus, of the total change in vertically

PAGE 151

I I I I I I I I I I I 0.5 Advection: -udT /dx-vdT /dy-wdT/dz Mea n :-7 .77e-04 Std :8.56e-02 0 :::;: ..... <"':"" -0. 5 -1 I l I 0 5 Diffusion:HFx+HFy+ VF Mean :1.88e-02 Std :4.45e-02 -0. 5 Advection+ Diffusion Mean :1.81e-02 Std : 8 .87e-02 0.5 0 A A C=-:"> <"'> f>. c=---...-... J\ C\ o """"'J v .....,.'\/ "' ""' =v-v 'V"""' 'V \...7V ... -0. 5 6 11 16 21 26 31 5 10 15 20 25 30 5 10 15 20 25 30 Mar Apr May 1 999 Fig. 4.18a The relative contributions to the depth averaged temperature balance by ocean circulation and diffusion at station A. Three time s e ries are shown: the advection the diffusion, and their sum (which equals the local rat e of chang e of d e pth-aver a g e d t e m perature). Accompanying each time series are their seasonal means and standard devia tions in units ofC day-1 as measures of the seasonal and synoptic scale variabil ity. 119

PAGE 152

0.5 Advection:-udT/dx-vdT/dy-wdT/dz Mean :7.34e-03 Std :1.52e-01 :e 0 1\ /\." A e= v "'""" -0.5 0.5 -0.5 0.5 -0.5 Diffusion: H Fx +H Fy+ VF Advection+ Diffusion I I I I I I I I I Mean :2.58e-02 Mean :3.32e-02 Std :4.27e-02 Std :1.58e-01 \] \ 6 11 16 21 26 31 5 10 15 20 25 30 5 10 15 20 25 30 Mar Apr May 1999 Fig. 4.18b Same as Fig. 4.18a except for station B. 120

PAGE 153

0.5 0 -0.5 -1 Diffusion:HFx+HFy+VF 0.5 -0.5 0 5 0 -0.5 Mean :6.52e-02 Std :9.42e-02 6 11 16 21 26 31 5 10 15 20 25 30 5 10 15 20 25 30 Mar Apr May 1999 Fig. 4.18c Same as Fig. 4.18a except for station C. 121

PAGE 154

122 I I I I I I 0.5 Advection:-udT/dx-vdT/dy-wdT/dz Mean :-3.03e-02 Std :9.45e-02 0 A b. /\.. /'\. __........_ ..::::::. v-v v v-v-= v -yi7"--JV-V'-.1' -...._.r v-----0.5 -1 I I I I I I i i i i i : i : i Diffusion:HFx+HFy+VF Mean :8.82e-02 Std :6.96e-02 0.5 0 -0.5 -1 -0.5 1 6 11 16 21 26 31 5 10 15 20 25 30 5 10 15 20 25 30 Mar Apr May 1999 Fig. 4.18d Same as Fig. 4.18a except for station D.

PAGE 155

123 averaged temperature (1.66C), the contributions by ocean circulation and surface heat flux are a cooling of -0.07C and a warming of 1.73C, respectively. In contrast with the seasonal change, for synoptic scale variability we see that the ocean circulation is twice as effective as the surface heat flux in changing the vertically averaged temperature at station A. Somewhat different results are found at station B. While surface heat flux and ocean circulation again control the seasonal and synoptic scale variability, respectively, the relative magnitudes and the signs change. Of the total change in vertically averaged temperature (3.05C), both the ocean circulation and the surface heat flux at the southern terminus of the cold tongue contribute to warming at 0.68C and 2.37C, respectively. The ocean circulation also plays a proportionately larger role in the synoptic scale variability at station B than at station A. The relative importance of the ocean advection versus surface heat flux continues to change approaching shallower water. At the 15m isobath, station C, of the total change in vertically averaged temperature (4.76C), the contributions by ocean circulation and surface heat flux are cooling of -1.24C and a warming of 6.00C, respectively. The shallower the water, the larger the total spring transition change. However, the role of the ocean circulation in this change varies in both magnitude and sign with location. For instance, at station D in the upwelling region north of the Florida Keys, the ocean circulation provides a cooling influence of 2.79C that partially offsets the surface heat flux warming influence of 8.11 C. Without the upwelling influence by ocean circulation the water temperature north of the Florida Keys would be much warmer. Along with these increased effects by the ocean circulation on the seasonal mean temperature change,

PAGE 156

124 the magnitude of the controlling ocean circulation effect on the synoptic scale variability also increases with decreasing water depth. The temperature budget on all time scales is a complex one requiring a fully three-dimensional description. 4.6.3. VERTICAL PROFILES OF THE TERM BY TERM BALANCE The temperature budget three-dimensionality is further explored through time series of the depth profiles of the individual terms that comprise the temperature balance at these four stations (Figs.4.19-4.22). In each of these figures the left hand panels show the horizontal and vertical components of the ocean advection and their sum, and the right hand panels show the diffusion, the diffusion plus the advection (which is nearly exactly equal to the negative of the local rate of change of temperature), and the temperature. With the exception of the initial portion of the record when the surface heat flux is out of the ocean and convective mixing is evident, the immediate impact of the surface heat flux is primarily in the near surface region (Fig. 4.19). Without convective mixing, it is then only through turbulence mixing, brought about by the ocean circulation dynamics, that the surface heat flux effect accumulates downward to make its contribution to the depth-averaged spring transition. From the vertical distributions of the seasonal means shown to the right of each panel we see that turbulence mixing distributes the surface heat flux over about 20m depth. Along with its role in turbulence mixing, the direct role of the ocean dynamics through advection on the synoptic scale variability is seen in both horizontal and vertical directions. Omission of any coordinate direction would compromise the model's ability to describe the temperature evolution.

PAGE 157

-udT/dx-vdT/dy VF+HFx+HFy -10 -10 -20 -20 -30 -30 -40 -40 -4. 1 -wdT/dz 0 1 -4 1 -udT/dx-vdT/dy-wdT/dz+VF+HFx+HFy -10 -20 -30 -40 -4. 1 0 1 -4.1 -udT/dx-vdT/dy-wdT/dz -10 \ -20 -30 -40 6 1116 21 26 31 5 10 15 20 25 30 5 1015 20 25 30 -4.1 0.1 6111621263151015202530 5 1015202530lD Mar May Mar May 1999 1999 Fig. 4.19. Time series of the depth profiles of the individual terms that comprise the temperature balance at station A. The left hand panels show the horizontal and vertical components of the ocean advection and their swn, and the right hand panels show the diffusion, the diffusion plus the advection, and the temperature. To the right of each panel is the seasonal mean profile. The contour interval for each of the budget terms is 0.05C day1 and the contour interval for temperature is 0.5C. Shading indicates warm ing and clear indicates cooling. 125 0 1 0.1 25

PAGE 158

126 -udT/dx-vdT/dy VF+HFX+HFy -10 -10 -20 0 0 -20 -30 -0 1 0.1 -0.1 0.1 -wdT/dz -udT/dx-vdT/dy-wdT/dz+VF+HFx+HFy -10 -20 -10 0 I 1 \ 0 1'--IB' A -20 i -0 1 0 1 -udT/dx-vdT/dy-wdT/dz \ -10 6 11 16 21 26 31 5 10 15 20 25 5 10 15 20 25 -0.1 0.1 6 11 16 21 26 31 5 10 15 3) 25 5 10 15 20 25 30 211 25 Mar lfK May Mat lfK May 1999 1999 Fig 4.20. Same as Fig. 4.19 except for station B

PAGE 159

127 -udT/dx-vdT/dy -2 -2 -4 -4 -6 -6 -8 -8 -10 -10 -12 -12 0.1 0.1 -wdT/dz -udT/dx-vdT/dy-wdT/dz+VF+HFx+HFy -2 -2 -4 -4 -6 -6 -8 -8 -10 -10 -12 -12 0.1 0.1 -udT/dx-vdT/dy-wdT/dz Temperature -2 -2 -4 -4 -6 -6 -8 -8 -10 -10 -12 -12 6 1116 21 26 31 5 1015 20 25 30 5 1015 20 25 30 0.1 6 1116 21 26 31 5 10 15 20 25 30 5 10 15 20 25 30 20 25 Mar Apr May Mar Apr May 1999 1999 Fig. 4.21. Same as Fig. 4.19 except for station C.

PAGE 160

128 -udT/dx-vdT/dy VF+HFx+HFy -2 -4 -4 -6 -6 \ -6 -8 -10 -10 -12 I -12 -a1 0 1 0.1 -wdT/dz -udT/dx-vdT/dy-wdT/dz+VF+HFx+HFy -2 -2 -4 -4 -6 -6 -8 -8 -10 -10 -12 -12 0.1 0.1 -udT /dx-vdT /dy-wdT /dz Temperature -2 -2 -4 -4 -6 -6 -8 -8 -10 -10 -12 -12 6 1116212631 5 1015202530 5 1015202530 0.1 6 1116 21 26 31 5 1015 20 25 30 5 1015 20 25 30 2D 25 Mar May Mar 1999 1999 Fig 4.22. Same as Fig. 4.19 except for station D.

PAGE 161

129 For instance, the strong stratification at the end of May at station A is due largely to the combined effects of horizontal and vertical advection. Similar conclusions follow from station B (Fig. 4.20). Station C (Fig. 4.21), in shallow water, shows the additional effect of turbulence mixing in the bottom Ekman layer. Here, temperature is elevated near the bottom by ocean circulation-induced mixing. The role of the ocean circulation in promoting mixing is seen in the covariability between the advection and diffusion terms. These interactions are even more evident at station D (Fig. 4.22) where warming by near bottom turbulence mixing is required to partially offset the cooling influences by horizontal advection and upwelling. Another interesting feature of the shallow water regions is the tendency for large changes in stratification by the ocean circulation. Two downwelling events culminating in destratification and warming by advection are seen at station C around April 5th and May lOth. Similarly, but in an opposite sense, stratification caused by horizontal advection is seen at this station at the end of May. 4.6.4. TERM BY TERM CONTRIBUTION TO THE SEASONAL CHANGE IN SST The contributions to the seasonal mean rate of change of sea surface temperature (SST) by each of the advection and diffusion terms are discussed with respect to Fig. 4.23. The left panels show the advection terms and their sum, and the right panels show the horizontal and vertical diffusion terms, and the sum of all terms. Of particular interest here is the spatial distribution of the processes controlling SST over the model domain. Horizontal diffusion is minimal everywhere except for north of the Florida Keys where it is still the smallest of the terms. Vertical diffusion controlled by the surface heat flux, warms everywhere with maximum warming tendencies stretching along mid-shelf

PAGE 162

udT / dx-vdT /dy 30 29 28 27 26 CI=0.02C/d 25 -wdT/dz 30 29 28 27 26 25 -udT/dx vdT/dy wdT/dz udT/dx-vdT/dy wdT/dz+HFx+HFy+V F 30 29 28 27 26 CI=0 02C / d 25 -90 -88 -as -84 -a2 90 -88 -88 -84 -a2 Fig. 4.23. The contributions made to the seasonal mean rate of change o f SST by each of the advection and diffusion terms. The left panels show the advection terms and th eir sum, and the right panels show the horizontal and vertical diffusion terms, the sum of all terms. The contour interval is 0.02C dar B o ld lines indicate warming and thin lines indicate. coo ling. 130

PAGE 163

131 from the Big Bend to the Florida Keys. This wanning is the seasonal signal associated with spring, which displays the spatial inhomogeneities mentioned earlier. The difference between this map and the local rate of change of SST is the ocean circulation. By advecting relatively cool water southward, advection provides a cooling influence over the WFS with largest values north of the Florida Keys With regard to total advection, Cape San Bias divides the entire domain into two sections. To the west, and extending along the north Florida coastline to the Mississippi River, advection tends to provide a warming influence, whereas to the east, along the west Florida coastline, advection tends to provide a cooling influence. Without advection, the center of the cold tongue on the WFS would be some 3-5C warmer. 4.7. SUMMARY Mid-latitude continental shelves undergo a spring transition as the net surface heat flux changes from cooling to wanning. Using in-situ data and a numerical circulation model we investigate the circulation and temperature budget on the WFS, including the northeast Gulf of Mexico shelf from the Mississippi River to the Florida Keys, for the spring transition of 1999. The data consist of sea level from coastal stations, velocity profiles from instruments moored across the shelf between the 50m and 1Om isobaths, and hydrography from ship surveys. The model is a regional adaptation of the primitive equation, POM forced by NCEP reanalysis wind stress and heat flux fields and by river inflows. Based on agreements between the modeled and observed fields we use the model to draw inferences on how the surface momentum and heat fluxes affect the seasonal and synoptic scale variability.

PAGE 164

132 Spring season features of the WFS include a mid-shelf southeastward current, cold and low salinity tongues, and a high chlorophyll plume. We account for the southeastward current in 1999 by the combined responses to local, shelf-wide wind and buoyancy forcing. Wind stress drives a circulation that tends to be strongest near-shore. Heat flux provides a cyclonic contribution that adds constructively (destructively) at mid shelf (near-shore), thus forming the observed mid-shelf jet. This heat flux-induced baroclinic circulation is related to the spring season cold tongue. By advecting Mississippi (and other) River water it forms the low salinity tongue that is displaced seaward of the cold tongue. Convergence of nutrients and associated phytoplankton growth then accounts for the high chlorophyll concentrations ('G-reen River') that are co located with these surface features. These findings support the hypothesis advanced by Weisberg et al. (1996) on the origin of the southeastward current and cold tongue through differential heating from the coast to offshore (by shoaling topography) and from south to north (by solar declination). Since we arrive at these features with a model experiment that explicitly omits the Gulf of Mexico Loop Current we argue that the Loop Current is not an essential element of these spring transition features. Through term-by-term analyses of the temperature budget we describe the evolution of WFS temperature in spring. Surface heat flux largely controls the seasonal transition, whereas ocean circulation largely controls the synoptic scale variability. These two processes are closely linked, however. Since the ocean circulation controls the turbulence mixing, the effects of the ocean circulation are of increasing importance with decreasing water depth. For instance, warming by turbulence mixing near the bottom only occurs when the bottom Ekman layer is well developed by strong currents. Thus,

PAGE 165

133 the water column can warm even under upwelling influence if the mixing is large enough. The water column is also found to either stratify or destratify in response to ocean circulation changes. Examples of these effects are shown. While our temperature analyses support the Morey (1999) one dimensional temperature balance argument advanced on the basis of a two dimensional model we demonstrate that the temperature balance is more complicated in time and space and requires fully three-dimensional thermodynamics. Bottom topography and coastline geometry are important in generating regions of convergence and divergence and hence upwelling centers. In particular, we show that the region north of the Rorida Keys has strong upwelling in spring and we speculate on the importance of the Rorida Big Bend as a region for communication between the deeper Gulf of Mexico and the WFS. The shelf break there is about 20m deeper than at the DeSoto Canyon thereby requiring less upwelling for deeper waters to broach the shelf. This may be one reason why the Rorida Middle Grounds are productive. The northeast Gulf of Mexico shelf from Cape San Blas to the Mississippi River also shows mean upwelling during spring 1999.

PAGE 166

REFERENCES Blumberg, A. F., and G. L. Mellor (1987), A description of a three-dimensional coastal ocean circulation model, Three-Dimensional Coastal Ocean Models, Vol. 4, N. Heaps (ed.), 208-233, AGU, Washington, D. C., 1987. Chew, F. (1955), On the offshore circulation and a convergence mechanism in the red tide region off the west coast of Florida. EOS Trans. AGU 36, 963-97 4 Chu, P. Edmons, N. and Fan, C (1999). Dynamical mechanisms for the south china sea seasonal circulation and thermohaline varibilities. J. Phys. Oceanogr., 29, 29712989 Chapman, D.C and G. Gawarkiewicz (1993), On the establishment of the seasonal pycnocline in the Middle Atlantic Bight, J. Phys. Oceanogr., 23, 2487-2492 Clarke, A. J., and K. H. Brink (1985), The response of stratified, frictional flow of shelf and slope waters to fluctuating large-scale, low-frequency wind forcing, J. Phys. Oceanogr., 15, 439-453 Cooper, C. (1987), A numerical modeling study of low-frequency circulation on the west Florida shelf, Coastal Engineering, 11, 29-56 Cragg, J., G. Mitchum, and W. Sturges (1983), Wind-induced sea-surface slopes on the West Florida shelf, J. Phys. Oceanogr., 13, 2201-2212 Dowgiallo, M.J., ed., Coastal oceanographic effects of the summer 1993 Mississippi River flooding, Special NOAA Report, March 1994, 77pp. Ezer, T ., and G. Mellor (1992), A numerical study of the variability and separation of the Gulf Stream, induced by surface atmopheric focing and lateral boundary flows. J. Phys. Oceanogr. 22, 660-682 Galperin, B., L. H. Kantha, S. Hassid, and A. Rosati (1988), A quasi-equilibrium turbulent energy model for geophysical flows, J. Atmos. Sci., 45, 55-62 Gilbes, F., C. Thomas, J. J. Walsh, and F. E. Muller-Karger (1996), An episodic chlorophyll plume on the West Florida Shelf, Continental Shelf Research, 16, 1201-1224

PAGE 167

Kourafalou, V. H., L.-Y. Oey, J.D. Wang, and T.L. Lee (1996), The fate of river discharge on the continental shelf. 1: Modeling the river plume and the inner shelf coastal current. J.Geophys. Res., 101, 3415-3434 Kundu, P.K. An analysis of inertial oscillations observed near the Oregon coast (1976). J. Phys. Oceanogr. 6, 879-893 Li, Z., and R. H. Weisberg (1999a), West Florida continental shelf response to upwelling Favorable wind forcing, part ll: kinematic description. J.Geophys. Res., 104, 13507-13527 Li, Z., and R. H. Weisberg (1999b), West Florida continental shelf response to upwelling favorable wind forcing, part ll: dynamical analyses, J. Geophys. Res., 104, 23427-23442 Marmorino, G. 0 (1983a), Variability of current, temperature, and bottom pressure across the West Florida continental shelf, winter, 1981-1982, J. Geophys. Res., 88, c7, 4439-4457 Marmorino, G. 0 (1983b), Summertime coastal currents in the Northeastern Gulf of Mexico, J. Phys. Oceanogr., 13, 65-77 Mellor, G.L. and T. Yamada (1974), A hierarchy of turbulence closure models for planetary boundary layer, J. Atmos. Sci., 13, 1791-1806 Mellor, G. L. and T. Yamada (1982), Development of a turbulence closure model for geophysical fluid problems, Rev. Geophys., 20, 851-875 Mitchum, T. G., and A. J. Clarke (1986a), The frictional nearshore response to forcing by synoptic scale winds, J. Phys. Oceanogr., 16, 934-946 Mitchum, T. G., and A. J. Clarke (1986b), Evaluation of frictional, wind forced long wave theory on the West Florida shelf, J. Phys. Oceanogr., 16, 1029-1037 Mitchum, T. G., and W. Sturges (1982), Wind-driven currents on the West Florida shelf, J. Phys. Oceanogr., 12, 1310-1317 Morey, S. L. (1999), The spring transition of thermal stratification on a mid-latitude continental shelf. Ph. D. dissertation, COAPS, Florida State University Tallahassee Niiler, P. P. (1976), Observations of low-frequency currents on the West Florida continental shelf, Memoires Societe Royale des Sciences de Liege, 6, X, 331-358, 1976. Orlanski, I. (1976), A simple boundary condition for unbounded hyperbolic flows, J.

PAGE 168

Comput. Phys., 21, 251-269, 1976. Pullen, J.D. (2000), Modeling studies of the coastal circulation off northern California, Ph.D. Dissertation, Oregon State University 136 Paluszkiewicz, T., L. Atkinson, E.S. Parmentier, and C.R. McClain (1983), Observation of a Loop Current front eddy intrusion onto the West Florida Shelf, J. Geophys. Res., 88, 9636-9651, 1983 Price, J. F. (1976), Several aspects of the response of shelf waters to a cold front passage, Memories Societe Royale des Sciences de Liege, 6, X, 201-208 Smagorinsky, J. (1963), General circulation experiments with primitive equations 1: The basic experiment. Mon. Weather. Rev., 91, 99-164 Sturges, W., and R. Leben (2000), Frequency of ring separation from the Loop Current in the Gulf of Mexico: A revised estimate. J. Phys. Oceanogr., 30, 1814-1819 Tolbert, W. H. and G.G. Salsman (1964), Surface circulation of the eastern Gulf of Mexico as determined by drifter bottle studies, J. Geophys. Res., 69, 223-230 Vargo, G. A. and E. Shanley (1985), Alkaline phosphates activity in the red-tide dinoflagellate ptychodiscus brevis, Mar. Ecol., 6, 251-264 Weisberg, R. H., B. D. Black, H. Yang (1996), Seasonal modulation of the West Florida continental shelf circulation, Geophys. Res. Lett., 23,2247-2250 Weisberg, R. H., B. D. Black, Z. Li (2000), A upwelling case study on the Florida's west coast. J. Geophys. Res. 105, 11459-11469 Weisberg, R. H., Z. Li, F. MullerKarger (2001), West Florida shelf response to local wind forcing, April1998. J.Geophy. Res., 106 (C12) 31,239-31,262 Williams, J., Grey, W.F., Murphy, E.B., J. J. Grane (1977), Memoirs of the Hourglass cruise. Report of the Marine Research Laboratory, Florida of National Research, St. Petersburg, IV, part ill, 134 pp Yang, H., R. H. Weisberg (1999), Response of the west-Florida continental shelf to Climatological monthly mean wind forcing, J. Geophy. Res., 104, 5301-5320 Yang, H, R. H. Weisberg, P. P. Niiler, W. Sturges, W. Johnson (1999), Lagrangian circulation and forbidden zone on the West Florida Shelf. Continental shelf Res. 19. 1221-1245

PAGE 169

CHAPTERS LOCAL AND DEEP-OCEAN FORCING CONTRIBUTIONS TO ANOMALOUS WATER PROPERTIES ON THE WEST FLORIDA CONTINENTAL SHELF 5.1 ABSTRACT Continental shelves are regions where land-drainage-derived and ocean waters mix to determine the material properties of the littoral zone. The overarching question regarding the origin of littoral zone water properties is the relative importance of local 137 and deep-ocean forcing. We address this question for the west Florida continental shelf (WFS) using in-situ data and a numerical circulation model for the years 1998 and 1999. The spring and summer seasons in these years show distinctively different water properties (cold and nutrient rich water deep ocean water intruded onto the shelf in 1998), both on the shelf and at the shelf-break. We account for these differences by a combination of local forcing, independent of the adjacent Gulf of Mexico Loop Current, and Loop Current interactions with the shelf. The primary role of the deep-ocean processes is to modulate the material isopleth distributions at the shelf break. Local, shelf-wide wind and buoyancy forcing determines whether or not these material isopleths broach the shelf. The subsequent along and across-shelf material property distributions are determined by a combination of local and deep-ocean forcing affects.

PAGE 170

137 5.2. INTRODUCTION Continental shelves are buffer regions between the coast and the deep-ocean. There, land-drainage-derived and ocean waters mix to determine the material properties of the littoral zone. The overarching question regarding the origin of littoral zone water properties is the relative importance of local, as opposed to deep-ocean forcing. Here we define local forcing as the shelf-wide inputs of momentum (by winds) and buoyancy (by surface heat flux, evaporation minus precipitation, and river inflows). Deep-ocean forcing is defined as momentum and buoyancy input at the shelf-break by boundary currents and eddies. Figure 5.1 helps to conceptualize these definitions. The WFS, for our purposes, extends from the Florida Keys to the Mississippi River. It is influenced from the deep ocean by the Gulf of Mexico Loop Current (LC), the precursor of the Gulf Stream, that enters through the Yucatan Channel and exits through the Florida Straits. While en route the LC migrates northward and sheds an anticyclonic eddy before collapsing back to the south and starting its northward migration cycle anew (e.g., Sturges and Leben, 2000). At times the LC and its eddies impinge on the WFS, and, at the very least, the baroclinic structure of the LC and its adjacent waters sets the thermocline height at the shelf-break and hence the potential for driving cold, nutrient rich waters onto the shelf. Whether the LC itself drives deep-ocean water properties onto the shelf or merely sets the stage from which local processes take over is the main question of our paper. This local versus deep-ocean forcing question is of general concern for all continental shelves. Recent reviews on the effects of local wind, buoyancy, and deep-ocean forcing on continental

PAGE 171

138 Fig. 5 .1. The West Florida Shelf and its associated local and offshore forcing influences.

PAGE 172

139 shelves are provided by Brink (1998a), Hill (1998), and Brink (1998b), respectively, and Boicourt et al. (1998) shows that distinguishing these effects remains a challenge. The WFS exhibits complex variable topography. It is wide and gently sloping in the south where it is partially closed by the Florida Keys. It remains wide with subtle variations to the north. It then narrows west of Cape San Bias on the Florida Panhandle where it pinches down to a minimum at DeSoto Canyon. The shelf then widens again in the Mississippi Bight before hitting a choke point at the Mississippi River delta. Local forcing throughout the WFS is by seasonal and synoptic scale weather systems, tropical cyclones, and distributed river inputs. The Mississippi River, draining the USA heartland, is the major source of fresh water. East of the Mississippi River are the cumulative inputs from other rivers of southeast USA origin and the inflows from Florida's estuaries and the Everglades. The river inputs also vary seasonally. Peak flows occur in spring for the rivers entering the northern portion of the domain and in summer for the rivers originating in Florida. In sum, the WFS is an excellent natural laboratory for studying continental shelf processes. It contains both end-members based on shelf slope, including regions of wide, gently sloping and narrow, steeply sloping isobaths; it has an adjacent western boundary current; and it is meteorologically forced by well defined seasons (with positive and negative surface heat fluxes) and seasonally modulated synoptic scale weather. WFS measurements suggests a seasonal cycle in which the circulation tends toward upwelling in winter and downwelling in summer, consistent with monthly mean wind stress climatology (Yang and Weisberg, 1999). The seasonal transitions in spring and fall begin when the net surface heat flux changes sign. Weisberg et al. (1996)

PAGE 173

140 hypothesize that baroclinicity imposed by surface heat flux is a necessary part of the seasonal circulation transitions, and this hypothesis is borne out by joint in-situ data/numerical model analyses for spring and fall by He and Weisberg (2002a,b), respectively. Not all years are alike, however. For instance, 1998 shows anomalously cold water on the shelf that lasts well into summer, along with a sharply defined thermocline that encroaches right to the beach. Can this behavior be accounted for by local forcing, or must offshore influence be invoked and, if so, how? This inter-annual variability and the ensuing questions are what motivate the present paper. Our objectives transcend physical oceanography since inter-annual variability is also important biologically. The WFS is an epicenter forK. brevis red-tide (e.g., Steidinger et al., 1999), and while the WFS is generally considered to be oligotrophic it supports major recreational and commercial fisheries. Unraveling the circulation influences that determine material property distributions is necessary to understand these and other societal important consequences. This chapter is organized as follows. We begin in section 5.3 with a description of relevant data in 1998, a time when we were gearing up for a multidisciplinary red-tide study. Anomalous conditions are observed in the currents and hydrography in 1998 when compared with similar observations in 1999. Some of these differences can be explained on the basis of local forcing, independent of the LC (or its eddies), as shown by the numerical model simulations of section 5.4. This is consistent with the 1999 spring transition study of He and Weisberg (2002a), hereafter referred to as HW02. However, not all of the1998 features can be accounted for by local forcing alone. Motivated by anomalous sea surface height gradients observed in satellite altimetry, the Hetland et al.

PAGE 174

141 (1999) hypothesis that LC impingement near the Florida Keys affects the WFS because of isobath convergence, and the LC case study of He and Weisberg (2002c ), an idealized LC is added to the previous model simulation in section 5.5. This allows us to account for the remaining anomalies, but in a complex, baroclinic way. Section 5.6 discusses the results, and how the interaction between local and offshore forcing provides a basis for inter-annual injections of nutrient rich, deep-ocean waters onto the shelf. It also identifies other times in the observational record when such behavior occurs. Section 5.7 provides a summary. 5.3. OBSERVATIONS The spring to fall seasons of 1998 show anomalous stratification and coastal upwelling Anomalous must be qualified as anecdotal since WFS data are insufficient to formally define deviations from climatologies. Nevertheless, sea surface temperature (SST) by satellite imagery was colder than normal along the west Florida coastline, hot summer bathers rejoiced to rivulets of cold water welling up at their feet, and recreational divers carped about cold, murky water below a sharp, shallow thermocline. Measurements from our Ecology of Harmful Algae Blooms (ECOHAB: Florida) Program began in June 1998, prior to which we have a more limited data set. Figure 5.2 shows the positions of the various time series and hydrographic sections that are used. We begin with wind velocity vectors sampled at NOAA buoy 42036 and water velocity vectors sampled at moorings on the 50 m and 20 m isobaths. These are shown for the periods January-September of 1998 and 1999, respectively, in Figures 5.3a and 5.3b. With the exception of subtle, but very important, differences the winds at first glance

PAGE 175

(I) "C 28 :::s .... ...J 27 26 25 -90 -89 -88 -87 -86 -85 -84 Longitude : J <) florida Bay J "" elK" -83 -82 -81 -80 Fig. 5. 2. The locations of the hydrographic and in-situ current measurements. Circles denote the Texas A&M NEGOM ship survey stations. The thin lines denote the USF ECOHAB hydrographic transects TA (Tampa) and SA (Sarasota). The thick line denotes the Mote Marine Lab hydrographic transect. The double cross denotes the NOAA meteo rological buoy 42036. Stars denote the moored ADCP sites PC, AS1/CM2, and LB3/EC4 at the 30 m, 50 m, and 20 m isobaths, respectively. 142

PAGE 176

tO a. E tO 10 Ui Q) L() 0 L() 0 L() 0 L() ,.... ,.... I ,.... ,.... ,.... I I I [SJW] [SJWO] [S/WO) 0 0 ,.... C\1 I I 0 C') I a. Q) (/) Cl ::J <( "3 ""') Fig. 5 .3a Time series ofNOAA buoy winds, AS1/CM2 mid-depth currents (50m isobath} and LB3/EC4 mid-depth currents (20m isobath) from January to September 1998, sampled daily after low-pass filtering to remove oscillations at time scales shorter than 36 hrs. 143

PAGE 177

"'0 -c: cU .0 0 >. !!2 0 E :::l a:l 0 <( L!) <( < 0 'E z cU c. :; E () cU I-Q) c. "'0 I "'0 L!) 0 0 0 0 0 0 0 ,.... ,.... C') C\J ,.... C\J C') I I I I I [S/W) [SfW::>) cu .0 0 !!2 E 0 C\J < 'E :::l () Q) c. "'0 I "'0 0 0 0 0 0 C') C\J ,.... ,.... I [SfW::>) 0 0 C\J C') I I c. Q) en s ...., c: :::l ...., c. <( (ij .0 Q) LL c: cU ...., 0> 0> 0> Fig. 5.3b Time series ofNOAA buoy winds, AS1/CM2 mid-depth currents (50m isobath) and LB3/EC4 mid-depth currents (20m isobath) from January to September 1999, sampled daily after low-pass filtering to remove oscillations at time scales shorter than 36 hrs. 144

PAGE 178

145 appear to be similar during these two years. The synoptic scale variability is generally largest in winter and early spring, tapering off to smaller values in summer. However, the wind speeds in 1998 are slightly larger than in 1999, and the durations over which the winds are upwelling favorable are slightly longer. Beginning in May 1998 there is also an approximate 2-month period in which the winds are upwelling favorable along the northern Florida Panhandle with no counterpart in 1999. The currents are also different, particularly at the 50 m isobath. During the first two months of 1998 these currents fluctuate with the winds whereas sometime in March they become unidirectional, flowing along-shore toward the southeast for the remainder of the record. No such behavior is seen in 1999. While there is a tendency in spring for southeast flow by local winds and surface heat flux (HW02), both the magnitude and duration of these spring transition southeast currents in 1999 are decidedly different than in 1998. The 20 m isobath currents also show different behaviors between these two years. They are generally stronger in 1998 than in 1999, May 1998 shows strong unidirectional currents in opposition to the winds, and the summertime currents are also stronger. Currents are also available at the 30 m isobath offshore of Panama City in the Florida Panhandle for the period March-September 1998 (Figure 5.4). There the currents in March and April show reversals consistent with the synoptic scale winds, but the currents from May to August are nearly unidirectional and in the same sense as the flows farther south (Figure 5.3a). How these anomalous WFS circulation behaviors partition between local and offshore forcing are the topics of the next two sections. The hydrography also differs between these two years. HW02 chose 1999 for their study of local forcing effects on the spring transition because the isotherms seaward

PAGE 179

Mid-dpeth Current At 30m Isobath off Panama City 1 6 11 16 21 26 31 5 10 15 20 25 30 5 10 15 20 25 30 4 9 14 19 24 2 9 4 9 14 19 24 29 3 8 13 18 23 2 8 1998 Fig. 5.4 Time series ofNOAA buoy winds and PA mid-depth currents (30 m isobath) from March to August 1998, sampled daily after low-pass filtering to remove oscillations at time scales shorter than 36 hrs 146

PAGE 180

147 of the shelf break were relatively flat. 1998 is different. Figure 5.5 provides temperature sections sampled offshore of Sarasota on March 30, May 18, June 10, July 7 August 5, September 9, and November 9. The first two show inner-shelf thermoclines that literally intercept the beach. The offshore extent of the thermocline structure is seen in June, when steeply sloping isotherms are observed at the shelf-break. While the isotherm slopes relax seaward of the shelf-break in July, the thermocline persists over the entire inner-shelf, and the coldest waters there are separated from similar temperature waters seaward of the shelf-break. We surmise that the origin of these cold inner-shelf waters is from farther north It is not until August that more typical summer conditions appear over the inner-shelf. However, upwelling of cold water is then observed seaward of the shelf-break. A relatively shallow thermocline persists into September, and November shows remnants of thermal stratification well into late fall (a tropical storm canceled our October cruise). With monthly cruises now available over three successive years we contend that the hydrography for the spring and summer seasons of 1998 (like the currents) are different from what may be considered normal. How can we reconcile these differences, and what are the implications? 6.4. MODEL SIMULATION WITH LOCAL FORCING ONLY 6.4.1 MOTIVATION We begin by examining the role of local forcing, independent of the LC. Motivation follows from the winds. Using the National Center for Environmental Predication (NCEP) r e an a ly s is, Figure 5 6 compares th e s pring season (March, April, May) average wind fields over the eastern Gulf of Mexico for climatology 1998, and

PAGE 181

0 0 -5 -5 -10 -10 -15 -1 5 -20 -20 -25 -25 ,, ;:. .... -40 -30 -20 -10 0 -40 -30 -20 -10 0 0 0 -50 -50 -100 -100 -150 -150 -150 -100 -50 0 -150 -100 -50 0 0 -50 E -20 -100 .c a. Q) 0-30 -150 -40 -150 100 -50 0 80 -60 -40 -20 0 Offshore Distance [km] 0 -50 -100 USF SA Transect -150 11/9 -150 -1 00 -50 0 Fig. 5.5. Temperature sections from the Mote Marine Lab (March 30th and May 1 81h) and the USF [June I Qth (SA) ; July 7th (SA), August 5th (SA), September 9th (TA), and November 91h (SA)] hydrographic transects. 148

PAGE 182

30 29 28 27 26 25 -90 -88 -86 -84 -82 -80 Fig. 5.6. Spring season averaged wind fields for 1998, 1999, and climatology (average of 50-year NCEP reanlysis wind fields) 149

PAGE 183

150 1999 Climatology (average of 50 years NCEP reanalysis wind fields) and 1999 are similar, whereas 1998 differs, showing southwesterly, upwelling favorable winds along the northern part of the shelf. These anomalous spring season average winds in 1998 result from the subtle differences in the otherwise regular synoptic scale variations during that year (Figures 5.3a and 5.3b). The larger magnitude and longer lasting periods of upwelling favorable winds in 1998 over those in 1999 account for the seasonal average differences. 5.4.2 MODEL AND FORCING FIELDS The model strategy is similar to the HW02 spring 1999 analysis. The primitive equation, Princeton Ocean Model (POM) [Blumberg and Mellor, 1987)] is forced by NCEP reanalysis winds and net surface heat fluxes, along with river inflows. The only role of the adjacent Gulf of Mexico (in this section) is to set the vertical distribution of temperature and salinity for initializing the model density field. Once begun, the integration proceeds solely on the basis of local forcing. The model domain (Figure 5.7) extends from the Florida Keys in the southeast to west of the Mississippi River in the northwest, and it has one open boundary arcing between where a radiation condition (Orlanski, 1976) is applied. The major rivers impacting the WSF are input along the coast using the technique of Kourafalou et al. (1996). Horizontal resolution on an orthogonal curvilinear grid varies from less than 2 km near the coast to 6 km near the open boundary A sigma coordinate system is used in the vertical with 21 layers distributed non-uniformly to accommodate surface and bottom Ekman layers. Horizontal diffusivities are parameterized by the Smagorinsky (1963)

PAGE 184

151 formulation with a coefficient of 0.2. Bottom stress, Tt,, is calculated by a quadratic law using a variable drag coefficient with a minimum value of 0.0025. Time steps of 6 seconds and 360 seconds are respectively employed for the model external and internal modes (Blumberg and Mellor, 1987). The model begins at rest on February 28, 1998, initialized with the same horizontally uniform stratification as in HW02 (based on climatology below 200 m depth and a March 1999 ECOHAB Program hydrographic survey above 200 m). Any deviations from the HW02 spring 1999 results must therefore be a consequence of changes in local forcing since everything else is the same. From this initial state of zero baroclinicity, the model spins up rapidly, generating baroclinicity in balance with the local wind and buoyancy forcing, as previously justified in HW02. Tidal forcing is excluded in the present application for two reasons. First, we are not considering high frequency variability, and second, with magnitudes of only a few em s-1 tidal currents on the WFS are not essential contributors to vertical mixing. A comparison between modeled and observed tidal currents and their mixing efficacy are given in He and Weisberg (2002d). The wind and heat flux forcing fields begin with NCEP daily reanalysis products. These are interpolated onto our model grid from coarsely resolved 2.5X2.5 maps. Comparisons between the NCEP winds and those observed by in-situ buoys are generally very good (the buoy winds are included in the NCEP reanalysis). An exception is the comparison at Venice (located midway between Tampa Bay and Charlotte Harbor) where the NCEP wind speeds from mid-April to early-August under-represent the Venice wind speeds by about a factor of two. This discrepancy corresponds to the period when SST is

PAGE 185

30 29 Q) 28 "'0 :::1 -...J 27 26 25 Model Grid -91 -90 -89 -88 -87 86 -85 -84 -83 -82 -81 -80 Longitude Fig. 5.7. West Florida Shelf model domain. 152

PAGE 186

153 anomalously cold along the coast between these two locations. Such coastal boundary layer problems for the NCEP fields are observed elsewhere (e.g., Baumgartner and Anderson, 1999). To correct for it we apply a spatial ramp-up in wind speed by a factor of two from just north of Tampa Bay to just south of Charlotte Harbor and offshore to about the 50m isobath. Elsewhere the winds are as given by NCEP. Heat flux forcing is also modified to account for SST features that are unresolved by the coarse NCEP fields. We use a relaxation scheme (e.g., Ezer and Mellor, 1992; Chu et al, 1999) whereby a correction term is added to the surface heat flux that tends to relax the model SST to an interpolated product of satellite monthly-mean SST. 5.4.3. RESULTS A comparison between the model end of spring (May 31) states for 1998 and 1999 (Figure 5.8) shows that local forcing can account for inter-annual variations in WFS circulation and water properties. For the velocity field we see that the season and depth average currents are larger and displaced farther offshore in 1998 than in 1999. As a consequence of the stronger spring currents in 1998 than in 1999, low salinity water (of Mississippi River origin) is transported farther south along the WFS in 1998 than in 1999 without any influence by the LC. The evolution of the fields is illuminating. Figures 5.9, 5.10, and 5.11 compare modeled and observed currents at the 50m, 20m, and 30m isobaths offshore of Tampa Bay, Sarasota, and Panama City, respectively, for the periods March 1998-August 1998. The formats are all the same. We show velocity vectors sampled in nature and in the

PAGE 187

154 model near the surface, at mid-depth, and near the bottom, along with an example of the winds (from NOAA buoy 42036). The model vectors are daily averages, as are the winds. The observed vectors are low-pass filtered to remove tides and inertial oscillations and then daily averaged. Sampled depths vary with location because of ADCP blanking and side lobe reflection distances. For each location we choose the bin closest to the surface or bottom deemed to be good (typically 3-4m from the surface or the bottom) and the one closest to mid-depth. The model is sampled to match these locations. The 50 m site shows excellent agreement during the first three synoptic scale weather cycles. Departures are evident by the end of March when the modeled currents continue to oscillate with the winds, whereas the observed currents tend to be more unidirectional. These departures are most evident at mid-depth and near the bottom. Near the surface, the model and data tendencies are similar, although the magnitudes are different. The more limited data at the 20m isobath also show that something is slightly amiss. Here the model and the data agree for a little longer, through the end of April (Weisberg et al., 2001), but at that point the data show a southward direction in opposition to the wind, whereas the modeled currents tum toward the north with the wind. In July and August, the model and data once again co-vary seemingly in response to the wind. This finding that discrepancies develop between the model responses to local forcing and the observations along the west Florida coastline contrasts with findings farther north in the Florida Panhandle. Offshore of Panama City at the 30 m isobath we see excellent agreement between the model and data throughout the record, including the period of unidirectional tendency from May through mid-July. Moreover, the velocity

PAGE 188

28 27 2s Seasonal and Depth Averaged Current 25 28 27 26 Surface Temperature 5/31/1998 25 30 29 28 27 26 Surface Salinity 5/31/1998 25 Surface Temperature 5/31/1999 Surface Salinity 5/31/1999 Fig. 5.8. Modeled currents, sea surface temperature and sea surface salinity fields under local forcing only, compared for the spring seasons of 199 8 (left panels) and 1999 (right panels). From top to bottom in both panels are the seasonal and depth averaged currents, and the modeled sea surface temperature and sea surface salinity fields sampled on May 31'1. 155

PAGE 189

10 I 0 -10 0 2 (i) 0 -0. 2 0 2 I 0 -0.2 0 2 Data CM2 Mid-depth I 0 -0.2 0 2 I 0 -0.2 0.2 Data CM2 Bottom I 0 -0.2 0.2 (i) l 0 0.2 1 6 11 16 21 26 31 5 10 15 20 25 30 5 10 15 20 25 30 4 9 14 19 24 29 4 9 14 19 24 29 3 8 13 18 23 28 Mar Apr May Jun Jul Aug 1998 Fig. 5.9 Comparisons between the observed and the modeled currents at the 50m isobath (ASIICM2) sampled near-surface, mid-depth, and near bottom along with the NOAA buoy 42036 winds for the case of local forcing only. The vector time series are sampled daily after low-pass filtering to remove oscillations at time scales shorter than 36 hrs 156

PAGE 190

Model EC4 Bottom 1 6 11 16 21 26 31 5 10 15 20 25 30 5 10 15 20 25 30 4 9 14 19 24 29 4 9 14 19 24 29 3 8 13 18 23 28 Mar Apr May Jun Jul Aug 1998 Fig. 5.10. Comparisons between the observed and the modeled currents at the 20m isobath (LB3/EC4) sampled near-surface, mid-depth, and near-bottom, along with the NOAA buoy 42036 winds for the case oflocal forcing only. The vector time series are sampled daily after low-pass filtering to remove oscillations at time scales shorter than 36 hrs. 157

PAGE 191

10 I 0 -10 0.2 I 0 0.2 -0.4 0.2 .., 0 0 2 -0.4 0 2 I 0 0 2 -0.4 0.2 I 0 -0. 2 -0.4 0 2 Data Mid-depth Current i 0 0 .2\ > 'I '\'I! '(' \\\ ....... 0 2 Model Bottom Current I -0. 2 0 4 +.--r-r-lr-r---r-r.--r-r-l--r-r-,--.,.--,...,r-r---r-r-.-.-,.---,----r-r-r--.--r-,--,----r--r-r-....--r-l1 6 11 16 21 26 31 5 10 15 20 25 30 5 10 15 20 25 30 4 9 14 19 24 29 4 9 14 19 24 29 3 8 1 3 18 23 28 Mar Apr May Jun Jul Aug 1998 Fig 5.11. Comparisons between the observed and the modeled currents at the 30m isobath (PC) sampled near-surface, mid-depth and near-bottom along with the NOAA buoy 42036 winds for the case of local forcing only. The vector time series are sampled daily after low-pass filtering to remove oscillations at time scales shorter than 36 hrs. 158

PAGE 192

159 and temperature fields observed by a shipboard survey over the northern half of the study region between May 5th-16th agree very well with the model fields sampled on May lOth. Figure 5.12 shows the bottom temperature distributions. The observed and modeled patterns are similar, each showing upwelling at DeSoto Canyon and relatively cold water penetrating onto the shelf in the Big Bend. A small discrepancy exits in magnitude with the model bottom temperatures being about a degree warmer than the observations. Figures 5.13 and 5.14 show the velocity fields at depths of 14m (the uppermost bin from the ship) and 50 m, respectively. The observed and modeled patterns at 14m depth compare very well, particularly along the northern part of the shelf and southward along the shelf-break. The model field is more coherent as expected since it is a synoptic sample, whereas the shipboard survey is sampled over an 11-day interval. Two findings are notable. First, the modeled currents offshore of Tampa Bay nearest the coast are flowing north whereas the currents farther offshore are flowing south. Given this across shelf gradient, a small shoreward shift in the current field would increase the magnitude of the modeled currents at the 50 m isobath to be in better agreement with the moored observations (Figure 5.9). Since the nature of the spatial distributions is largely baroclinic, we see that it is imperative to get the density field correct in order to get the model currents correct. From this we surmise that local forcing alone did not produce the baroclinic structure well enough. Second, note the DeSoto Canyon eddy in both the observations and the model, particularly at the 50 m depth. Such eddies are often attributed to the LC. Here we see that the canyon itself, by virtue of potential vorticity conservation, can generate its own eddy, independent of the LC. As fluid attempts to

PAGE 193

\l .. s :rrN I.JCPW !WW 88"W ST"W xsw 83"W 31 -., 30. 5 .. .. 5/10 (B) 30 : "' , 29.5 \ \ 29 .. 28.5 28 {':) 27.5 27 -90 -89 -88 -87 -86 -85 -84 -83 Fig. 5.12. Comparison between the NEGOM measured (May 6-15) and the modeled (sampled on May 1 0) bottom temperature fields for the case of local forcing only. 160 &r-'W -82

PAGE 194

161 JO'N 29'N 28'N 50 cms1 27'N 90'W 89'W 88'W 87'W 86'W 85'W B4'W 83'W 82'W -89 -88 -87 -86 -85 -84 -83 -82 Fig 5 13 Comparison between the NEGOM measured (May 6-15) and the modeled (sampled on May 10) subsurface velocity fields at 14m depth for the case oflocal forcing only.

PAGE 195

. \fi 'N i J + \ .. v \ I ., \ I \ \ \ l ' \ r:. ""' r .J F,lamp.1 \}f) ''-::7' N-t----,----,----,-----r-----i-----,------.--_i_---! M 7 '' W 30.5 > 30 29.5 28.5 27. 5 -89 -88 -87 -86 -85 -84 -83 -82 Fig. 5.14. Comparison between the NEGOM measured (May 6-15) and the modeled (sampled on May 10) subsurface velocity fields at 50 m depth for the case oflocal forcing only. 162

PAGE 196

163 navigate the right angle bend in bottom topography it flows uphill and generates negative relative vorticity, resulting in a clockwise eddy. Despite some discrepancies these temperature and velocity field comparisons show that the modeled patterns (in response to local wind and buoyancy forcing alone) are generally correct. To see why this is true and to draw a contrast with 1999, it is instructive to look at a particular upwelling event in March 1998. We choose the upwelling/downwelling sequence that begins on March lOth, a time by which the initial temperature distribution is adjusted to the surface fluxes. Sequences of across-shelf sections for temperature and for the along-shelf and across-shelf velocity components are shown for DeSoto Canyon, where the shelf is narrowest, in Figure 5 15a, and in the Big Bend, where the shelf is widest, in Figure 5.15b. These snapshots, sampled from a movie animation, demonstrate the appearance of cold water on the shelf by local forcing. March lOth is a transition day from a previous downwelling period At DeSoto Canyon we see onshore flow at all depths and an accelerating eastward coastal jet near-shore. By March 12th the coastal jet peaks, strong onshore flow continues at the shelf break, and deep waters of 18 19 C temperature are upwelled onto the shelf (where waters of this temperature are also evident near-shore by negative surface heat flux). Upwelling continues, and by March 14th the entire DeSoto Canyon shelf is in the range of 18-19C and waters of 17-18 C are broaching the shelf-break, seaward of which the isotherms slope down to their initialized depths Downwelling then commences, peaking on March 19th. The near-shore temperatures remain cold, however, by a combination of surface hear flux and along-shore Inspection of similar animations for spring 1999 shows no upwelling events of sufficient magnitude or duration to bring deep waters

PAGE 197

3/10 (J 3110 -200 -200 -150 -100 -50 0 -200 -150 -100 -50 0 -200 -150 -100 -50 0 -150 -150 -100 -50 0 -200 -150 -100 -50 0 -200 -150 -100 -50 0 -50 I 5 -100 -150 3/14 -200 -200 -150 -100 -50 0 -200 -150 -100 -50 0 -200 -150 -100 -50 0 150 100 -50 0 f ; -200 -150 -100 -50 0 -200 -150 -100 -50 0 Offshore Dlslance [km) Fig. 5.15a. Across-shelftransects ofthe modeled temperature, across-shelf(u) and alongshelf (v) velocity components sampled from a movie animation at DeSoto Canyon on March 1 Qth, 12th, 14th, and 19th for the case of local forcing only. 164

PAGE 198

Temperature [ C] -50 -150 3/10 -200 '------4---------4 -400 -300 -200 -100 -50 I = -100 -150 18 17 16 200 '------'---------------4 -400 -300 -200 -100 -50 I E. -100 0 150 0 3/14 -200 -400 -300 200 -100 50 I -100 -150 0 -200 -400 -300 200 -100 0 -400 -300 -200 -100 -400 -300 -200 -100 0 -400 -300 -200 -100 0 -400 -300 -200 -100 0 -400 -300 -200 -100 0 400 -300 200 -100 0 -400 -300 -200 -100 0 400 -300 -200 -100 0 Offshore Oislance [km] Fig 5.15b. Across-shelftransects of the modeled temperature, across-shelf(u) and alongshelf (v) velocity components sampled from a movie animation at the Big Bend on March 1 Qth, 12th, 14th, and 19th for the case of local forcing only. 165

PAGE 199

166 colder than 19-20C onto the shelf. The subtle changes in upwelling favorable wind magnitude and duration between spring 1998 and 1999 are sufficient to account for the markedly colder temperatures along the Aorida Panhandle in 1998 than in 1999. Local surface heat flux also contributes to SST. The Big Bend, where the shelf is widest, is the primary manufacturing site for cold water of shallow origin. Thus, 18C water is already present near-shore in the Big Bend on March 1Oth prior to upwelling commencement, and this water cools further as the cold front passes. The coldest waters found near-shore in the Big Bend remain separated from the coldest waters upwelled at the shelf-break. While not shown, a Cape San Blas section located between the DeSoto Canyon and Big Bend sections demonstrates that an advective connection does not occur over this synoptic scale event. Two sources of cold water therefore exist. A deep source by upwelling and a shallow source by surface heat flux. Waters found later in spring and early summer within the WFS cold tongue (e.g., HW02) result from a mixture of these to sources The nutrient content must also be a mixture of deep-ocean and land-derived sources From these findings we surmise that cold water by itself is neither indicative of upwelling, nor of high productivity potential. Moving farther south to where we have ECOHAB hydrographic sections (Figure 5.5) we find that the model forced by local winds and heat fluxes alone does not adequately account for the thermal structure. Figure 5.16 shows that we fail to account for stratification in March and April, the stratification in May is too weak and the bottom waters are too warm These and other discrepancies are discussed in the next section.

PAGE 200

-20 -40 -60 -80 .c a -1oo -120 -140 -160 -180 -40 -60 ',':..'.' -80 -120 -140 -160 -180 -250 -200 -150 -100 -50 0 Oflshore Distance (kmJ Fig. 5.16. Modeled temperatures along the Sarasota transect sampled on March 30th and May 181h for the case of local forcing only. 167

PAGE 201

5.5. MODEL SIMULATION WITH LOCAL AND IDEALIZED LC FORCING 5.5.i MOTNATION 168 The partial closure of the WFS by the Florida Keys in the south results in a convergence of isobaths west of the Dry Tortugas. Hetland et al. (1999) hypothesize that LC impacts at the shelf-break in the south set the shelf waters in motion by influencing relatively shallow isobaths; this being in contrast with LC impacts farther north where such shallow isobaths are more distant from the shelf break. The Meyers et al. (2000) analysis of velocity data from a 1995 across-shelf moored array neither supports, nor rejects this hypothesis. Three LC impacts are observed in this data set, but apparently none occurring sufficiently far south. He and Weisberg (2002c) describe a LC intrusion in June 2000 for which shipboard hydrography and moored velocity data are available. Currents at the shelf break due to the LC and currents on the inner-shelf due to the winds are both observed and modeled to be independent. Inspection ofTOPEX-Poseidon (TIP) satellite altimetry data for the period 1992 through the present (Figure 5.17) offers clarification. Both the 1995 and 2000 observational periods are absent large sea level gradient anomalies over the southern WFS. Only four southern comer impact events stand out robustly over the entire nine year record: October 1996-March 1997, August 1997 November 1997, March 1998-July 1998, and September 1998 December 1998, the third one of which corresponds to the analysis period herein. While anomalous, these post-LC eddy shedding impingement events on the southern comer as hypothesized by Hetland et al. (1999), appear to leave their mark on the shelf.

PAGE 202

Q) "'C -88 -86 -84 -82 -80 Longitude 30 jj!JI I \1 I I lr.}l It,' I 1/1 \J!\Jl IQ IJ I !(JI j\ I I Jlifi(JI I I Ill I I ... '-t'\ I I I I jl !j I I 1 1 jl 1 \'.....! 1 !j 29 28 27 j 26 25 NJ MMJ SNJ MMJ SNJ MMJ SNJ MMJ SNJ MMJ SNJ MMJ SNJ MMJ SNJ MMJ SNJ MMJ SN 19921993 1994 1995 1996 1997 1998 1999 2000 2001 Fig. 5 17. TOPEXIPOSEIDON sea surface height anomalies sampled along track 26 from November 1992 to November 2001 Shading denotes positive anomalies. The contour interval is 0.2 m. ,_. 0\ \0

PAGE 203

170 5.5.2 MODEL AND FORCING FIELDS We employ the technique of He and Weisberg (2002c) to assess the role of the LC on the WFS during the spring and summer seasons of 1998. Using the identical density field initialization and local wind and buoyancy forcing functions as in section 6.4, we control the surface height field along the open boundary in the south to allow for geostrophic inflows and outflows consistent with the TIP data. The mode splitting solution of the POM allows for rapid baroclinic adjustment since all of the vertically integrated transport must be in the external mode. Thus, relatively strong (weak) currents exist above (below) the thermocline offshore of the shelf break in addition to the net transport that is contained in the external barotropic mode. The model interior solution freely evolves in response to these fully baroclinic pressure and current fields imposed at the open boundary Interior adjustment is also rapid With celerity of order m s-1 the gravest baroclinic shelf waves traverse the model domain from the Florida Keys to the Mississippi River (about 1000 km) in about 10 days, a period comparable to the TIP repeat interval, and barotropic modes propagate much more rapidly By controlling the open boundary our LC model experiments are somewhat arbitrary. Three controlling factors are: 1) the amplitude of the imposed height field, 2) the location and width of the inflow-outflow window, and 3) temporal modulation Several experiments were performed in which we varied these factors to arrive at a solution sufficient to describe the salient observational features. Thus, we employed a time independent Gaussian shaped pressure perturbation of magnitude 0.075 m interpolated over 17 grid points west of the Florida Keys. Given the repeat interval and smoothing limitations of the TIP data and the fact that the combination of open boundary

PAGE 204

171 and bathymetry in the southern comer of our model domain was not optimally designed for this experiment, finer input tuning is neither justified, nor necessary. Further improvements will require a model of larger domain, of which there are several, but none that purport to get the LC evolution correct either. 5.5.3 RESULTS Adding the LC modifies the model end of spring (May 31) states for 1998 (Figure 5. 18) from case without the LC (Figure 5.8). The seasonal and depth averaged mean currents are more coherently developed over the entire shelf, near-shore SST is cooler, and the SSS core is positioned closer to the coast by virtue of the mean current distribution being shifted shoreward While the magnitude of the SSS pattern is largely unaltered the SSS is decreased over a larger portion of the shelf by virtue of larger mean advection. Adding the LC improves the fidelity between the modeled and observed currents, particularly at the 50 m isobath (Figure 5.19) The modeled and observed current speeds are in better agreement near the surface than under local forcing only (Figure 5.9), and, more importantly, the unidirectional tendency of the currents at d e pth also agrees now. Unlike the Hetland et al. (1999) barotropic argument, however our results are strongly baroclinic. At the 20m isobath (Figure 5 20) the modeled currents are also improv e d rel a tive to the data over the initial portion of record, but less dramatically so when compared with the 50 m isobath currents. These 20 m isobath currents remain largely locally forced with only a small increase in southward t en d enc y Over the l a tter portion of th e record th e

PAGE 205

28 27 26 Seasonal and Depth Averaged Current 25 28 27 26 Surface Temperature 5/31/1998 25 29 28 27 26 25 Surface Salinity 5/31/1998 Surface Temperature 5 / 31/1998 Surface Salinity 5/31/1998 -91 90 -89 -88 8 7 -86 8 5 -84 83 -82 -81 80 -91 90 -89 -88 87 -86 85 84 -83 82 -81 80 Fig. 5.18. Modeled currents, sea surface temperature, and sea surface salinity fields for the spring season of 1998. The left hand panels are for the case of local forcing only; the right panels are for the case of local plus Loop Current forcing. From top to bottom in both panels are the seasonal and depth averaged currents, and the modeled sea surface temperature and sea surface salinity fields sampled on May 3151 172

PAGE 206

10 i 0 -10 0.2 i 0 -0.2 0.2 i 0 -0.2 0.2 i 0 -0.2 0.2 fij' 0 -0. 2 0.2 Data CM2 Bottom fij' 0 -0. 2 0 2 fij' 0 -0. 2 1 6 11 16 21 26 31 5 10 15 20 25 30 5 10 15 20 25 30 4 9 14 19 24 29 4 9 14 19 24 29 3 8 13 18 23 28 Mar Apr May Jun Jul Aug 1998 Fig. 5 19. Comparisons between the observed and the modeled currents at the 50m isobath (AS 1/CM2) sampled near-surface mid-depth, and near-bottom, along with the NOAA buoy 42036 winds for the case oflocal plus Loop Current forcing. The vector time series are sampled daily after low-pass filtering to remove oscillations at time scales shorter than 36 hrs 173

PAGE 207

10 I 0 -10 0 2 7ii l 0 -0. 2 0.2 7ii l 0 -0. 2 0.2 I 0 -0.2 0.2 Model EC4 Mid-depth I 0 -0.2 0.2 I 0 -0. 2 0.2 Model EC4 Bottom I 0 -0.2 1 6 11 16 21 26 31 5 10 15 20 25 30 5 10 15 20 25 30 4 9 14 19 24 29 4 9 14 19 24 29 3 8 13 18 23 28 Mar Apr May Jun Jul Aug 1998 Fig 5.20. Comparisons between the observed and the modeled currents at the 20m isobath (LB3/EC4) sampled near-surface, mid-depth, and near-bottom, along with the NOAA buoy 42036 winds for the case oflocal plus Loop Current forcing The vector time series are sampled daily after low-pass filtering to remove oscillations at time scales shorter than 36 hrs. 174

PAGE 208

175 solution under local forcing only (Figure 5.10) actually looks better than the solution with the LC added. The explanation resides in the density field were we see an extremum in the thermocline height located between the 50 m and 20 m isobaths. The reversal in the thermocline slope across the extremum accounts for the different behaviors at these two isobaths. This thermocline convexity is an important aspect of the shelf response to both local and offshore forcing because the sign of the baroclinic circulation reverses across it. Modeling exactly where the convexity is centered is not a reasonable expectation given the limitations in both the TIP data and model geometry. The 30 m isobath location in the Panhandle shows minimal change in the modeled currents with or without the LC addition (Figure 5.21 with the LC compared to Figure 5.11 without). The anomalous unidirectional currents at this location are primarily locally forced. Nevertheless, the LC incident on the southwest comer of the WFS still plays an important role in the Panhandle and (by connection) over the entire shelf. It does this by increasing the slopes of the isotherms (or the slopes of any other material property isolines) at the shelf-break, thereby increasing the potential for deep-ocean water properties to be upwelled onto the shelf by local forcing These effects are evident in movie animations, which the following narrative attempts to describe. The upwelling/downwelling sequence beginning on March lOth (described in section 3 for the case of local forcing only) proceeds very similarly with the LC added. The water that is upwelle d onto the shelf is about 1C colder. By March 14th the entire DeSoto C a nyon shelf is in the range of 17-18C and waters of 16 -17C are broaching the shelf-break. The ascent of these waters from their initial state is about 100 m. Waters of 17 18 C also erupt to the surfac e in the Big Bend independent o f th e

PAGE 209

10 i 0 -10 0.2 i 0 -0.2 -0.4 0.2 Ui' 0 -0. 2 -0.4 0.2 Ui' 0 -0. 2 -0.4 0.2 Model Mid-depth Current i 0 -0. 2 -0. 4 0 2 Data Mid-depth Current i 0 '\'\'! 'f W'\t "' -0.2--0.4 II 0.2 Model Bottom Current i -0.2 -0.4 1 6 11 16 21 26 31 5 10 15 20 25 30 5 10 15 20 25 30 4 9 14 19 24 29 4 9 14 19 24 29 3 8 13 18 23 28 Mar Apr May Jun Jul Aug 1998 Fig. 5.21. Comparisons between the observed and the modeled currents at the 30m isobath (PC) sampled near-surface, mid-depth, and near-bottom, along with the NOAA buoy 42036 winds for the case of local plus Loop Current forcing. The vector time series are sampled daily after low-pass filtering to remove oscillations at time scales shorter than 36 hrs. 176

PAGE 210

177 DeSoto Canyon. Since the Big Bend shelf-break is about 20m deeper than the DeSoto Canyon shelf-break (60 m versus 40 m) less upwelling is required for a given isotherm to broach the Big Bend shelf-break (HW02). Each of the March frontal passages results in similar upwellings of deep-ocean water beyond the shelf-break so that vertical stratification is established on the WFS by March 30th, although not as strongly as observed. April shows two major upwellings culminating on the 12th and the 24th. Since the net surface heat flux is into the ocean in April all of the cold water either modeled, or observed during these events comes from depth. May is a month of nearly continuous upwelling and increasing vertical stratification as upwelling (surface heat flux) cools (warms) the water below (above) the thermocline. Figure 5.22 gives an example on May 15th where we see 18-19C waters in the Big Bend and a well-defined thermocline offshore of Sarasota (as in Figure 5.5). The Big Bend also appears to be a source of cold water for points farther south as seen in the May 15th near bottom velocity and temperature fields of Figure 5.23, the latter of which agree with the shipboard observations in Figure 5.12. Upwelling circulations persist throughout June as the thermocline continues to sharpen. Persistent eastward flow across the DeSoto Canyon region eventually leads to the generation of a semi-permanent eddy that straddles the right angle bend in topography there. In contrast to eddies that are directly spun off from the LC, this eddy is a consequence of persistent flows over abrupt topography. The magnitude of the upwelling circulation tendency begins to taper off in mid-July as a consequence of wind forcing. Continuing at smaller values, in part due to the continuing LC influence, stratified conditions persist across the WSF. By August, and as evident in Figure 5.24 for the

PAGE 211

-20 -40 -60 I -so -100 a. Q) 0 120 -140 -160 -180 -40 60 I -80 ..c a. -100 Q) 0 -120 -140 160 180 -300 Big Bend 5/15 Sarasota 5/15 -250 -200 150 -100 -50 Offshore Distance [km] 0 Fig. 5.22. Mode led temperatures along the Big Bend and Sarasota sections sampled on May 15'h for the case of loca l plus Loop Current forcing. 178

PAGE 212

-90 -89 -88 -87 -86 -85 -84 -83 -82 -81 30 29 28 27 26 25 -90 -89 -88 -87 -86 -85 -84 -83 -82 -81 Fig. 5.23. Modeled bottom velocity and temperature fields sampled on May 15th for the case of local plus Loop Current forcing. 179

PAGE 213

0 -20 -40 -60 -80 I .c: 15. 100 0 -120 -140 -160 -180 -300 -250 200 -150 -100 -50 0 Fig. 5.24. Modeled temperature sampled at the Sarasota transect on Aug. S1h for the case of local plus Loop Current forcing. 180

PAGE 214

181 Sarasota transect, both the model and the observations (Figure 5.5) show similar patterns and magnitudes. 5.6. DISCUSSION Anomalous circulation and water properties, manifest in colder temperatures and increased stratification, are observed on the WFS in 1998 relative to 1999. Through numerical model simulation in comparison with data, Section 6.4 demonstrates that much of this behavior is attributable to subtle variations in local, shelf-wide forcing (larger magnitude and longer lasting upwelling favorable winds, particularly along the Panhandle coast). Increased wind-driven upwelling brings deeper, colder water onto the shelf and maintains cold conditions well into summer, at least in the northern portion of the model domain. Another contributing factor is required farther south, however, to account for the persistent, deep southward advection and stratification found there. Thus, we revisited the Hetland et al. (1999) hypothesis in section 6.5 that a LC impingement on the southwest corner of the WFS (where isobaths converge due to the Florida Keys) can set the WFS currents in motion by influencing relatively shallow isobaths. Our results confirm this hypothesis. We note that the Hetland et al. (1999) hypothesis is a barotropic one, whereas our results (observed and modeled) show a highly baroclinic environment. These findings are not incompatible. Local forcing, independent of the LC, is what enables deep-ocean water properties to broach the shelf. Deep-ocean processes set the height of material isopleths in the vicinity of the shelf-break from which they may be upwelled onto the shelf by local forcing. Such deep-ocean processes include the LC and its eddies, either

PAGE 215

182 directly, or indirectly. The effect of the LC in 1998 was indirect. By impacting the shelf break in the southwest comer the LC set currents in motion along the entire shelf-break from the point of impact at the Florida Keys to the Mississippi River Delta consistent with the anticlockwise propagation of topographic Rossby waves (e.g., Gill, 1982, p408). The ensuing southward currents and sloping isopycnals along the shelf-break resulted in elevated material property isopleths requiring less upwelling to broach the shelf-break. This remote (or indirect) impact in the south facilitated increased cold water invasion by local wind forcing in the north Once carried onto the shelf the upwelled cold water conspired with surface heat flux to form the sharp thermocline that persisted into summer. As a consequence of these factors (local winds, surface heat fluxes, and the remote LC impact) the across-shelf thermocline height distribution developed a convex shape, which is the basis for the strong baroclinicity observed and modeled. Because the thermal wind shear sign reverses across the thermocline apex, the baroclinic and barotropic parts of the circulation can add either constructively or destructively, explaining why the observed and modeled currents at our 50 m and 20 m isobath locations behave differently from one another. Thermocline convexity is also a property of the spring transition in a normal year (HW02), but a LC impact to the south can shift the thermocline apex either landward or seaward depending upon the inner-most isobath impacted. A landward shift in the thermocline pattern increases the temperature at the shelf-break, and conversely. This is one reason why replicating the observed currents and thermocline is difficult and why we believe that further tinkering with the open boundary condition for the LC is unwarranted. The TIP data can neither detennine the extent to which isobaths are impacted nor the temporal amplitude modulation.

PAGE 216

183 The composition of the water beneath the thermocline apex is of major ecological importance. Since cold water originates both by upwelling over the shelf-break and by surface cooling near-shore the nutrient content of the cold tongue water beneath the thermocline apex is a mixture of deep-ocean and land-derived sources. Inter-annual variations in this source water mixture results in inter-annual variations in shelf ecology. During 1998 there was an anomalous amount of deep-ocean-derived nutrients on the WFS (Walsh et al., 2002) which is directly attributable to upwelling in the north. Why the anomalous nutrients relate to the deep-ocean source is easily explained on the basis of deep-water temperature and nutrient relationships. Figure 5.25 shows TIS, TIP, T/Si, and TIN relationships for the source waters of the Gulf of Mexico. The data are sub-sampled from the WOCE hydrographic section A22 along 66 W and between the latitudes of roughly 11 N to 25 N, i.e., the latitude range for waters that enter the Caribbean and hence the Gulf of Mexico. Waters of temperature below around 18 C show increasing nutrient levels at Redfield ratio that distinguish these from waters manufactured near shore by wintertime surface cooling. Walsh et al. (2002) show that during 1999 waters of such deep-ocean nutrient signatures are well below the shelf-break and that the nutrient isolines are co-linear with the isobaths. This contrasts with 1998 when these nutrient isolines cross isobaths and high values are found well on the shelf mimicking the bottom temperature findings of Figure 5.12. Recalling that the differences between our section 6.4 (local forcing only) and section 6.5 (local focing plus LC) bottom temperatures are about 1-2 C we may conclude that deep, nutrient rich waters would have upwelled onto the shelf with or without the indirect influence by the LC. Without indirect LC influence the nutrient levels would just be quantifiably lower based on the temperature-nutrient

PAGE 217

32 30 28 26 .. .... 024 i';i 20 __ -___ -___ __ __ -_____ ___ -_-_-__ __ a; 14 ' -.f,., 12 ,. 1f 33 34 35 36 37 38 Salinity [psu] 32 30 I.a.. 28 IT 26 8 6 4 2 \1--------------------.\,.., .1;. ... ...... .. r. : .'-'".....:.... _: ... ----0 5 10 15 20 25 30 35 40 45 50 Silicate [umoVKg] 32 30 28 26 .,. 024 l" ... ... 20 :. ::\._ :g_ 18 ______ E -:.."'(. (!!. 16 '""'"t -14 ..... E r" $ 12 0 D-1: :J\) : e 32 30 28 0 26 .. t.r.: 0 -;J" I ... :::J 20 : 1 2 3 Phosphate [umoVKg] 4 a.18 --E -:... .... (!!. 16 --------a; 14 = '. += .... 12 .,. 0 ...... D-10 e 8 6 4 2 0 10 20 30 Nitrate+Nitrite [umoVKg] 40 Fig. 5.25. Temperature/Salinity, Temperature/ Phosphate, Temperature/Silicate, and Temperature/(Nitrate+Nitrite) relationships based on WOCE hydrographic section A22 data sampled between the latitude range of 11 N to 25N. 184

PAGE 218

185 relationships. The subsequent distribution of these nutrients is another matter, however, which requires that we consider Lagrangian pathways. Synoptic wind forced variability alone leads to advective length scales of order tens of km on the WFS (Weisberg et al., 1996). Longer net displacements require persistent currents of seasonal or inter-annual origin. The seasonally varying circulation due to combined wind and buoyancy forcing leads to larger displacements, but we see in the 1998 model simulations (Figure 5.9) that these are limited to the near-surface at the mid-shelf, Tampa Bay location. Near-bottom, where the higher nutrient concentrations are found, large persistent currents are not achieved without adding the LC (Figure 5.19). To further quantify these findings we calculated both progressive vector diagrams (PVDs) and Lagrangian particle trajectories for neutrally buoyant particles advected by the fully three-dimensional flow field. Figure 5.26 compares modeled and observed results at the 50 m isobath mooring site offshore of Tampa Bay in the form of PVDs, each originating on March 151 and calculated through August 3151 1998. Comparisons are given at the near-surface, mid-depth, and near-bottom locations. The model without the LC (left hand panels) shows a substantial displacement near-surface only, whereas the data show substantial displacements at all depths. The near-surface model displacement is about half the observed amount. The LC addition (right hand panels) largely remedies these discrepancies. Displacements now exist at all depths with the near-surface displacements in quantitative agreement and the mid-depth and near-bottom displacements underestimated by about a third. A similar set of PVD comparisons for the 20m isobath offshore of Sarasota is given in Figure 5.27. While the differences between the results with and without the LC in the Eulerian presentations of Figures 5.10 and 5.19,

PAGE 219

186 -1000 (A) Near Surface -1500 (B) Near Surface -1500 -1000 -500 0 500 1000 -1 000 -500 0 500 1000 [km] [km] -500 -500 -1000 -1000 (A) Mid-depth -1500 (B) Mid-depth -1500 -1000 -500 0 500 1000 -1000 -500 0 500 1000 [km] [km] -500 -1000 (A) Near Bottom -1500 (B) Near Bottom -1500 -1 000 -500 0 500 1 000 -1 000 -500 0 500 1000 [km] [km] Fig. 5.26. Progressive Vector Diagram (PVD) comparisons between observed (thin lines) and modeled (thick lines) currents at the 50 m isobath (ASIICM2) for the six month period March to August 1998. The left hand panels are for the case of local forcing only; the right hand panels are for the case of local plus Loop Current forcing. In each case from top to bottom are the near-surface, mid-depth, and near bottom PVDs, respectively.

PAGE 220

-100 E .::.: -400 (A) Near Surface -400 -100 E .::.: -400 -200 0 200 400 [km) (A) Mid-depth -500 -400 E .::.: -400 -200 0 200 400 [km] (A) Near Bottom -500 -400 -200 0 200 400 [km) -100 E .::.: -400 (B) Near Surface -400 -100 E .::.: -400 -200 0 200 400 [km) (B) Mid-depth -400 -100 E .::.: -400 200 0 200 400 [km) (B) Near Bottom -WO -400 -200 0 [km) 200 400 Fig. 5.27. Progressive Vector Diagram (PVD) comparisons between observed (thin lines) and modeled (thick lines) currents a t the 20m isobath (LB3/EC4) for the six month period March to August 1998, irrespective of the data gap (Figs. 5.10 or 5.20). The left hand panels are for the case of local forcing only; the right hand panels are for the case of local plus Loop Current forcing. In each case from top to bottom are the near-surface, mid-depth, and near bottom PVDs, respectively. 187

PAGE 221

188 respectively, are subtle upon integration over three months they appear profound. Small persistent currents can give rise to large net displacements. However, these PVD displacements are unrealistically large, a known PVD limitation by virtue of assuming no along-track variation in velocity Although we are limited to PVDs for Eulerian comparisons with data, we can perform more realistic Lagrangian tracking with the model. Figure 5.28 shows modeled Lagrangian particle trajectories for release points in a near-bottom sigma layer (layer 18, or within 1 m of the bottom at the 50 m isobath) along the DeSoto Canyon, Cape San Bias, Big Bend, and Sarasota sections. Three-month interval trajectories are given in the left and right hand panels for particles released on March 151 and June 15 \ respectively The trajectories that feed the inner-shelf region between Tampa Bay and Charlotte Harbor are those that originate in the Big Bend consistent with the modeled near-bottom Eulerian representation on May 151h given in Figure 5.23 The displacements on the order of 3-4latitude degrees over three months correspond to persistent currents of order 0.1 m s -1 as observed. This trajectory "pipeline" corresponds to the region of coldest water formed by the mixture of upwelled nutrient rich deep water and near-shore shallow water. How the nutrients and phytoplankton vary en route following these trajectories from the shelf break to the mid-shelf and inner shelf regions under varying light conditions is the subject of Walsh et al. (2002) 6.7. SUMMARY AND CONCLUSION We consider the question of local versus deep ocean forcing in determining continental shelf water properties. The WFS, by virtu e of having both narrow (DeSoto C anyon) and wid e (Big Bend) regions and an adja ce nt bou n dary c u rre n t (th e Gulf o f

PAGE 222

Ensemble Tracks Ensemble Tracks 30 29 28 27 26 25 -90 -88 -86 -84 -82 -90 -88 -86 -84 -82 30 30 29 29 28 28 27 27 26 26 25 25 -90 -88 -86 -84 -82 -90 -88 -86 -84 -82 30 29 28 27 26 25 -90 -88 -86 -84 -82 -90 -88 -86 -84 -82 30 30 29 29 28 28 27 27 26 26 25 25 -90 -88 -86 -84 -82 -90 -88 -86 -84 -82 Fig. 5.28. Modeled trajectories for neutrally buoyant Lagrangian drifters released near the bottom along the DeSoto Canyon, Cape San Bias Big Bend and Sarasota transects on either March 1st (left hand panels) o r June 1st (right hand panels) and tracked for the respective three-month spring or summer seasons of 1998 under the influ ence of both local and Loop Current forcing. Bottom contours of20 m, 50 m, I 00 m and 200m are overlaid. 189

PAGE 223

190 Mexico LC) is an excellent natural laboratory in which to address this question of general continental shelf importance. We find that deep-ocean processes set the depths of material isopleths in the vicinity of the shelf-break. Local forcing is necessary for these materials to be upwelled onto the shelf. Once on the shelf local forcing is also the motive agent for distributing materials between the shelf-break and the inner-shelf with the bottom Ekman layer playing an important role. Distribution by local forcing may be augmented by deep-ocean influences under special circumstances. Such was the case during the anomalous spring through fall seasons of 1998. The spring transition in all years on the WFS features a "cold tongue" at mid-shelf that extends southward from Cape San Bias On the seaward side of the cold tongue are regions of low salinity and elevated ocean color. These features are explained on the basis of local, shelf-wide surface wind and heat flux forcing (Weisberg et al., 1996 and He and Weisberg, 2002a), independent of the LC. The cold tongue has a subsurface expression that extends into summer after surface heating paves over the surface layer. This subsurface cold water is of a mixture of waters of deep-ocean (by upwelling across the shelf-break) and near-shore (by wintertime cooling) origin The nutrient content and hence the ecology of the WFS in any given year depends on the proportions of these water sources. 1998 was anomalous with respect to WFS water properties. The cold tongue was colder and longer lasting, it had elevated nutrient levels, and the thermocline was more sharply defined, even in shallow water. We account for these anomalous conditions as a consequence of both anomalous local forcing and anomalous LC influence. Local forcing alone, particularly over the Florida Panhandle region was sufficient to account for anomalous upwelling there The LC influence added to this by

PAGE 224

191 elevating the material isopleths in the vicinity of the shelf-break, thereby allowing somewhat colder/higher nutrient concentration water to be upwelled in and east of DeSoto Canyon, but the essential features observed would have occurred independent of this LC influence. The LC influence was indirect in that it occurred remotely by a shelf-break impact on the southwest comer of the WFS where the isobaths converge (due to the Florida Keys), confirming the hypothesis advanced by Hetland et al. (1999). By contacting relatively shallow isobaths the LC set a persistent southward current in motion that facilitated the transport of waters upwelled in the north (particularly from the the DeSoto Canyon to the Big Bend region) to inner-shelf locations between Tampa Bay and Charlotte Harbor. Such cold water even reached the beach. The transport mechanism, being fully baroclinic, is more complex than the Hetland et al. (1999) barotropic theory. By a combination of local wind and heat flux forcing, irrespective of the LC, the thermocline has a convex shape that encompasses the subsurface cold tongue. The effects of the LC are 1) to increase the sharpness of the thermocline by causing colder water to be upwelled on to the shelf and 2) to cause the thermocline apex to sidle landward by setting shallower isobath waters in motion. The baroclinic structure of the circulation tends to mitigate (enhance) the LC influence on the currents landward (seaward) of the thermocline apex. This explains why comparisons between our modeled and observed currents are improved much more so at the 50 m isobath when the LC is added than at the 20 m isobath. The 1992 to present TIP satellite altimetry record shows that southwest comer impacts by the LC are rare, having occurred in fall/winter and again in summer 1997 and

PAGE 225

192 as described here for 1998. Longer-term measurements from the 20m isobath mooring of Figure 5.9 (not shown) also demonstrate the 1997 effects. LC impacts occurring farther north (e.g., He and Weisberg, 2002c) do not have as marked affect on the WFS because, by virtue of the Taylor-Proudman theorem, only the relatively deep isobaths are influenced. These more northern impacts do raise material isopleth heights, thereby facilitating upwelling beyond the shelf-break, but local forcing is the necessary in order for these deep-ocean water properties to broach the shelf-break and to subsequently be transported across the shelf. When this occurs (either by local forcing alone, or by local forcing augmented by LC impacts) the WFS receives deep-water nutrient injections that fuel inter-annual variations in shelf ecology Our findings bear upon the question of how to monitor a continental shelf. The bottom Ekman layer, being a conduit for across-shelf transport, requires measurements for circulation and water properties. Measurements are also required in the vicinity of and at, the shelf-break because deep-ocean properties must first broach the shelf-break if they are to be subsequently transported across the shelf. Measurements in deeper-water are of lesser value because they can not be used to imply transport onto the shelf. Measurements near a suspected upwelling center (like DeSoto Canyon) are also of limited value. For instance the 30 m isobath time series offshore of Panama City are much more valuable to assessing the effects of upwelling on the WFS than any collection of moorings in or seaward of the Canyon itself. We conclude that a limited array of moorings that emphasizes the shelf break, the mid shelf, and the near-shore regions (specifically bracketing the location of anticip a t e d thermocl i n e con v exity and also accounting for changes across near-shore salinity fronts by large estuarine discharges)

PAGE 226

193 would provide one portion of a multi-faceted monitoring program. Such program should also include radars for surface current and wave fields, profiling floats for interior density and water property fields, and surface meteorological stations for local forcing functions, along with all remotely sensed satellite information. In summary, we addressed the issue of local versus deep-ocean forcing of the WFS circulation through a coordinated program of in-situ measurements and numerical circulation modeling. Local forcing by winds and surface heat fluxes are the primary controlling factors. Deep-ocean forcing by the LC (or its eddies) are generally of secondary importance. 1998 was an anomalous year for both local and deep-ocean forcing factors, and we investigated how these conspired in combination to alter WFS water properties. A related paper by Walsh et al. (2002) continues these analyses through their impact on WFS ecology by comparing in-situ data with a coupled physical/biological model. ACKNOWLEDGMENTS Support was derived from several sources. Field work was initiated under a cooperative agreement between the USGS Center for Coastal Geology and USF. Measurements and modeling progressed with support from the Minerals Management Service and the Office of Naval Research. We are presently supported by the Office of Naval Research, Grant#N00014-98-1-0158; the National Oceanic and Atmospheric Administration, Grant# NA76RG0463; and by the State of Florida. Ruoying He is the recipient of an endowed research fellowship from the family of Elsie and William Knight. Rick Cole continues to provide the Ocean Circulation Group with excellent data returns

PAGE 227

194 from our moored program. A systems engineering overview and real time data capabilities has been added by Cliff Merz. Jeff Donovan is involved in all aspects of our work including editing, preparing, and managing our data sets, and managing our acquisition and computer systems. J Walsh provided many stimulating discussions G. Mitchum and B.G. Hong provided the TIP analyses. F. Muller-Karger provided satellite SST data W. Sturges (Florida State University) provided the Panama City moored ADCP velocity data collected under MMS cooperative agreement #14-35-0001-30787 M. Howard (Texas A&M University) provided the MMS sponsored NEGOM ship survey data. G. Kirkpatrick provided the Mote Marine Lab hydrographic data. NCEP reanalysis information were provided by NOAA-ClRES Climate Diagnostics C e nter, Boulder CO available from their web-site at http://www.cdc.noaa.gov.

PAGE 228

195 REFERENCES Baumgartner, M.F., S.P. Anderson (1999), Evaluation of NCEP regional numerical predication model: surface fields over the middle Atlantic Bight. J Geophys Res. 104 (C8): 18, 141-18,158 Blumberg, A.F., and G.L. Mellor (1987), A description of a three-dimensional coastal ocean circulation model, Three-Dimensional Coastal Ocean Models, Vol. 4, N. Heaps (ed.), 208-233, AGU, Washington, D. C., 1987. Brink, K.H. (1998a), Wind-driven currents over the continental shelf, in The Sea, 10, 320, K.H. Brink and A.R. Robinson eds., Wiley N.Y. Brink, K.H. ( 1998b ), Deep-sea forcing and exchange processes, in The Sea, 10, 21-62, K.H. Brink and A.R. Robinson eds., Wiley, N.Y. Boicourt, W C W.J Wiseman, Jr., A Vaile-Levinson, and L.P. Atkinson (1998), Continental shelf of the southeastern United States and the Gulf of Mexico in the shadow of the western boundary current, in The Sea, 11, 135-182, A.R. Robinson and K.H. Brink eds., Wiley, N Y Chu, P. Edmons, N. and Fan, C (1999). Dynamical mechanisms for the south china sea seasonal circulation and thermohaline varibilities J. Phys. Oceanogr., 29, 29712989, 1999 Ezer T. and G. Mellor, (1992); A numerical study of the vat*lbility and separation of the Gulf Stream, induced by surface atmospheric forcing and lateral boundary flow. J. Phys. Oceanogr. 22 660-682 Gill, A. E (1982). Atmosphere-Ocean Dynamics, p408. Academic Press. He, Rand R.H. Weisberg (2002a). West Florida shelf circulation and temperature budget for the 1999 spring transition. Cont. Shelf Res., 22,5, 719-748. He, Rand R.H. Weisberg (2002b) West Florida shelf circulation and temperature budget for the 1998 fall transition. Cont. Sh elf Res. (submitted) He, Rand R. H. Weisberg (200 2 c) A Loop Cu rre nt intrusion case study on th e West Florida Shelf, J. Phys. Oceanogr. (submitted)

PAGE 229

195 He, Rand R.H. Weisberg (2002d). Tides on the west Aorida shelf. J. Phys. Oceanogr. (submitted). Hetland, R. D., Y. Hsueh, R.R. Leben, and P. P. Niiler (1999), A Loop Current-induced jet along the edge of the West Florida shelf, Geophys. Res. Lett. 26, 2239-2242 Hill, A.E. ( 1998), Buoyant effects in coastal and shelf seas, in The Sea, 10, 21-62, K.H. Brink and A.R. Robinson eds., Wiley, N.Y. Kourafalou, V. H., L.-Y. Oey, J.D. Wang, and T.L. Lee (1996), The fate of river discharge on the continental shelf. 1: Modeling the river plume and the inner shelf coastal current. J.Geophys. Res., 101, 3415-3434, 1996 Li, Z. and R.H. Weisberg (1999a). West Aorida Shelf response to upwelling favorable wind forcing, Part 1: Kinematics. J. Geophys. Res., 104, 13,507-13,527. Li, z. and R. H. Weisberg ( 1999b ). West Aorida Shelf response to upwelling favorable wind forcing, Part 2: Dynamics. J. Geophys. Res., 104, 23427-23442. Meyers, S.D., E.M. Siegel, and R.H. Weisberg (2001). Observations of currents on the west Aorida shelfbreak. Geophys. Res. Lett., 28, 2037-2040. Orlanski, 1.(1976) A simple boundary condition for unbounded hyperbolic flows, J. Comput. Phys., 21, 251-269, 1976. Smagorinsky, J.(1963). General circulation experiments with primitive equations. I. The basic experiments. Mon. Weather Rev., 91, 99-164 Steidinger, K.A., Vargo, G.A., Tester, P.A. and Tomas, C.R., (1998). Bloom dynamics and physiology of Gymnodinium breve with emphasis on the Gulf of Mexico. In: Anderson, D.M., Cembella, A.D. and Hallegraeff, G.M., Editors, 1998. Physiological Ecology of Harmful Algal Blooms, Springer, Berlin, pp. 135-153. Sturges, W., and R. Leben, (2000), Frequency of ring separation from the Loop Current in the Gulf of Mexico: A revised estimate. J. Phys. Oceanogr., 20, 1814-1819. Walsh, J.J. K.D. Haddad, D.A. Dieterle, R.H. Weisberg, Z. Li, H. Yang, F.E. Muller Karger, C.A. Heil, and W.P. Bissett (2001). A numerical analysis of the landfall of 1979 red tide of Gymnodinium breve along the west coast of Aorida the fall/winter upwelling mode. Cont. Shelf Res., 22(1), 15-38 Walsh, J.J., R.H. Weisberg, D.A. Dieterle, R. He, B. P. Darrow, J. K. Jolliff, G. A. Vargo, G. Lirkpatrick, K. Fanning, T. T. Sutton, A. Jochens, D. C Biggs, B. Nababan, C. Hu and F. MullerKarger (2002), The phytoplankton response to intrusion of slope

PAGE 230

water on the west Florida shelf: models and observations. To be submitted Weisberg, R.H., B. Black, Z. Li (2000). An upwelling case study on Florida's west coast, J. Geophys. Res., 105, 11459-11469 Weisberg, R.H., Z. Li, and F.E. MullerKarger (2001) West Florida shelf response to local wind forcing: April1998. J. Geophys. Res., 106 (C12) 31,239-31,262 196 Yang, H. and R. H. Weisberg ( 1999). West Florida continental shelf circulation response to climatological wind forcing, J. Geophys. Res. 104, 5301-5320. Yang, H., R.H. Weisberg, P.P. Niiler, W. Sturges, and W. Johnson (1999). Forbidden zone over the West Florida Shelf, Cont Shelf. Res., 19, 1221-1245.

PAGE 231

CHAPTER6 WEST FLORIDA SHELF CIRCULATION AND HEAT BUDGET IN FALL TRANSITION OF 1998 6.1. ABSTRACT 197 Mid-latitude continental shelves undergo a fall transition as the net heat flux changes from warming to cooling Using in-situ data and a numerical model we investigate the circulation on the west Florida shelf (WFS) for the fall transition of 1998. The model is a regional adaptation of the primitive equation, Princeton Ocean Model forced by NCEP reanalysis wind, air pressure, and heat flux fields, plus river inflows. Based on agreements between the modeled and observed fields we use the model to draw inferences on the seasonal and synoptic scale features of the shelf circulation. By running twin experiments, one without and the other with an idealized Loop Current (LC) interaction, we explore the relative importance of local versus deep-ocean forcing in affecting shelf circulation. It is found that local forcing largely controls the inner-shelf circulation, which is very similar for both cases. The effects of the Loop Current in fall 1998 are to reinforce the mid shelf currents and to increase the across-shelf transports in the bottom Ekman layer, thereby accentuating the shoreward transport of cold, nutrient rich water of deep-ocean origin. A three-dimensional analysis of the temperature budget reveals that surface heat flux largely controls both the seasonal and synoptic scale

PAGE 232

temperature variations. Surface cooling leads to convective mixing that rapidly alters temperature gradients. One interesting consequence is that upwelling can result in near shore warming as warmer offshore waters are advected into cooler shallower regions. The temperature balances on the shelf are found to be both complex and fully three dimensional. 6.2. INTRODUCTION The circulation on the continental shelf controls the exchange of materials between the coastal and deep-ocean regions. The circulation is therefore a primary determinant of continental shelf ecology. Influenced by both local (surface winds, air pressure, surface buoyancy fluxes, and river inflows) and deep-ocean (boundary currents and eddies) forcing, the shelf circulation is dynamically linked to the varying water properties, particularly the temperature which exerts the primary control on density. Continental shelf temperature budgets are known to vary with location. Lentz (1987), Lentz and Chapman (1989), and Dever and Lentz (1994) studied the temperature variations of the northern California shelf and found the balance to be primarily two dimensional (controlled by across-shelf and surface heat fluxes), independent of season. With an array of instruments deployed to the north of Cape Hatteras from August through November 1994, Austin (1999) found distinctive seasonal variations in the temperature balance of the North Carolina shelf that are closely linked to water column stratification From May to August (under stratified conditions), across-shelf advection dominates the vertically integrated temperature budget. From October to March (under well mixed conditions), surface heat flux and along-shelf advection dominate the variations. He and

PAGE 233

Weisberg (2002a), hereafter referred to as HW02, describe the fully three dimensional temperature budget on west Florida shelf (WFS) during the spring transition of 1999. They found that the WFS thermodynamics are fully three-dimensional and that when vertically integrated the surface heat flux and ocean circulation dynamics control the seasonal and synoptic scale variability, respectively. 199 Opposite from the spring transition, which begins when the net surface heat flux changes from cooling to warming and the water column tends toward stratification, the fall transition begins when the net surface heat flux changes from warming to cooling and the water column tends toward de-stratification. How the evolution of the three dimensional temperature structure of the WFS differs between the spring and fall transitions is topic of the present paper. We follow the approach ofHW02 in which we combine in-situ data with an adaptation of the primitive equation, Princeton Ocean Model (POM) of Blumberg and Mellor (1987). The model is forced by National Center for Environment Prediction (NCEP) reanalysis wind, air pressure, and net surface heat flux fields, plus river inflows to study the seasonal circulation and temperature budget for the period September 151 to November 30th, 1998. By comparing twin model experiments, one independent of the Gulf of Mexico Loop Current (LC) and the other including an idealized version of the LC, we also investigate the relative importance of local versus deep-ocean forcing in determining the shelf circulation and temperature budget for the fall transition of 1998. The model and the forcing fields are described in section 2. Section 3 presents the twin experiments and compares the model results with the in-situ observations. Based upon these comparisons, the model is used in section 4 to describe the seasonal mean

PAGE 234

200 circulation on the WFS for fall 1998 and the corresponding temperature and salinity fields. Section 5 presents a term-by-term analysis of the three-dimensional temperature budget. The results are summarized and discussed in Section 6. 6.3 MODEL AND FORCING FIELDS 6.3.1. MODEL In parallel with HW02 the domain (Fig 6.1) extends from the Florida Keys to west of the Mississippi River delta and it has one open boundary that arcs between these two locations. An orthogonal curvilinear grid is used in the horizontal with resolution that varies from about 2 km near the coast to 6 km near the open boundary. A sigma coordinate is used in the vertical with 21layers non-uniformly distributed to better resolve the near-surface and near-bottom frictional boundary layers. Horizontal diffusivities are parameterized according to Smagorinsky (1963) w i th a dimensionless coefficient of 0.2. Vertical diffusivities follow the Mellor andY amada (1982) level 2.5 closure scheme. Bottom stress follows a quadratic law using a variable drag coefficient with a minimum value of 0.0025. A mode splitting technique is employed with external and internal time steps of 12 seconds and 360 seconds respectively. The model is initialized at rest with horizontally uniform stratification Above 200m, str a tification is based on t e mperature and s a linity observ e d in a Septemb e r 1998 trans-shelf [Ecology of Harmful Algal Blooms (ECOHAB: Florida) Program ] hydrographic survey. Below 200m stratification is based on climatology From this initial zero baroclinicity state the model s pins up r a pidly, generating b ar oclini ci ty in

PAGE 235

30 29 28 ...J 27 26 25 -90 1-NA2./ 2-EC4 3-ECS -88 ,;''' 201 -86 -84 -82 -80 Longitude Fig. 6.1. The regional model grid and bathymetry (upper panels) and observational station locations (lower panels). Sea level comparisons are with Florida tide gauges at Pensacola, Apalachicola, St. Petersburg, and Naples. Velocity comparisons are with acoustic Dop pler current profiles from instruments moored at 25, 20 and 10m isobath (1-3). Tempera ture budget is diagnosed at station A, B. Seasonal mean temperature budget is diagnosed along a transect off Sarasota, FL.

PAGE 236

202 balance with the wind, air pressure, and buoyancy forcing. Justification for this approach is given in HW02. Tidal forcing is excluded since, with tidal currents of only a few em s-1 the related tidal mixing is weak in comparison with other sources for mixing. A WFS tidal analysis along with discussions on tidal mixing efficacy are given in He and Weisberg (2002b). 6.3.2. ATMOSPHERIC FORCING The relatively coarse resolution (2.5X2 5) NCEP wind, air pressure, and surface heat flux fields are interpolated onto the model grid Whereas the coarse wind and pressure fields are representative of nature, the surface heat flux field does not match the spatial scales of sea surface temperature (SST) variability. We attempt to remedy this by applying a surface heat flux correction that relaxes the modeled SST toward a monthly mean SST analysis derived from satellite AVHRR images (e.g., Ezer and Mellor 1992; Chu et al, 1999). This procedure facilitates realistic baroclinic structures since the density perturbations originating at the surface are mixed vertically through time and spatially varying eddy viscosity (K111) and eddy diffusivity (Kh) coefficients produced by the model's turbulence closure scheme. 6.3.3. LATERAL BOUNDARY FORCING Examination of TOPEX/POSEIDON (TIP) time series indicates a possible impact by the LC on the southwest comer of the WFS in fall 1998. LC impacts on the WFS and their effects on the shelf circulation are the subj e cts o f recent studi e s by Hetland et al

PAGE 237

203 (1999); Meyers et al. (2000); He and Weisberg (2002c) ; Weisberg and He (2002). Hetland et al. (1999) hypothesize that LC impacts on the southwest comer of the WFS can set currents in motion on the WFS because it is there, by virtue of the Florida Keys, that the isobaths converge. Weisberg and He (2002) confirm this hypothesis using the technique of He and Weisberg (2002c) for including an idealized version of the LC in our regional model domain. The latter paper also shows that LC impacts occurring farther north along the shelf break do not set currents in motion on the shelf, consistent with the findings of Meyers et al. (2000). Here we apply the same technique in a twin experiment for the purpose of quantifying the relative importance of local forcing versus the LC during the fall 1998 seasonal transition. A LC impact is mimicked by imposing an approximately Gaussian shaped sea surface height perturbation along the open boundary west of the Florida Keys. Geostrophic inflow on one side is balanced by geostrophic outflow on the other side and the ensuing current freely evolves within the model domain. The mode splitting technique by which all of the vertically integrated transport is contained in the barotropic mode ensures a rapid baroclinic adjustment so that the shelf-break is impacted by a strong baroclinic current. We use TIP data to estimate the location and magnitude of the pressure perturbation and for simplicity (and lack of sufficient data) the pressure perturbation is held constant over the three-month simulation. Other than the imposed pressure perturbation the open boundary is treated with an Orlanski (1976) radiation boundary condition. Seven major rivers are included for land-drainage-derived buoyancy forcing These are the Mississippi, Mobile, Apalachicola, Suwannee, Hillsborough, Peace, and Shark rivers. Interpolated monthly mean fresh water mass fluxes are input to the top

PAGE 238

sigma level at the grid cells closest to the rivers' locations using the technique of Kourafalou et al. (1996) (also see Pullen, 2000). 6.4. MODEL AND DATA COMPARISION 204 Twin model experiments, one with local forcing only (Case I) and the other with both local forcing and the idealized Loop Current (Case II) are performed. This section compares the Case I and Case II results against with in-situ sea level and velocity data. 6.4.1. SEA LEVEL Since the model is forced without tides, all of the model and data comparisons are shown after low-pass filtering to exclude the tidal and inertial period oscillations. Case I sea level comparisons are given in Fig. 6.2 at four different NOAA tide-gauge stations distributed along the coast from Pensacola in the northwest to Naples in the southeast. Three hurricanes transited the model domain in fall 1998: Earl (August 31 to September 3), Georges (September 16 to 29) and Mitch (October 22 to November 5) The storm surge simulations for these and other synoptic scale events are reasonably good with squared correlation values of around 0.8 at all stations Case II comparisons (not shown) are similar to Case I, essentially in agreement with Marmorino (1982) who found that the coastal sea level response to winds is insensitive to the LC configuration near the shelf break. We conclude that coastal sea level variability in fall 1998 is primarily in response to local, shelf-wide forcing. 6.4.2. CURRENTS

PAGE 239

205 1 6 11 16 21 26 1 6 11 16 21 26 31 5 10 15 20 25 30 Sep Oct Nov 1998 Fig. 6.2. Comparisons between modeled (bold lines) and observed (thin lines) sea level at Pensacola, Apalachicola, St. Petersburg and Naples as quantified by squared correlation long with NCEP wind velocity sampl e d at station A. Modeled results are from Case I.

PAGE 240

206 Comparisons are made between modeled and observed velocity vector time series at the 25m, 20m, and 10m isobaths (moorings NA2, EC4, and EC5, respectively, as shown in Fig. 6.1 ). The Case I, comparisons are given in Figs 6.3-6.5. The observations are from moored Acoustic Doppler Current Profilers (ADCP), and comparisons are provided at three different depths: near-surface, mid-water column, and near-bottom. We quantify these comparisons using a complex correlation analysis (i.e., Kundu, 1976; HW02), the results of which are given as two sets of numbers on each plot. The left hand sets are the seasonal mean east and north velocity components for each vector time series. The right-hand sets are the squared correlation coefficients, the relative orientation angles (measured anti-clockwise), and the regression coefficients between the modeled and the observed velocity vectors. Like sea level, the modeled and observed currents also compare reasonable well. At all three stations and depths, the squared correlation coefficients range between about 0.59 and 0.78, and the orientations agree to between about -3 to +12. Disparities are seen in the time series on an event by-event basis. First, the model fails to pick up some of the reversals in the observed currents. Second, the regression coefficients show that the model tends to underestimate the amplitude of observed velocity fluctuation by about 20 to 50 percent. Such deficiencies may result of the low resolution of NCEP reanalysis fields that exclude smaller-scale coastal boundary layer structures in the winds. Coarse resolution problems are exacerbated further through smoothing to interpolate the reanalysis fields onto the model grid. Nevertheless, the model captures the general features of the observations and gets the sense of the velocity rotation correct in both surface and bottom Ekman layers. Improved forcing functions will lead to improved quantitative metrics.

PAGE 241

207 I o 6 11 16 21 26 6 11 16 21 26 31 5 10 15 20 25 30 Sap Oct Nov 1998 Fig. 6.3. Comparisons between modeled and observed currents at the 25m isobath (mooring NA2) sampled at surface, mid-depth and near the bottom, along with the NCEP wind sampled at Station A. Each vector current time series is accompanied by its seasonal mean east and north velocity components (left-hand couplet) and each model/data com parison is quantified by its squared complex correlation coefficient, phase angle (or angular deviation of the model vectors from the data vector measured c ounterclockwise), and regression coefficient (right-hand triplet). Modeled results are from Case I.

PAGE 242

208 I o Data Bottom ) 0 \ ,,y,.,\\\, .\.. j.\\ /d, .2. ,, ( >77(\ '\ "' '', "'t' "''" '\\"' .. 6 11 16 21 26 6 11 16 21 26 31 5 10 15 20 25 30 Sep Oct Nov 1998 Fig. 6.4. Comparisons between modeled and observed currents at the 20m isobath (moor ing EC4) sampled at surface, mid-depth and near bottom along with the NCEP wind velocity sampled at Station A. Quantitative comparisons are as in Fig. 6.3. Modeled results are from Case

PAGE 243

209 I o 'I ... '1\\W" -2o 6 11 16 21 26 6 11 16 21 26 31 5 10 15 20 25 30 Sap Oct Nov 1998 Fig. 6.5. Comparisons between modeled and observed currents at the 10m isobath (moor ing EC5) sampled at surface, mid-depth and near bottom along with the NCEP wind velocity sampled at Station A. Quantitativ e comparisons are as in Fig. 6.3. Modeled results are from Case

PAGE 244

210 As with sea level, the Case II comparisons (not shown) are similar to the Case I comparisons. Adding LC provides only subtle differences in the modeled currents of the inner-shelf. As shown for the 25 m isobath location in Fig. 6.6 the modeled southeastward currents in Case II are a little stronger than in Case I, but the synoptic scale variability between the two cases are nearly the same. These results are summarized in Figure 6.7 wherein seasonal mean velocity vectors and seasonal mean velocity hodograph ellipses are provided at mid-depth for the observations overlaid on the model results for Case I (upper panel) and Case II (lower panel). We can conclude that for fall 1998 the inner-shelf currents were primarily in response to local forcing Velocity measurements at deeper isobaths are unavailable during fall 1998. Based on these inner-shelf results and the spring/summer period findings of Weisberg and He (2002) when deeper measurements are available, we now use the model to draw inferences on the midto outer-shelf circulation. 6.5. MODELED SEASONAL MEAN CIRCULATION 6.5.1. CIRCULATION FIELDS The seasonal mean circulation fields, obtained by averaging the model flow vectors from September 1st to November 30th, are presented in Figs. 6.8 and 6.9 for Cases I and II, respectively. Each figure shows the depth-average, near-surface (sigma layer 2), mid-depth (sigma layer 10), and near-bottom (sigma layer 20) currents. Three dimensionality arises by geometrical factors such as blocking by the Aorida Keys, coastline changes from the Big Bend to the Panhandle, near-shore penetration of deep isobaths at DeSoto Canyon, and by the effects of baroclinicity and surface and bottom

PAGE 245

6 11 16 21 26 6 11 16 21 26 31 5 10 15 20 25 30 Sep Oct Nov 1998 Fig 6.6. Comparisons between modeled currents of Case I and Case II at the 25 m isobath (mooring NA2) sampled at surface, mid-depth and near bottom along with the NCEP wind velocity sampled at Station A. 211

PAGE 246

tg 27.5 e.g, '-. ' 27 -83. 5 -83 -82. 5 -82 Fig. 6.7. Comparisons between modeled (bold) and observed (thin) seasonal mean veloc ity vectors and hodograph ellipses (variance principal component axes) at mid-depth for all three mooring locations on the WFS between 25 to 10m isobath Upper (lower) panel is for Case I (Case II). 212

PAGE 247

Depth Averaged Current Near Surface Current Mid-depth Current Near Bottom Current Fig. 6.8. Modeled (Case I) seasonal mean velocity vectors for the depth averaged and near surface, mid-depth, and near bottom sigma levels, k=2, 10, 20, respectively. 213

PAGE 248

Near Surface Current Mid-depth Current Near Bottom Current ' Fig. 6.9. Modeled (Case II) seasonal mean velocity vectors and their variance ellipses for near surface, mid-depth, and near bottom sigma levels, k=2, 10, 20, respectively. 214

PAGE 249

215 Ekman layers. The mid-depth and depth-average fields are similar (and with the same scale), and they show the general nature of the mean currents exclusive of Ekman layer effects. The near-surface (near-bottom) currents are larger (smaller) and these differences in magnitude are accommodated for illustration purposes by the scale changes noted on the figures Commonality exists for some of features in both cases. Near-shore the depth average and mid-depth currents show a divergence about the Big Bend region in which the currents flow southward between Tampa Bay and the Florida Keys while flowing westward west of Cape San Bias. At the shelf break these currents tend to be southward. The near-surface and near-bottom currents are different. Near-surface we see a northward jet at the shelf-break that is continuous with the westward flows along the Panhandle and the Florida Keys. Near-shore off Tampa Bay (Florida's west coast) we see an offshore directed flow All of these surface features are attributable to a general easterly wind pattern on average (Fig 6.10). Near bottom off Florida's west coast we see a shoreward directed across-shelf flow pattern in the bottom Ekman layer which is consistent with the southward flow at mid-depth While modulated in amplitude and phase this upwelling favorable circulation offshore of Florida's w e st coast appears to be a general feature of the fall season that shows up each year in our extended observational records This upwelling favorable circulation has water approaching the coast at depth from the north, upwelling near shore, and then flowing southward and away from the coast at the surface. An upwelling case study for a specific synoptic scale event is described by Weisberg et al. ( 2 000) and these simple E k man-geostrophic adjustment conc e pts coupl e d with the WFS geom e try account for the a bsence of su rfa ce dri f t e rs

PAGE 250

September Mean ,,,,,, '\\\\\, \\\\\'\ November Mean ...-' ;' "' y y y ..,.., #" / / """ October Mean e--e---e---/ /. Seasonal Mean --.r ---y ,;tM "' ,;tM ..Fig. 6.10. Monthly(September, October and November) and seasonal mean surface wind fields in fall, 1998. 216

PAGE 251

217 near-shore south of Tampa Bay as described by Yang et al. (1999) In contrast with Florida's west coast, the bifurcation of the depth-average currents in the Big Bend results in a downwelling favorable regime along the Florida Panhandle coast. We see that coastline and isobath geometry relative to the seasonal wind patterns largely dictates the structure of the regional flow fields. Along with commonality there are also distinct differences between the two cases (with and without LC influence) particularly over the midto outer-shelf regions. With the LC influence (Fig. 6.9), we see a broader and stronger southward mid-shelf current along with a broader and stronger across-shelf flow pattern in the bottom Ekman layer everywhere south of Cape San Bias. This is attributed to the constructive interference that occurs between the LC and local forcing effects over this portion of the model domain. In contrast we see destructive interference along the Panhandle coast west of Cape San Bias. The result is a diminution of the downwelling favorable circulation there. While remote, we find that the impact of the LC (on relatively shallow isobaths at the shelf-break west of the Florida Keys) is propagated throughout the entire model domain to the Mississippi River delta. This occurs quickly through the anticlockwise propagation of continental shelf waves (e.g., Gill, 1982, p408), which with celerity of order m s -1 can transit the model domain in a matter of days. Weisberg et al. (1996) hypothesized that the evolution of the seasonal mean fields involves both wind and surface heat flux forcing. HW02 demonstrat ed this for the spring 1999 transition, and similar factors come to play in the fall transition. The difference being the spring and fall transitions is that the former entails a change in the net surface h eat flux from cooling to warming the ocean while the latter change is from warming to

PAGE 252

218 cooling. Figs. 6.11 and 6.12 provide the September, October, November, and seasonal mean depth-average velocity fields for Cases I and ll, respectively. Beginning in September and consistent with the mean wind field (Fig. 6.10) we see a northward directed current with largest magnitude near the coast. As the winds switch from southeasterly to northeasterly in October the currents change markedly, in part due to heat flux. Summer months are when the near-shore temperatures are highest and the baroclinic geostrophic circulation tends to be northward and downwelling favorable. Hence the September currents show largest magnitude near-shore where the wind and buoyancy-induced circulations add constructively. By October the near-shore waters are cooled by the net surface heat flux, which reverses the sign of the thermal wind and causes a constructive interference with the now slightly upwelling favorable winds. By November the onshore-directed density gradient is positioned closer to the shelf-break so the near-shore currents are primarily wind-driven, whereas the shelf-break currents are largely buoyancy driven. Hence the region of maximum current magnitude sidles seaward over the fall season march with the region of maximum density gradient. Strongest currents directed to the north (south) occur prior to the change in net surface heat flux when density gradient points seaward (landward). Adding the LC results in constructive/destructive interference as previously described relative to Figs. 6.8 and 6.9. Flows directed from the Mississippi River region toward the Florida Keys are increased, whereas flows in the opposite direction are decreased. As a result we see what appears to be a shoreward translation of the region of southward directed currents from the Big Bend region to points located south of Tampa Bay. Under LC influence the entire shelf from the Big Bend to the south shows

PAGE 253

219 September Mean October Mean November Mean Seasonal Mean Fig. 6.11. Evolution of the monthly mean depth averaged velocity vectors for September, October and November along with the fall 1998 seasonal mean for Case I.

PAGE 254

220 September Mean October Mean November Mean Seasonal Mean Fig. 6.12. Evolution of the monthly mean, depth averaged velocity vectors for September, October and November along with the fall1998 seasonal mean for Case II.

PAGE 255

southward-directed flow, and this has a profound impact on across-shelf transport as shown next. 6.5.2. LAGRANGIAN TRAJECTORIES 221 The bottom Ekman layer is a major conveyance for the transport of materials across the broad, gently sloping WFS (Weisberg et al., 2001 and Weisberg and He, 2002). Such across-shelf bottom Ekman layer transport, further amplified by LC impacts as in fall 1998, accounts for the cold, nutrient rich water of deep-ocean origin reported at that time by Walsh et al. (2002). To illustrate this finding, we show Lagrangian particle trajectories calculated for neutral buoyant particles advected by modeled three dimensional flow fields for both Case I and Case II. These are given in Fig. 6.13 for particles released on September 151 in the near-bottom sigma layer 18 along the 25m, 50 m, and 100m isobaths between Panama City and the Rorida Keys. The left and right hand panels are for Cases I and II, respectively. Particles released at the 25m iosbath all find their way to the near-shore zone in either case. From the 50 m isobath we see some additional asistance by the LC impact-induced flows, especially from positions originating within the Big Bend. Constructive interference between locally forced and LC-induced along-shelf flows within the water column increases the bottom Ekman layer responses and hence the across-shelf transports. This effects is most evident in the 100 m isobath panels where across-shelf transports are only evident under the influence of the LC. Thus the delivery of cold, nutrient rich waters from the shelf-break to the near-shore in fall 1998 was a consequence of both the LC and local wind and buoyancy forcing. The LC-induced flows and their bottom Ekman layer responses were necessary to transport

PAGE 256

30 29 28 27 26 25 -90 -88 -86 -82 30 29 28 27 26 25 -90 30 29 28 27 26 25 -90 -88 -86 -84 -82 -88 -86 84 -82 30 29 28 27 26 25 -90 -88 30 29 28 27 26 25 -86 -84 -82 -90 -88 30 29 28 27 26 25 -90 -88 -86 -84 -82 -86 -84 82 222 Fig. 6.13. Modeled Lagrangian trajectories tracked over three month of the drifters released along 25m, 50m, and 1OOm isobath near the bottom. Left and right panels are for Case I and Case II respectively.

PAGE 257

223 properties from the shelf break to the mid-shelf. The locally driven flows and their bottom Ekman layer responses (further amplified by the LC effect) then took over to transport properties to the near-shore. Moreover the trajectory paths tend to intercept the near-shore region between Tampa Bay and Charlotte Harbor consistent with the local upwelling maximum argument advanced by Weisberg et al. (2000) 6.6. SCALAR FIELDS AND TEMPERATURE BUDGET 6.6.1. TEMPEARUTE AND SALINITY FIELDS The combination of local plus LC forcing also affects the scalar fields of temperature and salinity Figs 6.14 and 6.15 show the modeled sea surface temperature (SST) and sea surface salinity (SSS) fields at the end of the model run on November 30th for Cases I and II, respectively. Since the model is initialized in either case with spatially uniform SST and SSS, these figures show the relative effects of local and deep water momentum and buoyancy fluxes in changing SST and SSS. Both sets of figures are similar in that SST is cooler and SSS is fresher near-shore. The SST distributions are largely accounted for by the net surface heat flux transition from warming to cooling in fall. Surface cooling causes convective mixing which, when combined with shoaling topography, produces a seaward directed SST gradient. Ocean circulation dynamics, through upwelling is also a factor however. This accounts for the Case II SST being somewhat cooler than the Case I SST since relatively deeper, colder water is advected shoreward under Case II. SSS shows less inter-case difference than SST This is because low salinity derives from coastal river plumes ad v ec ted by the n ear -shor e flow fie lds that are very

PAGE 258

28 27 26 Temperature Nov.30 25 28 27 26 Salinity Nov. 30 25 -90 -88 -86 -84 -82 -80 Fig. 6.14. Modeled (Case I) sea surface temperature and sea surface salinity fields at the end of fall 1998 model simulation (November 30, 1998) 224

PAGE 259

28 27 26 Temperature Nov.30 25 29 28 27 26 Salinity Nov 30 25 -90 -88 -86 -84 -82 -80 Fig. 6.15. Modeled (Case II) sea surface temperature and sea surface salinity fields at the end of fall 1998 model simulation (November 30, 1998) 225

PAGE 260

226 similar in both cases. However, the fall SSS fields differ markedly from their spring transition (HW02) counterparts. This is due to the change that occurs along-shelf between the upwelling (downwelling) circulation found to the east (west) of Cape San Blas on average in fall 1998. Thus, waters of Mississippi River, Mobile, and Florida Panhandle origin are limited to the western part of the study region in fall, whereas they are advected southeastward in spring. As a consequence we can now explain why the seasonal "green river" phenomenon (Gilbes et al., 1997) occurs in spring and not in fall. 6.6.2. THE TEMPERATURE EQUATION The three-dimensional temperature budget for the fall 1998 transition is diagnosed in parallel to the HW02 spring 1999 analyses. The temperature equation is recast from its modeled flux divergence, a-coordinate form to an advective, z-coordinate form. Thus, we diagnose ()T oT ()T ()T () ()T () oT () oT -= -u--v--w-+-(AH -)+-(AH -)+-(KH -) at ax ()y az ax ax ()y oy az az (6.1) which equates the local rate of change of temperature (oT jot) to a combination of the flow field advective rate of change ( -u oT fox-voT joy-woT joz), and the rates of change by horizontal [ ojax(AH aT fax)+ ojoy(AH oTjoy)] and vertical [ ajaz (K H ()T jaz)] diffusion. The temperature budget is diagnosed with respect to time series of both vertical averages and depth profiles at two different isobaths, and within an across-shelf section for the seasonal means. The two time series analysis locations, A and B (see Fig. 6.1) are on the 50m isobath in the Big Bend and on the 15m isobath offshore of Sarasota, respectively.

PAGE 261

227 Through a term-by-term analysis of equation (6.1) we quantify the contributions by each physical process in changing temperature. For brevity we only consider Case IT since comparisons between the modeled and observed temperature sections suggest that Case II is closer to the reality than Case I in fall 1998. The differences between the temperature balances between Cases I and II will be discussed in section 5.5. 6.6.3. DEPTH-AVERAGE BALANCE Vertically integrating equation (6.1) provides a depth-average temperature equation. The depth-average diffusion is essentially the depth-average vertical diffusion [ Qj(pCpH), where Q is the net surface heat flux, p and CP are the seawater density and specific heat, and His the water depth] since horizontal diffusion is generally an order of magnitude smaller than the vertical diffusion. Vertically integrated water column temperature variations thus depend on the heat flux divergence by the ocean circulation and the surface heat flux. Time series of these two depth-integrated terms (advective and surface heat fluxes) and their sum are shown in Fig. 6.16 where panels labeled A and B correspond to points A and B, respectively. While not shown, we checked our budget analysis by noting that the sum is nearly identical to the depth-average local rate of temperature change. Means and standard deviations are provided for each term to account for their contributions to the seasonal and synoptic scale variations, respectively. The magnitudes of the advective and diffusive contributions to the depth-average rates of change are similar at the mid-shelf point A location. The seasonal change (given by the three-month mean values), being negative (cooling), is primarily controlled by

PAGE 262

(a) 0.5 0 Advection: -udT /dx-vdT /dy-wdT /dz Mean :-5.97e-03 Std :9.23e-02 J\n (\ /\__ ------== vv <;:;> --0. 5 0.5 : : Diffusion:HFx+HFy+VF Mean :-3.88e-02 Std : 5 78e-02 0 v C7 -0. 5 0 5 : : Advection+ Diffusion Mean :-4.48e-02 Std : 1 .20e-01 'g 0 !:2. -0. 5 !\r'\1\ 1\ !\ 1\ v 'JL....v \I '-" 'V 31 5 10 15 20 25 30 5 10 15 20 25 30 4 9 14 19 24 29 Aug Sep Oct Nov 1998 (b) Advection:-udT/dx-vdT/dy-wdT/dz Mean :-2.22e-04 Std :4.07e-02 0 /">.. ......_ Diffusion:HFx+HFy+VF Mean :-7.40e-02 Std : 8 73e-02 0 31 5 10 15 20 25 30 5 10 1 5 20 25 30 4 9 14 19 24 29 Aug S e p Oct Nov 1998 Fig. 6.16. (a) The relative contributions to the depth-averaged temperature balance by ocean circulation and diffusion at station A. Three time series are shown: the advection, the diffusion and their sum (which is nearly exactly equals the local change of depth average temperature). Accompanying each time series are their seasonal means and standard deviations in units of oc day-1 as measures of the seasonal and synoptic scale variability. (b) Same as (a), but for station B. 228

PAGE 263

229 surface heat flux, whereas the synoptic variability (given by the standard deviations) relies on both the surface heat flux and the ocean advection. Interestingly, the ocean advection and surface heat flux terms tend to counteract one another mitigating temperature change over the first half of the season, whereas they add constructively over the second half. On seasonal average, advection enhances the total water column cooling (-4.16) by about 15%. For the synoptic scale variability advection is slightly more effective than the surface heat flux in changing the water column temperature. The impacts of three major storm events are also noteworthy, i.e., Hurricanes Earl (August 31 to September 3), Georges (September 16 to September 29), and Mitch (October 22 to November 5). The relative importance between ocean advection and surface heat flux changes approaching shallower water. Closer to the coast at point B the seasonal water column cooling ( -6.90) is almost entirely by surface heat flux, and surface heat flux also plays a larger role than ocean advection at synoptic scales. Thus, the shallower the water the larger the fall transition temperature changes and the proportionately larger the role of surface heat flux over ocean advection. Advection does serve to mitigate change in an interesting way. Once a seaward temperature gradient is established upwelling favorable wind events can lead to counter-intuitive warming as warmer offshore-located waters are advected and upwelled into the colder near-shore region. This three-dimensional effect of advection is expanded upon in the following vertical profile analysis. 6 6.4 VERTICAL PROFILES OF THE TERM-BY -TERM TEMPERATURE BALANCE

PAGE 264

230 The temperature budget is further explored in Figures 6.17 A and 6.17B by a term by-term time series analysis of the vertical structure of the temperature variability at points A and B, respectively. In each of these figures the left-hand panels show the horizontal and vertical components of the ocean advection and their sum, and the right hand panels show the diffusion (primarily vertical), the diffusion plus the advection (nearly identical to the local rate of change of temperature), and the temperature itself. Shaded and clear contours are used to denote warming and cooling tendencies, respectively. Destratification due to convective mixing in response to net surface cooling distinguishes the fall transition from the spring transition when net surface heating leads to stratification (HW02). Convective mixing renders the vertical temperature gradient ( oT joz) smaller in fall than in spring so the contribution to the temperature budget by vertical advection ( woT joz) in fall is less than in spring, especially in shallow water. Nevertheless, and depending on location, the role of ocean advection on both the synoptic and seasonal scales may be comparable to that of diffusion and oftentimes it is of opposite sign As with spring, the thermodynamics are fully three-dimensional, and the omission of any of the coordinate directions would compromise a model's ability to describe the temperature evolution. The balances at the deeper and shallower isobath stations show interesting evolution differences. At the 50 m isobath (Figure 6 .17 A) we see that at synoptic scales all of the terms are of comparable magnitude. Initially, while the water column is stratified, advection tends toward warming at all depths while diffusion tends toward cooling near the surface. Upon addition, we see that the thermocline erodes due to the simultaneous coo lin g near-surface and warming at depth. Once the hori z ontal

PAGE 265

-(b) .... "' .... Fig. 6.17. (a) Time series of the depth profiles of the individual terms that compose the temperature balance at Station A. The left-hand panels show the horizontal and vertical components of the ocean advection and their sum, and the right-hand panels show the diffusion, the diffusion plus the advection, and the temperature. To the right of each panel is the seasonal mean profile. The contour interval for each o f the budg e t term is 0 1 oc day-1 and th e contour interval for temperature is 1 C. Shadin g indic a t e s w arming and clear indicants cooling. (b) Same as (a), but for station B. 231

PAGE 266

232 temperature gradient reverses to become directed seaward ocean advection by itself tends toward cooling near-surface and warming at depth. This is evident at the shallow water location B in Figure 6.17B. The fall 1998 season predominantly upwelling favorable circulation then advects cooler water seaward near the surface and warmer water shoreward at depth. The opposite tendency is mirrored in the diffusion term as convective mixing is driven from both above and below. Such mixing tends to warm near the surface (despite negative surface heat flux) and cool at depth. The net result is a nearly vertically uniform temperature that decreases monotonically over the fall transition and into winter. The counter-intuitive result of the ocean advection not only provides for more effective mixing (from below in addition to from above) ; it also leads to a positive feedback that increases the net surface heat flux from the ocean to the atmosphere. It does this by replenishing the near surface with relatively warmer water from below than would exist in the absence of ocean advection. Ocean advection therefore mitigat e s and lengthens the fall season transition. 6.6.5. ACROSS SHELF TRANSECTS FOR THE SEASONAL MEAN TEMPERATURE BUDGET Here we draw comparisons between the fall 1998 Case I and C a se II seasonal mean temperature budg ets and th e t e mper a ture bud ge t for sprin g 1999 (HW02) with respect to an across-shelf section sampled offshore of Sarasota, FL. (Figure 6 1). The purposes are to illustrate the relative effects of local versus LC forcing of the WFS in fall 1998 a nd th e f undam e ntal diff e r e n ce s b e tw ee n th e f all a nd sprin g tran s itions. Fig ur e 6 1 8

PAGE 267

233 -udT/dx-vdT/dywd T/dz HFX+Hfy+ VF udT/dx-vdT/dy-wdT /dz+H f x+ HFy+VF 0 0 .. ... -20 -20 -20 [ -40 -40 ..... -"'"--40 "' 15. -2: -60 -60 -60 / -80 CI=0 .01C/ d C I =0. 0 1C/d CI=0 01C/d -80 -80 Fall98 Case I Fall98 Case I Fall98 Case I -100 100 -100 udT/dx-vdT/dy-wd T/ dz H FX+Hfy+VF -udT/dx-vdT / dy-wdT /dZ+Hfx+Hfy+VF 0 -20 -20 -20 ,.r .. -. I -40 -40 -40 5 -60 0 -60 -60 -80 C I =0.0 1C/d CI=0.01C/d CI=0.01C/d Fall98 Case II -80 Fall98 Case II -80 Fall98 Case II -100 -100 -100 udT/dx-vdT /dywdT /dz Hfx+Hfy+VF udT/dx-vdT /dy-wdT/dZ+Hfx+H fy+VF 0 0 20 -20 -20 I -40 -40 -40 5 0. 2: -60 -60 -60 -80 CI=0.01C/d -80 CI=0 01C/d CI=0 .01C/d Spr99 Spr99 -80 Spr99 100 100 100 200 150 100 50 -200 -150 100 50 -200 150 100 -50 ONshore D istance lkml ONshore Distance lkml OHshore Dlalancelkml Fig 6 .18 Across-shelf transect (off Sarasota, FL) profiles of seasonal means of ocean advection, diffusion terms an d their sums for Fall 98 Case I (upper panels), Fall98 Case II (middle panels) and Spring 99 (lower panels) The contour interval is indicated in each panel. The shading and clear contour denote warming and cooling effects, respectively

PAGE 268

234 is organized into three panels labeled A, B, and C for fall 1998 Case I, fall 1998 Case II, and spring 1999, respectively. From left to right, each panel contains three seasonal mean elements: 1) ocean advection, 2) diffusion, and 3) their sum (which mirrors the seasonal mean local rate of temperature change). The seasonal means of each of these terms are obtained by averaging the model-diagnosed temperature budget fields from September 151 to November 30th for fall 1998 and from March 151 to May 3151 for spring 1999. Consider first the relative importance in the fall 1998 season temperature budget between local forcing only (Case I) and local forcing plus the remote LC influence (Case II) Within the near-shore region (isobaths less than about 20 m) the two are nearly identical. Upwelling on seasonal average provides tendencies for warming (cooling) near the bottom (surface) for the non-intuitive reasons explained earlier. Diffusion on seasonal average provides a cooling along this transect which exceeds the magnitude of the advective warming. The net result is near-shore cooling over the entire water column. The effects of the LC are seen most clearly at mid-shelf and seaward, especially over the lower portion of the water column. There, enhanced across-shelf transport and upwelling of cold, deep water (e.g., Figure 6.9) leads to an enhanced cooling effect over the entire shelf. In either of the fall cases, despite counteracting advective and diffusive influences over some portions of the domain, cooling wins out over the upper ocean (depths less than 40m in Case I and at all depths shown in Cas e II) The spring 1999 seasonal average differs from either of the fall 1998 cases in that advection and diffusion tend to add constructively almost everywher e. Also the season a l mean tempe r a ture advection tend e ncy in spring 1999 i s smaller in m ag nitud e th a n the

PAGE 269

235 advection tendencies under either of the fall 1998 cases. We attribute these fundamental differences between fall and spring transitions to the state changes from destratified to stratified conditions in spring versus stratified to destratified conditions in fall, coupled with the predominantly upwelling favorable conditions in fall 1998 leading to an increased circulation dynamics influence on that seasonal transition. In other words, both the temperature gradients and the currents are persistently larger in fall 1998 than in spring 1999 leading to a larger advective influence. Another fundamental difference entails the counter-intuitive finding of upwelling leading to near-shore warming in fall This is unlikely in spring because once the temperature gradient is directed landward upwelling cools the near-shore region. Consistent with the vertically integrated temperature analyses off the North Carolina coast of Austin (1999), the WFS seasonal transitions also entail a combination of surface heat flux and ocean advection. Here, however, the thermodynamics are fully three-dimensional, spatially inhomogeneous, and dependent on both the local and deep-ocean forcing influences that determine the temperature gradients and the circulation. 6.7. SUMMARY AND CONCLUSION Mid-latitude continental shelves undergo a fall transition as the net surface heat flux changes from warming to cooling. Using in-situ data and a numerical circulation model, we investigate the circulation and temperature budget on the WFS, including the northeastern Gulf of Mexico shelf from the Mississippi River to the Florida Keys, for the fall transition of 1998. The data consist of sea level from coasta l stations, velocity profiles from instruments moored across the shelf between the 25 m and the 10 m

PAGE 270

236 isobaths. The model is a regional adaptation of the primitive equation, POM forced by NCEP reanalysis wind stress, air pressure, and heat flux fields and by river inflows. By running twin experiments, one by local forcing only (Case I) and the other by both local forcing and an idealized Loop Current effect (Case m, we investigate the relative importance of the local and deep-ocean forcing in affecting the shelf circulation. Based on the fact that both cases provide similarly good agreements with the observed data (on sea level and currents) we conclude that the inner-shelf circulation of the WFS primarily occurs in response to local forcing. Remote forcing by the LC as occurred in fall 1998 does have important effects nevertheless. Distinctive differences between two cases are found in the model over the middle to outer portions of the shelf. By intensifying the mid-shelf currents the LC influence leads to a stronger on shore transport within the bottom Ekman layer. Lagrangian particle tracking over the three month fall season with and without LC influence shows that the LC is instrumental in causing cold, nutrient rich water of deep ocean origin to be translated between the shelf-break and the inner-shelf Such increased landward bottom Ekman layer transport appears to be responsible for elevated primary productivity and anomalous coastal ecology interactions that occurr e d in fall 1998 (Walsh et al., 2002). While significant, such remote LC interactions that facilitate large scale elevations of nutrients levels on the shelf appear to be rare Of the 10 years of TOPEX!Poseidon sat e llite altim e try records examin e d by W e isber g a nd He (2002) such events only occurred in two of these years (1997 and 1998). Inter annual variability in WFS ecology due to remote LC events may be a very important aspect of what i s otherwise de sc ribed as an oli g otrophic s yst e m.

PAGE 271

237 Measurements on the WFS suggest a seasonal cycle in which the circulation tends toward upwelling in winter and downwelling in summer. Between these two phases are the spring and fall transitions due to the sign changes in the net surface heat flux. Weisberg et al. (1996) hypothesized that the circulations observed during the seasonal transitions are largely due to the baroclinicity imposed by surface heat flux. The present (fall transition) study together with HW02 (spring transition) confirm this hypothesis. Due to the combined effects of surface heat flux and shoaling topography, the across shelf temperature gradient car ;ax) changes from seaward (landward) to landward (seaward) during spring (fall) transition. The associated baroclinic circulation when fully developed, flows northwestward (southeastward) in spring (fall). Such baroclinic circulation, adding either constructively or destructively to the wind-driven circulation, provides both season and location-dependent along and across-shelf current distributions. Through a fully three-dimensional and term-by-term analysis of the temperature budget, we describe the evolution of the WFS temperature in the fall season complementing the HW02 spring analyses. Consistent with Austin ( 1999), the WFS temperature balance also has distinctive seasonal variations. Through a depth-average temperature analysis, we found that surface heat flux tends to dominate the temperature variations for both the seasonal and synoptic scales in fall. This differs from spring when the temperature variations on seasonal and synoptic scales are controlled by surface heat flux and ocean advection, respectively (HW02). The shallower the water, the larger the fall transition temperature change, and the larger the role of surface heat fluxes over ocean advection.

PAGE 272

238 Net surface cooling during fall leads to efficient convective mixing that rapidly alters temperature gradients. One unique and counter-intuitive consequence, revealed by analyzing the depth dependence of the temperature budget, is that an upwelling circulation can provide warming (cooling) tendencies near the bottom (the surface) as it transports warmer water landward at depth and cooler water seaward at the surface. Inner-shelf temperature diffusion and ocean advection, by tending to offset each other throughout the water column, results in a nearly vertically uniform (and negative) local rate of change of temperature that leads to a typical de-stratified shelf in fall. The differences in the temperature balances between the fall and spring transitions are found to be closely related to the water column stratification. While ocean advection and diffusion tend to add constructively in spring, they tend to add destructively at many places in fall, thereby mitigating and lengthening the fall transition. Similar to the spring transition, the WFS temperature balances are complex and fully three-dimensional. A corollary statement applies to all other water properties as well.

PAGE 273

REFERENCES Austin, J. A. ( 1999) The role of the alongshore wind stress in the heat balance of the north Carolina inner shelf. J. Geophys. Res. 104, 18187-18203 239 Blumberg, A. F., and G L. Mellor (1987), A description of a three-dimensional coastal ocean circulation model, Three-Dimensional Coastal Ocean Models, Vol. 4, N. Heaps (ed.), 208-233, AGU Washington, D. C. Chu, P. Edmons, N. and Fan, C. (1999), Dynamical mechanisms for the south china sea seasonal circulation and thermohaline varibilities. J. Phys Oceanogr., 29, 29712989 Dever, E., and S. J Lentz (1994), Heat and salt balances over the northern California shelf in winter and Spring, J Geophys. Res., 99, 1,001-16,017 Ezer, T., and G. Mellor (1992), A numerical study of the variability and separation of the Gulf Stream, induced by surface atmospheric forcing and lateral boundary flows. J Phys. Oceanogr. 22, 660-682 He, R. and R. H. Weisberg (2002a), West Florida Shelf circulation and temperature budget for the 1999 spring transition. Cont. Shelf Res. 22, 5, 719-748 He, R. and R. H. Weisberg (2002b), Tides on the West Florida Shelf, J. of Phys. Oceanogr., in press He, R. and R. H. Weisberg (2002c), A Loop Current intrusion case study on the West Florida shelf. J. of Phys. Oc eano gr., revis e d and resubmitted Hetland, R. D., Y. Hsueh, R.R. Leben and P. P. Niiler (1999), A Loop Current induced jet along the edge of the West Florida Shelf Geophys. Res. Lett. 26, 2239-2242 Kourafalou, V. H., L.-Y. O e y J.D. Wang, and T.L. Lee (1996), The fate of river discharge on the continental shelf 1: Mod e ling the river plume and the inner shelf coastal current. J. Geophys. Res., 101, 3415-3434 Kundu, P.K (1976), An analysis of inertial oscillations observed near the Oregon coast. J. Phys. Oceanogr. 6, 879-893

PAGE 274

Lentz, S. J. (1987), A heat budget for the northern California shelf during CODE 2, J. Geophys. Res., 92, 14,491-14,509 Lentz, S. J. and D. C. Chapman (1989), Seasonal difference in the current and temperature variability over the northern California shelf during the Coastal Ocean Dynamics Experiment. J. Geophys. Res., 94, 12,571-12,592 240 Marrnorino, G.O (1982), Wind-forced sea level variability along the West Florida shelf (winter, 1978), J. Phys. Oceanogr. 12,389-40519982 Meyers, S.D., E. M. Siegel, R. H. Weisberg (2000) Observation of currents on the West Florida shelf. Geophys Res. Lett., 28 (10):2037-2040. Mellor, G. L. and T. Yamada (1982), Development of a turbulence closure model for geophysical fluid problems, Rev. Geophys., 20, 851-875, 1982 Orlanski, I. (1976), A simple boundary condition for unbounded hyperbolic flows, J Comput. Phys., 21, 251-269, 1976. Pullen, J.D. (2000), Modeling studies of the coastal circulation off northern California, Ph.D. Dissertation, Oregon State University Smagorinsky, J. (1963), General circulation experiments with primitive equations 1: The basic experiment. Mon. Weather. Rev., 91, 99-164 Walsh, J. J R. H. Weisberg, D. A. Dieterle, R. He., B.P. Darrow, J. K. Jolliff, K. M. Lester, G A. Vargo, G. J. Kirkpatrick, K A. Fanning, T. T. Sutton, A. E. Jochens, D. C. Giggs, B. Nababan, C. Hu, and F. E. MullerKarger, (2002). The phytoplankton response to intrusion of slope water on the west Florida shelf: model and observations. J. Geophy. Res. Submitted Weisberg, R. H, B D. Black, H Yang ( 1996), Seasonal modulation of the West Florida continental shelf circulation, Geophys. Res. Lett., 23, 2247-2250 Weisberg, R. H., B.D. Black, Z. Li (2000), An upwelling case study on the Florida west coast. J. G e ophys. Res. 105(C5) 11459-11469 Weisberg, R. H., Z. Li and F MullerKarger (200 1 ), West Florida Shelf response to local wind forcing: April1998. J. Geophys. Res. 106 (C12), 31239-31262 Weisberg, R. H. and R. He (2002), Anomalous circulation on the west Florida shelf: Local and deep-ocean forcing contribution, J. Geophys. Res. Submitted Yang H, R. H. Weisberg, P P. Niller W. Sturges W Johnson (1999), Lagrangian circulation and forbidden zone on the West Florida Shelf. Cont. Sh elf R es. 19(9), 1221-1245

PAGE 275

CHAPTER 7 SUMMARY 241 Observations and numerical model simulations are used to study the responses of the west Florida shelf to forcing at tidal, synoptic, seasonal and inter-annual time scales. The observations include coastal sea level data, in-situ moored ADCP current measurements, and the current and hydrographic data taken from ship surveys. The model is a regional adaptation of three dimensional, primitive equation sigma coordinate Princeton Ocean Model (POM) which uses an orthogonal curvilinear coordinate system in the horizontal and a sigma coordinate system in the vertical. The model domain extends from the Mississippi River in the northwest to the Florida Keys in the southeast with one open boundary that arcs between these two locations. The principal scientific goals that are pursued include: (1) to describe the west Florida shelf circulation at different time scales; (2) to differentiate between the shelf wide contributions of momentum and of buoyancy fluxes in determining the shelf circulation responses to local forcing; (3) to determine the relative importance of local and deep water forcing, and (4) to describe how the circulation is related to the biological (primary productivity), che mical (nutrient distribution), and geological (s e diment resuspension) processes on the west Florida shelf.

PAGE 276

242 Chapter 2 focuses on barotropic tides and on the four major tidal constituents (Mz, Sz, K1 and 01) that account for the bulk ( -90% )of the WFS tidal variance. The questions answered in that study include how tides distribute over the entire shelf and what their contributions are to mixing and material property transports. It is found that Apalachicola Bay, in Florida Panhandle, is a dividing point between appreciable semi-diurnal tides to the east over the WFS and minimal semi-diurnal tides from there to the Mississippi River delta. In contrast with the semi-diurnal tides, the diurnal tides are spatially more uniform. Co-amplitude maps show largest tides in the regions where the shelf is widest, i.e., in the Florida Big Bend and in Florida Bay. Such spatial distributions of the tides on the WFS are the result of local geometry as contrasted with previous ideas about the basin-wide tidal resonance. Tidal currents on the WFS, especially the semi-diurnal constituents, are primarily barotropic. Tidal residual current in both Eulerian and Lagrangian frameworks show much smaller contribution to the material properties transports, when compared with shelf motions at either synoptic or seasonal scale. With the Mz tide having the largest tidal current magnitude, it is used to examine the spatial distribution of the bottom stress and the role that this may play in the water column mixing. Except for the very shallow regions of the Florida Big Bend and Florida Bay, the potential for mixing by tides is rather weak. Based on these findings, tides are not considered in the following studies of the west Florida shelf motions observed at synoptic, seasonal, and inter-annual time scales. Interactions between the deep water Loop Current and the WFS circulation are the topics in Chapter 3, where, by using in situ current and hydrographic data,

PAGE 277

243 along with satellite analyses of sea surface temperature and sea surface height and idealized numerical model experiments, we study a Loop Current intrusion event that occurred on the west Florida shelf in June 2000. The main question that we explore for this northern deep-water Loop Current intrusion event is: what is the relative importance between contributions of the Loop Current and of the local wind in determining the shelf wide circulation. We find that the Loop Current can produce anomalous hydrography and strong currents near the shelf break, but because of the planetary vorticity constraint of the sloping bottom (the Taylor-Proudrnan theorem), the middle to inner-shelf regions are only slightly affected. Being largely baroclinic, the across-shelf scale of the Loop Current induced current is compatible with the baroclinic Rossby radius of deformation evaluated near the shelf break (20-30 km). The counteracting effects of the baroclinic and barotropic portions of the flow field further inhibit penetration onto the shelf. Therefore, the shelf break currents are largely Loop Current controlled, whereas the middle and inner-shelf currents are largely controlled by local forcing. The Loop Current may interact with the shelf break at any location. Hetland et al. (1999) hypothesize that Loop Current impacts at the southwestern comer of the west Florida shelf (west of Florida Keys, where the shallow isobaths are convergent) can excite currents at shallower depths. Our examinations of in-situ measurements and idealized numerical simulations confirm this hypothesis. Although such southwest comer interactions are significant, they do not appear to occur frequently. Based on long-term current measurements and TOPEXIPOSEIDON time series, we identify four such events during the period between 1992 and 2001, including the

PAGE 278

244 events in spring and summer 1998 (Weisberg and He, 2002) and in fall 1998 (He and Weisberg, 2002d), which are described in detail in Chapters 5 and 6 of this dissertation, respectively. While infrequent, such southwest comer Loop Current impacts can lead to large inter-annual variation in shelf water properties and ecological responses. Spring season features of the west Florida shelf include a mid-shelf southeastward current, cold and low salinity tongues, and a high concentration chlorophyll plume (i.e., the "green river" reported by Gilbes et al., 1996). We hypothesize that these annually occurring features are detennined by the shelf wide local forcing. To test this, Chapter 4 considers the shelf response to the shelf wide local forcing alone in spring 1999, a year when the Loop Current, as evidenced in relatively flat isopycnals topography at the shelf break, did not have strong direct impact on the shelf. The Loop Current is excluded in this study by applying a pure radiation condition along the model open boundary. We find that driven by the local forcing alone, the model results agree well with the in-situ observations, indicating that the shelf circulation indeed responds primarily to the shelf wide local forcing. The mid-shelf southeastward current is accounted for by the combined response to local shelf-wide wind and buoyancy forcing. Wind drives a circulation that tends to be strongest near-shore. Heat flux provides a cyclonic contribution that adds constructively (destructively) at the mid to outer shelf (inner -s helf) thus fonning the observed mid-shelf jet. By advecting nutrient rich river flows (the Mississippi river and others) southeastward, this mid shelf current also fonns the low salinity tongue and the high -c oncentration chlorophyll plume (i.e., the "green river") that both extend

PAGE 279

245 southeastward from Cape San Bias. These annually occurring physical and biological features are produced independent of the Loop Current, indicating that the Loop Current is not a primary factor. In addition, the temperature budget analyses for the spring transition reveal that the surface heat flux largely controls the seasonal transition in the shelf water temperature, whereas ocean advection largely controls the synoptic scale variability in the shelf water temperature. These two processes (surface heat flux and ocean advection) are closely linked, however, and the shelf water temperature balance is fully three-dimensional. An approach similar to that used in the spring 1999 study (Chapter 4) is also applied to the spring 1998 situation. The spring and summer seasons between 1998 and 1999 show distinctively different circulations and water properties, both on the shelf and at the shelf-break. How the inter-annual variability of the shelf circulation and water properties are determined is the question studied in Chapter 5, where we use in situ data and a series of numerical model experiments (with and without the effect of the Loop Current). We account for the observed inter-annual variability by the inter-annual variability in both the local forcing and Loop Current interactions with the shelf Upwelling favorable wind events in 1998 are found to be stronger and longer lasting than those in 1999. The net effect is that more cold, nutrient rich deep waters are upwelled, which in tum, sharpen the thermocline and support stronger shelf wide current fields. Spring to summer 1998 also experiences a Loop Current southwestern comer impact event, which does two things. First, it raises the material isopleth at the shelf break, making it easier for the local, shelf-wide wind and buoyancy forcing to cause these material isopleths to broach th e shelf. Second, it sets

PAGE 280

246 currents in motion on the shelf thereby increasing the landward flow in the bottom Ekman layer and facilitating the landward transport of deep-water material properties to the near shore region. Therefore, the anomalous conditions of 1998 are accounted for by a combination of local and deep-ocean forcing effects. Complementing the spring transition studies discussed in Chapters 4 and 5, Chapter 6 examines the circulation and temperature budget during the fall transition of 1998. Along with the net surface heat flux changing from warming to cooling, the fall season of 1998 also features a Loop Current southwestern comer effect, whose effect is accounted for by twin model experiments, one with and the other without the Loop Current. We find that local forcing largely controls the inner-shelf circulation, which is very similar for both cases. The effect of the Loop Current in fall 1998 is to reinforce the mid-shelf currents and increase the bottom Ekman layer across-shelf transports, thereby contributing to the landward transport of cold, nutrient rich water of deep-ocean origin. Temperature budget analysis shows that in contrast with the spring transition, surface heat flux in the fall transition largely controls the temperature variability at both the seasonal and synoptic scales Such seasonality of the shelf water temperature budget is closely related to the seasonality of the water stratification. The net surface cooling during fall leads to convective mixings that rapidly alter the across shelf temperature gradients. One counter intuitive consequence is that the upwelling circulation, by transporting relative warmer (cooler) water landward (seaward) at depth (surface), provides warming (cooling) tendencies near the bottom (the surface), thereby mitigating and l e ngthening the fall

PAGE 281

247 transition. As in the spring, the shelf water temperature budgets in fall are also fully three-dimensional. In summary, by combining both in-situ measurements with numerical model simulations, this dissertation provides an improved description and understanding of the west Florida shelf circulations that occur at different time scales and how these circulations determine the material property distributions found on the shelf. Building upon previous modeling studies that were limited to idealized wind forcing (e.g., climatological or spatially uniform wind stresses), this dissertation considers the effects of spatial and temporal varying winds, surface heat fluxes, river runoffs, and idealized Loop Current interactions with the shelf. By comparing model experiments with and without the deep ocean forcing against the actual data, this dissertation begins to develop a quantitative understanding of the shelf circulation and address the importance of both local, shelf-wide forcing and forcing by deep ocean processes. Despite the fact that the model simulations in this dissertation represent the observed data reasonably well, several major limitations are identified. Compared with the in-situ data, the model generally underestimates the synoptic scale variability of the currents by 20-50 %. This may be due to the model surface forcing fields derived from the NCEP reanalysis product. The coarse resolution of the product renders the net surface heat flux and the boundary layer transition from winds over land to winds over water incorrect. Along with these external surface boundary conditions, potential problems internal to the model are its pararneterizations and numerical schemes. For instance, bottom stress fields, subject to the choices of both sigma layer distributions and the roughness thickness, might be overestimated. Also,

PAGE 282

248 the MellorYamada turbulence parameterization, being semi-empirical, might not be sufficient for certain physical processes. Moreover, since turbulence is critically tied to the stratification failure of the model to maintain proper stratification, either through parameterizations or through the surface boundary condition, will feedback negatively on model performance. Numerically, the simple center difference advection scheme used in the model provides greater numerical diffusion than would exist in the reality, thereby failing to maintain a sharp thermocline that might otherwise support stronger current fields. Improving model simulations will therefore be part of future research work. This will necessitate the use of more in situ measurements and remotely sensed observations, including current profiles, dynamic heights and density fields, and the surface winds and fluxes across the entire shelf and the shelf break. These measurements are important by themselves since they provide better descriptions of the shelf circulation and the deep ocean interactions with the shelf. They also provide the initial and boundary conditions for the model and the basis for testing model performance and for guiding model improvements when discrepancies occur. Sensitivity experiments and comparisons with higher order advection schemes (e.g., Smolarkiewicz anti-diffusion scheme) and different turbulence mixing parameterizations (e.g. KPP) are necessary. Similarly, comparisons between this model and other coastal ocean models (e.g ROMS and FVM) will be of signific a nce. Techniques including data assimilation, inverse calculations, and regional nesting (to better account for the Loop Current interactions with the shelf) must b e adv a nced to further improve this (or any) model's ability to simulate the west Florida sh elf

PAGE 283

249 circulation. Finally, the coupling of ocean and atmosphere interaction models and the coupling of physical and biogeochemical models will be needed to more fully account for the shelf physics and how these physics contribute to the myriad of interesting and societally important ecological questions.

PAGE 284

250 APPENDICES

PAGE 285

251 APPENDICE A. THREE-DIMENSIONAL TEMPERATURE BUDGET DIAGNOSIS IN THE PRINCETON OCEAN MODEL For an incompressible fluid in a right-hand general orthogonal coordinate system, in a frame of reference rotating at an angular velocity f (f = 2Q sin 8) about the vertical coordinate z. Invoking hydrostatic and Boussinesq approximation to effect simplification, the temperature (or other state variable) equation is: ar ar ar ar a ar a ar a ar -=-u--v--w-+-(A -)+-(A -)+-(K -) (A-1) at ax ay az ax H ax ay H ay az H az which equates the local rate of change of temperature ar ;at to a combination of the flow field advective rate of change -u ar jax-vaT jay-waT jaz, and the rates of change by horizontal diffusion ajax(AH aT jax) + ajay(AH aT jay) and vertical diffusion a;az(KH aTjaz). In the Princeton Ocean Model (POM), the above equation is transformed to a topographically conformal a coordinate system in the vertical using the following transformations to go from the (x, y, z, t) to (x', y', a, t') system, where x=x', y=y' and t=t' and T/(X, y)-Z (j = ---'--'--"--'--H (x, y) + 77(x, y) (A-2) Since all the derivatives in POM are calculated in term of a coordinate, they are recast to z coordinate to form the temperature equation (A-1) The derivatives of temperature in the two coordinate systems are related thus:

PAGE 286

APPENDICE A (Continued) aT= aT_ aT}) ax ax aa D ax D ax aT aT aT (a aD 1 aT} ay-ay-aa Day+ Day,) aT 1 aT -=--252 (A-3) where D=H + TJ, the total depth of the water column, H is the bottom depth, and T} is the free surface deflection. It is noted that pseudo-vertical velocity w in crcoordinate are different from the "real" vertical velocity w in z-coordinate Therefore w is recalculated as aD aT} aD aT} aD aT} w = w+u(a-+-)+v(a-+-)+(a-+-) ax ax ay ay at at (A-4) before it is used to calCulate the vertical advection term waT ja z The source code of the subroutine named heatZf is attached. This subroutine should be called after calculating the temperature and salinity fields (i. e after subroutines ADVT and PROFr). It then performs the temperature (or salinity) diagnosis at each time step.

PAGE 287

cc*********************************************** cc== Temperature/Salt Budget Diagnostic Analysis cc== cc== Version 11/25/01. Ruoying He. cc== cc== References: cc== 1. Mellor & Blumberg, 1985 cc== 2. Kantha & Clayson, 2001 cc== 3. He & Weisberg, 2001 cc== cc== Please direct any questions or comments to cc== Ruoying He ruoying@marine.usf.edu c************************************************ subroutine heatZ(FB,F,FD,FCLIM,DTI2) include 'comblk98.h' DIMENSION FB(IM,JM,KB),F(IM,JM,KB) DIMENSION FF(IM,JM,KB),FCLIM(IM,JM,KB) DIMENSION FD(IM,JM,KB) REAL RHDIFX(IM,JM,KB),RHDIFY(IM,JM,KB) c--------------------------------------------------------c-Initialize the values DO 500 I=1,IM DO 500 J=1,JM TPS(I,J) DO 500 K=1,KB HADVX(I,J,K) HADVY(I,J,K) HADVZ(I,J,K) HAMX(I,J,K) HAMY(I,J,K) HDTDT(I,J,K) 500 CONTINUE DO 529 J=1,JM =O.EO =O.EO =O.EO = O.EO = O.EO = O.EO =O.EO DO 529 I=1,IM F(I,J,KB)=F(I,J,KBM1) 529 FB(I,J,KB)=FB(I,J,KBM1) DO 2 K=1,KBM1 Do 100 J = 1, JM Do 100 I = 1, IM TPS(I,J) = ZZ(K) DT(I,J) + ELF(I,J) 100 CONTINUE DO 10 J=2,JMM1 DO 10 I=2,IMM1 DXR 2. I (DX(I+l,J)+DX(I,J)) DXL 2. I (DX(I,J)+DX(I-l,J)) DYT 2. I (DY(I,J+1)+DY(I,J)) DYB 2. I (DY(I,J)+DY(I,J-1)) cc>> Additional terms form sigma2Z transformation cc=========================================================== 253

PAGE 288

cc 1/D*dS/dsigma AAl=(F(I,J,K)-F(I,J,K+l))/(DT(i,j)*DZ(k)) cc sigma*dD/dx+dEl/dx AA2=0.5*((TPS(I+l,J)-TPS(I,J))*DXR + +(TPS(I,J)-TPS(I-l,J) )*DXL) cc sigma*dD/dy+dEl/dy AA3=0.5*((TPS(I,J+l)-TPS(I,J))*DYT + +(TPS(I,J)-TPS(I,J-l))*DYB) cc sigma*dD/dt+dEl/dt AA4=(l.+ZZ(K))*(ELF(I,J)-EL(I,J))/DTI cc-->> Diagnostic analysis cc=========================================================== cc-->>1. Horizontal +vertical advections HADVX(I,J,K)=0.5*(U(I+l,J,K)*(F(I+l,J,K)-F(I,J,K))*DXR + +U(I,J,K)*(F(I,J,K)-F(I-l,J,K))*DXL) + -0.5*(U(I,J,K)+U(I+l,J,K))*AAl*AA2 HADVY(I,J,K)=0.5*(V(I,J+l,K)*(F(I,J+l,K)-F(I,J,K))*DYT + +V(I,J,K)*(F(I,J,K)-F(I,J-l,K))*DYB) + -0.5*(V(I,J+l,K)+V(I,J,K))*AAl*AA3 HADVZ(I,J,K)=0.5*(WZ(I,J,K)+WZ(I,J,K+l)) + *(F(I,J,K)-F(I,J,K+l))/(DT(I,J)*DZ(K)) cc-->>2. Local change HDTDT(I,J,K)=(FD(I,J,K)-FB(I,J,K))/DTI2 + -AAl*AA4 10 CONTINUE 2 CONTINUE cc-->>3. Horizontal diffusion DO 99 K=l,KB DO 99 J=l,JM DO 99 I=l,IM 99 FB(I,J,K)=FB(I,J,K)-FCLIM(I,J,K) do 300 K=l,KBMl do 300 J=2,JM do 300 I=2,IM RHDIFX(I,J,K)= 1 -.5EO*(AAM(I,J,K)+AAM(I-l,J,K))*(H(I,J)+H(I-l,J))*TPRNI 2 *(FB(I,J,K)-FB(I-l,J,K))*DUM(I,J)/(DX(I,J)+DX(I-l,J)) RHDIFY(I,J,K)= 1 -.5EO*(AAM(I,J,K)+AAM(I,J-l,K))*(H(I,J)+H(I,J-l))*TPRNI 2 *(FB(I,J,K)-FB(I,J-l,K))*DVM(I,J)/(DY(I,J)+DY(I,J-1)) RHDIFX(I,J,K)=RHDIFX(I,J,K)*.5EO*(DY(I,J)+DY(I-l,J)) RHDIFY(I,J,K)=RHDIFY(I,J,K)*.SEO*(DX(I,J)+DX(I,J-1)) 300 continue DO 101 K=l,KB DO 101 J=l,JM DO 101 I=l,IM 254

PAGE 289

101 FB(I,J,K)=FB(I,J,K)+FCLIM(I,J,K) DO 102 K=1,KBM1 DO 102 I=2,IMM1 DO 102 J=2,JMM1 AREA=((H(I,J)+ETF(I,J))*ART(I,J)) HAMX(I,J,K)=(RHDIFX(I+1,J,K)-RHDIRX(I,J,K))/AREA HAMY(I,J,K)=(RHDIFY(I,J+1,K)-RHDIFY(I,J,K))/AREA 102 CONTINUE cc-->> Times 86400, Convert c/second to c/day DO 103 K=1,KB DO 103 I=1,IM DO 103 J=1,JM HADVX(I,J,K)=HADVX(I,J,K)*86400 HADVY(I,J,K)=HADVY(I,J,K)*86400 HADVZ(I,J,K)=HADVZ(I,J,K)*86400 HAMX(I,J,K) =HAMX(I,J,K)*86400 HAMY(I,J,K) =HAMY(I,J,K)*86400 HDTDT(I,J,K)=HDTDT(I,J,K)*86400 103 CONTINUE RETURN END 255

PAGE 290

ABOUT THE AUTHOR Ruoying He received his Bachelor of Science Degree in Oceanography from Ocean University of Qingdao, China in 1996. He joined College of Marine Science and continued his Ph.D. program in physical oceanography at the University of South Florida in 1998. He is an author of six research papers and has presented at more than 10 conferences during his four-year graduate study. His research interests include continental shelf circulation and dynamics, boundary layer dynamics, ocean tides, inverse methods in oceanography, coupled physical-biological dynamics and remote sensing oceanography. He has expertise in numerical modeling of ocean circulation at different time scales, particle and tracer tracking, data assimilation, and coastal ocean nowcast and forecast system. He has participated in twelve research cruises and has been actively involved in several major research projects, including EcoHAB, HyCODE and COMPS. He is the recipient of Marine Science Paul Getting Fellowship, Robert Gerral Fellowship and Knight Oceanographic Fellowship in 1999, 2000 and 2001, respectively.