USF Libraries
USF Digital Collections

Effects of extreme events on residual circulation for Tampa Bay, Florida

MISSING IMAGE

Material Information

Title:
Effects of extreme events on residual circulation for Tampa Bay, Florida
Physical Description:
Book
Language:
English
Creator:
Wilson, Monica
Publisher:
University of South Florida
Place of Publication:
Tampa, Fla
Publication Date:

Subjects

Subjects / Keywords:
Hurricane
Estuary
Ocean model
Hurricane Frances
Hurricane Jeanne
Dissertations, Academic -- Marine Science -- Masters -- USF   ( lcsh )
Genre:
bibliography   ( marcgt )
theses   ( marcgt )
non-fiction   ( marcgt )

Notes

Abstract:
ABSTRACT: A numerical circulation model of Tampa Bay, Florida is used to simulate the flow field and tidal residual circulation for 2001-2004. This model is used to investigate the effects of extreme events on the residual circulation of the bay. The three extreme events that are used in this study are: Hurricane Frances, Hurricane Jeanne, and an extreme winter frontal passage that occurred on December 26, 2004. Each extreme event was divided into phases that were chosen by wind peaks and times of velocity inflow and outflow. There were three phases to the hydrodynamics effect of Frances on Tampa Bay. Hurricane Jeanne and the winter frontal passage each had two phases. An important difference between the three extreme events is the duration of each; Hurricane Frances lasted approximately two and a half days, Hurricane Jeanne affected the bay area for about twenty-four hours, and the extratropical storm passed within 16 hours.^ ^Winds were six standard deviations higher than the 2004 mean (4.06 m s-1) during Hurricane Frances, and seven standard deviations higher during both Hurricane Jeanne and the extratropical storm. Water levels reached four standard deviations during Hurricane Frances and the extratropical storm, and two standard deviations during Hurricane Jeanne. The difference between these results is due to the timing of each event with the tides, whether it was in or out of phase with the tides. During phase 2 of Hurricane Frances there was a total volume inflow of m3, for an increase of 60% in bay volume. There was a total volume outflow during phase 3 of m3, a 28% decrease. During Hurricane Jeanne there was a total volume inflow of m3 (30% increase) and total volume outflow of m3 (14% decrease). The extratropical storm showed a total volume inflow of m3 (29% increase) and a total volume outflow of m3 (31% decrease).^ ^Though the increase and decrease of volume for each event was different, they all had the same affect on the bay, causing changes in the residual circulation over time scales of these extreme events.
Thesis:
Thesis (M.S.)--University of South Florida, 2007.
Bibliography:
Includes bibliographical references.
System Details:
System requirements: World Wide Web browser and PDF reader.
System Details:
Mode of access: World Wide Web.
Statement of Responsibility:
by Monica Wilson.
General Note:
Title from PDF of title page.
General Note:
Document formatted into pages; contains 145 pages.

Record Information

Source Institution:
University of South Florida Library
Holding Location:
University of South Florida
Rights Management:
All applicable rights reserved by the source institution and holding location.
Resource Identifier:
aleph - 001921345
oclc - 191048684
usfldc doi - E14-SFE0001888
usfldc handle - e14.1888
System ID:
SFS0026206:00001


This item is only available as the following downloads:


Full Text
xml version 1.0 encoding UTF-8 standalone no
record xmlns http:www.loc.govMARC21slim xmlns:xsi http:www.w3.org2001XMLSchema-instance xsi:schemaLocation http:www.loc.govstandardsmarcxmlschemaMARC21slim.xsd
leader nam Ka
controlfield tag 001 001921345
003 fts
005 20080122142821.0
006 m||||e|||d||||||||
007 cr mnu|||uuuuu
008 080122s2007 flu sbm 000 0 eng d
datafield ind1 8 ind2 024
subfield code a E14-SFE0001888
035
(OCoLC)191048684
040
FHM
c FHM
049
FHMM
090
GC11.2 (ONLINE)
1 100
Wilson, Monica.
0 245
Effects of extreme events on residual circulation for Tampa Bay, Florida
h [electronic resource] /
by Monica Wilson.
260
[Tampa, Fla] :
b University of South Florida,
2007.
3 520
ABSTRACT: A numerical circulation model of Tampa Bay, Florida is used to simulate the flow field and tidal residual circulation for 2001-2004. This model is used to investigate the effects of extreme events on the residual circulation of the bay. The three extreme events that are used in this study are: Hurricane Frances, Hurricane Jeanne, and an extreme winter frontal passage that occurred on December 26, 2004. Each extreme event was divided into phases that were chosen by wind peaks and times of velocity inflow and outflow. There were three phases to the hydrodynamics effect of Frances on Tampa Bay. Hurricane Jeanne and the winter frontal passage each had two phases. An important difference between the three extreme events is the duration of each; Hurricane Frances lasted approximately two and a half days, Hurricane Jeanne affected the bay area for about twenty-four hours, and the extratropical storm passed within 16 hours.^ ^Winds were six standard deviations higher than the 2004 mean (4.06 m s-1) during Hurricane Frances, and seven standard deviations higher during both Hurricane Jeanne and the extratropical storm. Water levels reached four standard deviations during Hurricane Frances and the extratropical storm, and two standard deviations during Hurricane Jeanne. The difference between these results is due to the timing of each event with the tides, whether it was in or out of phase with the tides. During phase 2 of Hurricane Frances there was a total volume inflow of m3, for an increase of 60% in bay volume. There was a total volume outflow during phase 3 of m3, a 28% decrease. During Hurricane Jeanne there was a total volume inflow of m3 (30% increase) and total volume outflow of m3 (14% decrease). The extratropical storm showed a total volume inflow of m3 (29% increase) and a total volume outflow of m3 (31% decrease).^ ^Though the increase and decrease of volume for each event was different, they all had the same affect on the bay, causing changes in the residual circulation over time scales of these extreme events.
502
Thesis (M.S.)--University of South Florida, 2007.
504
Includes bibliographical references.
516
Text (Electronic thesis) in PDF format.
538
System requirements: World Wide Web browser and PDF reader.
Mode of access: World Wide Web.
500
Title from PDF of title page.
Document formatted into pages; contains 145 pages.
590
Adviser: Mark E. Luther, Ph.D.
653
Hurricane.
Estuary.
Ocean model.
Hurricane Frances.
Hurricane Jeanne.
690
Dissertations, Academic
z USF
x Marine Science
Masters.
773
t USF Electronic Theses and Dissertations.
4 856
u http://digital.lib.usf.edu/?e14.1888



PAGE 1

Effects of Extreme Events on Residual Circulation f or Tampa Bay, Florida by Monica Wilson A thesis submitted in partial fulfillment of the requirements for the degree of Master of Science Department of Marine Science College of Marine Science University of South Florida Major Professor: Mark E. Luther, Ph.D. Steven D. Meyers, Ph.D. Robert H. Weisberg, Ph.D. Date of Approval: March 22, 2007 Keywords: hurricane, estuary, ocean model, Hurrican e Frances, Hurricane Jeanne Copyright 2007, Monica Wilson

PAGE 2

i Table of Contents List of Tables..................................... ................................................... ..............................ii List of Figures.................................... ................................................... .............................iii Abstract........................................... ................................................... .................................x Chapter One Introduction.......................... ................................................... .....................1 Chapter Two Background............................ ................................................... ...................8 Tampa Bay.......................................... ................................................... .................8 ECOM-3D............................................ ................................................... ..............13 Hurricane Frances.................................. ................................................... ............15 Hurricane Jeanne................................... ................................................... .............18 December 26, 2004 Extratropical Storm.............. ................................................21 Chapter Three Data Collection and Methods......... ................................................... .......23 Data Collection.................................... ................................................... ..............23 Methods............................................ ................................................... ..................32 Chapter Four Results.............................. ................................................... .......................40 Hurricane Frances.................................. ................................................... ............40 Hurricane Jeanne................................... ................................................... .............75 Extratropical Storm................................ ................................................... ..........103 Chapter Five Summary and Discussion............... ................................................... .......134 References......................................... ................................................... ...........................142

PAGE 3

ii List of Tables Table 1. Locations for each model freshwater point sources (Meyers et al., 2007)........31 Table 2. Time duration, maximum and minimum wind s peeds, mean wind direction, and total volume flow for each phase dur ing all three extreme events............33 Table 3. Tidal constituents gathered from www.tide sandcurrents.noaa.gov, St. Petersburg station 8726520, used in least square an alysis............................................. ...36 Table 4. Depth average, depth varying, and ration at five different locations throughout the bay at times of maximum inflow (top) and maximum outflow (bottom) during Hurricane Frances.................. ................................................... ..............67 Table 5. Depth average, depth varying, and ratio a t five different locations throughout the bay at times of maximum inflow (top) and maximum outflow (bottom) during Hurricane Jeanne................... ................................................... ..............99 Table 6. Depth average, depth varying, and ratio a t five different locations throughout the bay at times of maximum inflow (top) and maximum outflow (bottom) during the Extratropical Storm............ ................................................... ..........128

PAGE 4

iii List of Figures Figure 1. Map of Tampa Bay........................ ................................................... ..................9 Figure 2. Model bathymetry map of Tampa Bay....... ................................................... ...11 Figure 3. Complete path of Hurricane Frances...... ................................................... .......17 Figure 4. Complete path of Hurricane Jeanne....... ................................................... .......20 Figure 5. Map of Tampa Bay PORTS stations......... ................................................... ....25 Figure 6. Map of COMPS stations................... ................................................... .............26 Figure 7. CCUT wind speeds and directions versus f ive airport sites around Tampa Bay.......................................... ................................................... ...........................29 Figure 8. Map of model freshwater point sources (i n red)............................................. .30 Figure 9. ADCP echo amplitudes as a function of ra nge for all four beams...................34 Figure 10. Model velocity and elevation (black) ov erplotted with the non-tidal component (blue) and a 25-hr low pass filter (red). ................................................... .......37 Figure 11. Meteorological data for the month of Se ptember 2004: wind speed (top panel), wind direction (middle panel), and water le vel (bottom panel)............................41 Figure 12. Instantaneous ADCP velocities (dotted l ine) versus model axial current (solid line) during Hurricane Frances. Yearly mean for each level (thin solid line) and (dashed line) aer shown as well................... ................................................... ......42 Figure 13. Residual ADCP velocities (dotted line) overplotted with model residual axial current (solid line) during Hurricane Frances Yearly mean for each level (thin solid line) and (dashed line) are shown as well.................... ...............................44 Figure 14. Model velocity (black) overplotted with tidal fit (blue) and the difference between the model and tidal fit (red)... ................................................... .........45

PAGE 5

iv Figure 15. Vertical RMS velocity for all model lay ers during Hurricane Frances. Vertical RMS velocity was calculated for the 3 days during Hurricane Frances (red), the tidal component of the same 3 days (blue ), for averaged Septembers 2001-2003 (green), phase 3 (orange), phase 2 (yello w), and phase 1 (brown).................47 Figure 16. Maximum surface inflow during phase 2 o f Hurricane Frances. The tidal component is show on the left, the total component is in the middle, and the difference between total and tidal componen ts is on the right.............................48 Figure 17. Maximum mid-depth inflow during phase 2 of Hurricane Frances. The tidal component is show on the left, the total component is in the middle, and the difference between total and tidal componen ts is on the right.............................49 Figure 18. Maximum bottom inflow during phase 2 of Hurricane Frances. The tidal component is show on the left, the total component is in the middle, and the difference between total and tidal componen ts is on the right.............................50 Figure 19. Maximum surface outflow during phase 3 of Hurricane Frances. The tidal component is show on the left, the total component is in the middle, and the difference between total and tidal componen ts is on the right.............................52 Figure 20. Maximum mid-depth outflow during phase 3 of Hurricane Frances. The tidal component is show on the left, the total component is in the middle, and the difference between total and tidal componen ts is on the right.............................54 Figure 21. Maximum bottom outflow during phase 3 o f Hurricane Frances. The tidal component is show on the left, the total component is in the middle, and the difference between total and tidal componen ts is on the right.............................55 Figure 22. Baywide elevation at the end of phase 2 during Hurricane Frances..............56 Figure 23. Baywide elevation at the end of phase 3 during Hurricane Frances..............57 Figure 24. Grid points used to evaluate salinity a t different locations in the bay............58 Figure 25. Line plots of surface salinity at diffe rent locations in the bay. The 2004 mean for each location is shown in the thin bl ack line, and 1 are shown by the dashed lines................................ ................................................... .........................59 Figure 26. Baywide view of surface salinity at the end of phase 2 during Hurricane Frances............................................ ................................................... ...............................61 Figure 27. Baywide view of surface salinity at the end of phase 3 during Hurricane Frances............................................ ................................................... ...............................62

PAGE 6

v Figure 28. Salinity cross sections at the end of p hase 2 during Hurricane Frances. Cross sections are from the Sunshine Skyway, Middle Tampa Bay, the mouth of Hillsborough, and the mouth of Old Tampa Bay....... ................................................... ....63 Figure 29. Salinity cross sections at the end of p hase 3 during Hurricane Frances.........64 Figure 30. Vertical velocity profiles (thick solid line) at five locations in the bay at the time of maximum inflow during Hurricane Fran ces. Depth average is shown by the thin solid line............................. ................................................... ..........................66 Figure 31. Vertical velocity profile (thick solid line) at the time of maximum outflow during Hurricane Frances. The depth averag e is shown by the thin solid line............................................... ................................................... ...................................68 Figure 32. Particles released during phase 2. Par ticles were released throughout the water column at the mouth of Tampa Bay, Middle Tampa Bay, and in the center of Hillsborough Bay and Old Tampa Bay....... ................................................... ....70 Figure 33. Particles released during phase 3...... ................................................... ...........71 Figure 34. Observed water level from the St. Peter sburg station (blue line) and Port Manatee station (red line) compared to model d ata at St. Petersburg (black line) and Port Manatee (green line). The top panel represents water levels during Hurricane Frances, the middle panel represents wate r levels during Hurricane Jeanne, and the bottom panel represents the Extratr opical Storm. Observed water levels are from www.tideandcurrents.noaa.gov....... ................................................... ......72 Figure 35. Total model bay volume during Hurricane Frances (top panel), Hurricane Jeanne (middle panel) and the Extratropic al Storm (bottom panel). The thick solid black line represents the total compone nt of the extreme events and the blue line is the tidal component................... ................................................... ..................74 Figure 36. Instantaneous ADCP (dotted line) veloci ties versus model axial current (solid line) during Hurricane Jeanne. Yearly mean for each level (thin solid line) and (dashed line) are shown as well................... ................................................... ......76 Figure 37. Residual ADCP (dotted line) overplotted with model residual axial current (solid line) during Hurricane Jeanne. Year ly mean for each level (thin solid line) and (dashed line) are shown as well................... ........................................78 Figure 38. Model velocity (black) overplotted with tidal fit (blue) and the difference between model and tidal (red)........... ................................................... ...........79

PAGE 7

vi Figure 39. Vertical RMS velocity for all model lay ers during Hurricane Jeanne. Vertical RMS velocity was calculated for the 3 days during Hurricane Jeanne (red), the tidal component of the same 3 days (blue), for averaged Septembers 2001-2003 (green), phase 2 (orange), and phase 1 (yellow).... ................................................... ........80 Figure 40. Maximum surface inflow during phase 1 o f Hurricane Jeanne. The tidal component is show on the left, the total comp onent is in the middle, and the difference between total and tidal components i s on the right....................................8 2 Figure 41. Maximum mid-depth inflow during phase 1 of Hurricane Jeanne. The tidal component is show on the left, the total component is in the middle, and the difference between total and tidal componen ts is on the right.............................83 Figure 42. Maximum bottom inflow during phase 1 of Hurricane Jeanne. The tidal component is show on the left, the total comp onent is in the middle, and the difference between total and tidal components i s on the right....................................8 4 Figure 43. Maximum surface outflow during phase 2 of Hurricane Jeanne. The tidal component is show on the left, the total component is in the middle, and the difference between total and tidal componen ts is on the right.............................85 Figure 44. Maximum mid-depth outflow during phase 2 of Hurricane Jeanne. The tidal component is show on the left, the total component is in the middle, and the difference between total and tidal componen ts is on the right.............................86 Figure 45. Maximum bottom outflow during phase 2 o f Hurricane Jeanne. The tidal component is show on the left, the total component is in the middle, and the difference between total and tidal componen ts is on the right.............................87 Figure 46. Baywide elevation at the end of phase 1 during Hurricane Jeanne................89 Figure 47. Baywide elevation at the end of phase 2 during Hurricane Jeanne................90 Figure 48. Line plots of surface salinity at diffe rent locations in the bay. The 2004 mean for each location is shown in the thin bl ack line, and 1 are shown by the dashed lines................................ ................................................... .........................91 Figure 49. Baywide salinity at the end of phase 1 during Hurricane Jeanne...................93 Figure 50. Baywide salinity at the end of phase 2 during Hurricane Jeanne...................94 Figure 51. Salinity cross sections at the end of p hase 1 during Hurricane Jeanne..........95 Figure 52. Salinity cross sections at the end of p hase 2 during Hurricane Jeanne..........96

PAGE 8

vii Figure 53. Vertical velocity profile (thick solid line) at the time of maximum inflow during Hurricane Jeanne. The depth average is shown by the thin solid line......97 Figure 54. Vertical velocity profile (thick solid line) at the time of maximum outflow during Hurricane Jeanne. The depth average is shown by the thin solid line............................................... ................................................... .................................100 Figure 55. Particles released during phase 1. Par ticles were released throughout the water column at the mouth of Tampa Bay, Middle Tampa Bay, and in the center of Hillsborough Bay and Old Tampa Bay....... ................................................... ..101 Figure 56. Particles released during phase 2...... ................................................... .........102 Figure 57. Meteorological data for the month of De cember 2004. The top panel shows winds speed with yearly mean (thin solid hori zontal line) and (dashed line), the middle panel shows wind direction, and t he bottom panel shows water level with mean and ................................................... ................................................105 Figure 58. Instantaneous ADCP (dotted line) veloci ties versus model axial current (solid line) during Hurricane Jeanne. Yearly mean for each level (thin solid line) and (dashed line) are shown as well................... ................................................... ....106 Figure 59. Residual ADCP (dotted line) overplotted with model residual axial current (solid line) during the Extratropical Storm Yearly mean for each level (thin solid line) and (dashed line) are shown as well................... .............................107 Figure 60. Model velocity (black) overplotted with tidal fit (blue) and the difference between model and tidal (red)........... ................................................... .........108 Figure 61. Vertical RMS velocity for all model lay ers during the Extratropical Storm. Vertical RMS velocity was calculated for th e 3 days during the Extratropical Storm (red), the tidal component of t he same 3 days (blue), for averaged Septembers 2001-2003 (green), phase 2 (ora nge), and phase 1 (yellow).......110 Figure 62. Maximum surface inflow during phase 1 o f the Extratropical Storm. The tidal component is show on the left, the total component is in the middle, and the difference between total and tidal componen ts is on the right...........................111 Figure 63. Maximum mid-depth inflow during phase 1 of the Extratropical Storm. The tidal component is show on the left, th e total component is in the middle, and the difference between total and tidal components is on the right..............113 Figure 64. Maximum bottom inflow during phase 1 of the Extratropical Storm. The tidal component is show on the left, th e total component is in the middle, and the difference between total and tidal components is on the right..............114

PAGE 9

viii Figure 65. Maximum surface outflow during phase 2 of the Extratropical Storm. The tidal component is show on the left, th e total component is in the middle, and the difference between total and tidal components is on the right..............115 Figure 66. Maximum mid-depth outflow during phase 2 of the Extratropical Storm. The tidal component is show on the left, th e total component is in the middle, and the difference between total and tidal components is on the right..............116 Figure 67. Maximum bottom outflow during phase 2 o f the Extratropical Storm. The tidal component is show on the left, th e total component is in the middle, and the difference between total and tidal components is on the right..............117 Figure 68. Baywide elevation at the end of phase 1 of the Extratropical Storm...........118 Figure 69. Baywide elevation at the end of phase 2 ................................................... ...119 Figure 70. Line plots of surface salinity at diffe rent locations in the bay for December 2004. The 2004 mean for each location is shown in the thin black line, and 1 are shown by the dashed lines..................... ................................................... ..121 Figure 71. Baywide salinity at the end of phase 1 during Hurricane Jeanne.................122 Figure 72. Baywide salinity at the end of phase 2 during Hurricane Jeanne.................123 Figure 73. Salinity cross sections at the end of p hase 1 during the Extratropical Storm.............................................. ................................................... ..............................124 Figure 74. Salinity cross sections at the end of p hase 2 during the Extratropical Storm.............................................. ................................................... ..............................125 Figure 75. Vertical velocity profile (thick solid line) at the time of maximum inflow during the Extratropical Storm. The depth a verage is shown by the thin solid line......................................... ................................................... ..............................127 Figure 76. Vertical velocity profile (thick solid line) at the time of maximum outflow during the Extratropical Storm. The depth average is shown by the thin solid line......................................... ................................................... ..............................129 Figure 77. Particles released during phase 1. Par ticles were released throughout the water column at the mouth of Tampa Bay, Middle Tampa Bay, and in the center of Hillsborough Bay and Old Tampa Bay....... ................................................... ..131 Figure 78. Particles released during phase 2...... ................................................... .........132

PAGE 10

ix Figure 79. Total number of particles in Tampa Bay during September 2002 (top) and September 2004 (bottom)........................ ................................................... ..............139 Figure 80. Total number of particles in Tampa Bay during December 2002 (top) and December 2004 (bottom)......................... ................................................... .............140

PAGE 11

x Effects of Extreme Events on the Residual Circulati on for Tampa Bay, Florida Monica Wilson ABSTRACT A numerical circulation model of Tampa Bay, Florida is used to simulate the flow field and tidal residual circulation for 2001-2004. This model is used to investigate the effects of extreme events on the residual circulati on of the bay. The three extreme events that are used in this study are: Hurricane Frances Hurricane Jeanne, and an extreme winter frontal passage that occurred on December 26 2004. Each extreme event was divided into phases that were chosen by wind peaks and times of velocity inflow and outflow. There were three phases to the hydrodynam ics effect of Frances on Tampa Bay. Hurricane Jeanne and the winter frontal passage eac h had two phases. An important difference between the three extreme events is the duration of each; Hurricane Frances lasted approximately two and a half days, Hurricane Jeanne affected the bay area for about twenty-four hours, and the extratropical stor m passed within 16 hours. Winds were six standard deviations higher than the 2004 mean ( 4.06 m s-1) during Hurricane Frances, and seven standard deviations higher during both Hu rricane Jeanne and the extratropical storm. Water levels reached four standard deviatio ns during Hurricane Frances and the extratropical storm, and two standard deviations du ring Hurricane Jeanne. The difference between these results is due to the timing of each event with the tides, whether it was in or out of phase with the tides. During phase 2 of Hurricane Frances there was a total

PAGE 12

xi volume inflow of 910 02.2 m3, for an increase of 60% in bay volume. There was a total volume outflow during phase 3 of 910 02.1 m3, a 28% decrease. During Hurricane Jeanne there was a total volume inflow of 910 12.1 m3 (30% increase) and total volume outflow of 810 0.5 m3 (14% decrease). The extratropical storm showed a total volume inflow of 910 04.1 m3 (29% increase) and a total volume outflow of 810 1.1 m3 (31% decrease). Though the increase and decrease of vol ume for each event was different, they all had the same affect on the bay, causing changes in the residual circulation over time scales of these extreme events.

PAGE 13

1 Chapter One Introduction An estuary is defined as “a semi-enclosed coastal body of water having a free connection to the open sea and within which the sea water is measurably diluted with fresh water deriving from land drainage” (Cameron a nd Pritchard, 1963). Estuaries are regions of transition from rivers to the open ocean and are characterized by the possibility of tidal motions from the ocean and gradients of sa linity and density associated with the mixture of river water and sea water. The force of gravity on the density difference between seawater and freshwater tends to cause vert ical salinity stratification and a convective flow, known as “estuarine circulation”. The geomorphology, freshwater flow, and tides are all dominant variables in determining salinity distribution and circulation within an estuary (Hansen and Rattray, 1966). Estu aries are traditionally classified according to their geomorphology; however, the clas sification as a sequence of mixing types is usually applied to coastal plain estuaries Stommel (1951) classified estuaries as vertically mixed, slightly stratified, highly strat ified and salt wedge estuaries. Since salinity plays a major role in determining the dens ity structure in an estuary, these classifications are based on the salinity distribut ions. Tampa Bay is an example of a vertically well mixed drowned coastal plain river v alley. Extreme weather events may alter the flushing of th e Tampa Bay estuary by increasing exchange with the Gulf of Mexico, theref ore resulting in a decrease of

PAGE 14

2 residence time over the periods of the storms. In 2004, Tampa Bay was affected by three hurricanes, approximately a week apart, in Septembe r and an extratropical storm on December 26. The three extreme events chosen for t his study are: Hurricane Frances, Hurricane Jeanne, and the extreme winter frontal pa ssage in December. Hurricane Ivan was not included in this study because there were n o current observations during this time. Extreme events are defined as storms with wi nd speeds that are two to three standard deviations above the yearly mean (4.062 m s-1 in 2004). These events are typically 24-48 hours in duration. The specifics o f each extreme event will be discussed in later sections. Hurricanes are known to impact oceanic systems. Va lle-Levinson et al. (2002) studied the response of the lower Chesapeake Bay to Hurricane Floyd. They used water density and velocity data from a 70 day period depl oyment across the entrance to Chesapeake Bay in conjunction with wind velocity an d sea level records. The forcing associated with Floyd consisted of northeasterly wi nds prior to the passage of the storm eye, which caused a net inflow over the shallow nor thern half of the bay entrance and outflow in the deep channel to the south of the ent rance. After the passage of the eye the winds shifted rapidly to the northwest and peaked a t 27 m s-1. The change of the winds coincided with a pulse of freshwater that caused sa linity to drop throughout the water column of up to 8 units in 1 day. Wind and river d ischarge set up a seaward barotropic pressure gradient force that drove net outflow ever ywhere across the entrance to the bay allowing no inflow and effectively flushing water o ut. It was estimated that approximately one-third of the net outflow was caus ed by wind forcing and two-thirds by freshwater discharge. Walker (2001) studied chang es in circulation, water level, salinity,

PAGE 15

3 suspended sediments and sediment flux resulting fro m Tropical Storm Frances and Hurricane Georges in the Vermiliona-Atchafalaya Bay during September 1998. Tropical Storm Frances made landfall approximately 400 km we st of the study area. The strong and long-lived southeasterly winds resulted in the highest water levels and salinity values of the year. Water levels were high across the coa stal bay system, while salinity impacts varied spatially. Hurricane George made landfall 2 40 km east of the study area, winds were from the north on the west side of the storm a nd reached a maximum of 14 m s-1. During a period of strong wind stress, coastal wate r levels fell, and salinity decreased. As the winds subsided, a pulse of relatively saline wa ter entered Vemilion Bay increased salinity from 5 to 20 psu over a 24 hour period. T he results demonstrated that remote storm systems can have substantial impacts on the p hysical process that control circulation. Forristall (1980) used wind-driven current measure ments off the coast of Louisiana made during Hurricanes Carmen and Eloise in 1975 and incorporated them into a numerical model. Near-surface waters on the continental shelf off Louisiana are usually strongly stratified by river runoff in the summer. The passage of hurricanes provides the energy to mix the surface layer down t o a depth between 30 and 45 m. During this time a two-layer current system develop s, with the mixed layer responding more directly to the wind shear than the bottom lay er. The two-layer system was modeled by parameterizing the mixed layer with a re latively high eddy viscosity and the lower layer with a much lower eddy viscosity. Comp arisons of the model and the storm measurements showed that the model was reasonably a ccurate.

PAGE 16

4 Li et al. (2006) showed how Hurricane Isabel affect ed Chesapeake Bay. They used a numerical model prediction and real-time obs ervations to show a slab-like sloshing in Chesapeake Bay. The high winds of Hurr icane Isabel forced the entire water column up the Bay at speeds in excess of 1.5 m s-1. After the storm passed, the Bay relaxed with a rapid movement of the entire column in the opposite direction. The strong mixing caused by the winds removed the stratificati on in the water column and caused the bay to temporarily transform from a partially m ixed estuary into a vertically homogeneous one. Weisberg and Zheng (2006a) used a finite volume coa stal ocean model with flooding and drying capabilities to investigate the storm surge responses for Tampa Bay as well as to simulate Hurricane Charley in the Cha rlotte Harbor vicinity (2006b). The model-simulated surge agreed well with the observat ions at four stations for which data existed which allowed them to use the model to expl ain the surge evolution and to account for the inlet breach that occurred at North Captiva Island. They found that even though Charley was a category 4 hurricane, the surg e associated with it was only of nominal magnitude and the damage was primarily wind -induced. They explained the relatively small storm surge on the basis of the di rection and speed of approach, point of landfall, propagation up the bay axis, and the coll apse of the eye radius as the storm came ashore. Wilson et al. (2006) studied the changes in circula tion in Tampa Bay due to Hurricane Frances. They used ECOM-3D (discussed l ater in the background) and observation data to simulate how the hydrodynamics of Tampa Bay and the exchange of water with the Gulf of Mexico was altered during th e passing of Hurricane Frances.

PAGE 17

5 They found that the high winds associated with Hurr icane Frances directly affected the circulation of the bay by increasing wind stress, a nd these changes in circulation produced a significant flushing of Tampa Bay. As H urricane Frances approached it displaced a total volume into the bay of 1.5 billio n m3 and -895 million m3 out of the bay as it left the area. Ross (1973) and Goodwin (1980, 1987, and 1989) were the first to do numerical modeling of the Tampa Bay circulation with two-dime nsional, vertically-averaged studies. They assumed that the baroclinic circulat ion may be neglected since Tampa Bay is well-mixed. The models where able to achieve re asonable results with sea level observations and tidal currents, but they were not able to address fluxes over time scales larger than the tidal cycle since they omitted the mean circulation by gravitational convection. The first attempt at a three-dimension al model for Tampa Bay was done by Galperin et al. (1991) who used the Princeton Ocean Model of Blumberg and Mellor (1987). In these simulations, forcing by rivers, t ides, and winds were considered. They demonstrated that the baroclinicity related to the horizontal salinity gradient is enough to drive a non-tidal circulation by gravitational conv ection. They were able to illustrate that baroclinicity is important in Tampa Bay model studi es by comparing the barotropic and baroclinic runs using the same model. They also fo und that the salinity fields changed between the barotropic and baroclinic runs, proving that it is gravitational convection that controls the distribution of salinity in Tampa Bay (Weisberg and Zheng, 2006b). Vincent et al. (1999) integrated two real-time oceanographi c data acquisition networks, the Physical Oceanographic Real Time System (PORTS) and the Coastal Ocean Monitoring and Prediction System (COMPS), into the Blumberg-Me llor ECOM-3D model of Tampa

PAGE 18

6 Bay. These systems were used to drive a three dime nsional hydrodynamic circulation model of Tampa Bay in a real-time nowcast-forecast mode. In this study the instantaneous and residual (will be used interchangeably with non-tidal), meaning demeaned and de-tided, circulat ion in Tampa Bay are examined using a realistic numerical model of the estuarine circulation in Tampa Bay and hourly velocity in the shipping channel measured by an ADC P at the Sunshine Skyway Bridge. The model is used to simulate the flow field and ti dal residual circulation for the year of 2004 (See Meyers et al. (2007) for details). Insta ntaneous circulation in Tampa Bay is obtained from hourly model fields of archived model output. Once the tides were removed using a least square analysis, calculations of the residual circulation fields at every model grid cell is performed. Total volume c hanges are calculated to yield the volume of water that was being flushed in and out o f the bay. The non-tidal maximum in/outflow for each extreme event was studied and c ompared to discuss the effects of each extreme event on the circulation of Tampa Bay. A Lagrangian method was used to examine the flushin g of the bay during September of 2002 and 2004 as well as December of 2 002 and 2004. It is based on particle tracking, where neutrally buoyant dimensio nless particles are advected by the model velocity field (Burwell, 2001). Particle ret ention will be examined for the months of September and December of 2004 and compared to t he same months from 2002. 2002 was chosen for the comparison because there were no extreme events that occurred during the months of September and December during the year. Particle tracking was used to simulate the routes most likely taken by di fferent water parcels during each

PAGE 19

7 extreme event. These simulations were helpful in viewing how much new/old water entered/exited the bay during the three extreme eve nts. The high winds that occurred during these extreme events altered the circulation of the bay. Additionally, the residual circulation is dominated by the horizontal overturning of the bay waters driven by density dif ferences between the mouth and the head of the bay created by the inflow of freshwater The wind driven response is consistent throughout the water column, therefore, the focus will be on the near surface flow which will be used as a representative for the water column. The high winds cause surface waters to flow in and out of the bay at fas ter velocities therefore altering the circulation of the bay. The focus of this study i s to use a model to simulate changes in the residual circulation caused by extreme events.

PAGE 20

8 Chapter Two Background Tampa Bay Tampa Bay is located on the central part of the wes t coast of Florida. It is the largest estuary and port in Florida and is also the seventh largest U.S. commercial port in terms of tonnage handled (Estevez et al,. 1985). I t is a significant marine resource for the State of Florida and provides major ports of commer ce, supports a variety of fisheries and offers important recreational opportunities for Flo rida’s residents and visitors (Weisberg and Williams, 1992). It also accommodates the comm unity needs of power generation, fresh water consumption and sanitation requirements (Weisberg and Zheng, 2006b). Tampa Bay begins at the Gulf of Mexico near 82.50 W and 27.60 N, and extends in a northeast direction approximately 53 km (Figure 1). The bay has natural channels that follow the main core of the Y shaped estuary with t ypical depths of up to 10 meters. The estuary has two branches and lower and middle stem segments that are referred to as Old Tampa Bay, Hillsborough Bay, Lower and Middle Tampa Bay (Lewis and Whitman, 1985). Tampa Bay covers approximately one thousand square kilometers and has an average depth of approximately four meters (Goodwin 1987). The width of the bay is about 15 km at its midsection. Dredged navigation channels lead to many of the main port facilities. The depths of the channels have i ncreased to 15 meters to maintain the minimum depth required for shipping to occur in the bay (Zervas, 1993). The bay has a

PAGE 21

9 Figure 1. Map of Tampa Bay.

PAGE 22

10 maximum depth of about 27 meters in Egmont Channel near the mouth of the bay (Figure 2). Tampa Bay has a circulation that is 3-dimensio nal and time dependent. Tides, winds, and rivers all have a significant effect on the cir culation (Galperin et al., 1991, Weisberg and Zheng, 2006b). Tampa Bay sea level and current variations are cont rolled by the tides. Tides in the bay consist of mixed semidiurnal and diurnal ti des, with a tide range of less than a meter at the mouth to over a meter at the head. Th e tides experience two high and two low unequal tides during a period of one day. The tidal wave entering the bay can be characterized as a progressive wave, which transiti ons into a standing wave in Hillsborough Bay. Tidal epochs indicate that the t ide travels from the mouth of the bay to the head of Old Tampa Bay and Hillsborough Bay i n approximately 4.6 hours and 3.2 hours (Zervas, 1993). Ocean water from the Gulf of Mexico enters through the mouth located in the southwest portion of Tampa Bay. With the amount of freshwater inflow into the bay, its shallow depths, and strong tidal mixing, the salini ty is well mixed vertically. Although the bay is vertically well mixed, it does have sign ificant horizontal salinity gradients due to the distribution of fresh water inflow. These h orizontal gradients and surface wind forcing maintain the fully three-dimensional circul ation of the bay (Li, 1993). Freshwater from the north together with saltwater coming from the south produce a strong horizontal salinity gradient (Wilson et al., 2006). Estuaries that are vertically homogenous but have strong horizontal salinity gradients have been show n to have baroclinic density driven flows, which contribute to the residual circulation (Pritchard, 1956).

PAGE 23

11 Figure 2. Model bathymetry map of Tampa Bay.

PAGE 24

12 The bay has major sources of fresh water that are l ocated primarily on the east and south sides. The Alafia and Hillsborough river s drain into the bay from the northeast, near the head of the bay. The Little Ma natee enters on the eastern side and the Manatee on the south near the mouth of the bay. Us ing flow rates from table 4 in Meyers et al. (2007), these four rivers account for 27% of the average total freshwater input to the bay. The salinity of the bay is regulated by the fresh w ater sources and the Gulf of Mexico water at the open boundary. Evaporation als o plays a role in removing water from the bay, causing the salinity of the bay to in crease. Salinities in the bay vary from a high of approximately 35 at the entrance of the bay to a low of 20 ppt or less in the northern and eastern parts of Hillsborough Bay and the northwest part of Old Tampa Bay (Boler, 1992). Salinities in the bay are lowest in the summer, when the freshwater inflow is the greatest, and highest in the winter. The ve rtical stratification is strongest in Hillsborough Bay (Zervas, 1993). The average total freshwater into Hillsborough Bay is 610 29.3 m3 d-1 (Meyers et al., 2007). The lower salinities at th e head of the bay and the higher salinities at the mouth of the bay cause an axial pressure gradient force to exist that drives a non-tidal, gravitational convection m ode of circulation, known as estuarine circulation (Weisberg and Zheng, 2006b). The resi dual circulation speed can vary by a factor of 3 and alter in direction (Meyers et al., 2007). Water temperatures throughout the bay range from 11 .7C in the winter to 32.8C in the summer (Boler, 1992). At the Sunshine Skyway, tidal currents are nearly u niformed with depth, and have peak amplitudes ranging from 50 cm s-1 during a neap cycle and 100 cm s-1 during a

PAGE 25

13 spring cycle (Li, 1993). The tidal currents have a lso been observed to have maximum speeds on the order of 1.0 to 1.5 m s-1 in the Egmont Channel and the channel leading to Old Tampa Bay (Vincent, 2001). Burwell (2001) found the residual circulation in Ta mpa Bay appears to be a mix of classical two layer flow over the shipping chann els, with denser ocean water flowing in at depth, and fresher water flowing out of the b ay near the surface and along the relatively shallow sides of the bay. Flushing of t he bay occurs through the deep navigational channels running northeast/southwest f rom the mouth. Residence times in this area were found to be short on the order of 15 days to one month and increase up to over three months in regions outside the channels n ear the edges of the bay and in persistent eddies. ECOM-3D The Estuarine and Coastal Ocean Model (ECOM-3D) use d here was developed by Blumberg and Mellor (1987) and Blumberg (1990). It is a version of the Princeton Ocean Model (POM) that has been modified and adapte d for Tampa Bay applications by Galperin et al. (1992) and Vincent et al. (1997). ECOM-3D is a three-dimensional numerical model for the near-shore environment (Bur well, 2001). It is a non-linear finite difference model that uses a curvilinear-orthogonal grid in the horizontal and a bottom following sigma coordinate system in the vertical ( Burwell et al., 1999). ECOM-3D solves the three-dimensional time dependent equatio ns for conservation of mass, momentum, heat, and salinity. It uses a split time step for the solution of the baroclinic 3D mode and the barotropic 2-D mode and has an embed ded second moment Mellor-

PAGE 26

14 Yamada turbulence closure model to provide vertical mixing coefficients. This mode splitting separates the external fast surface gravi ty wave calculation from the internal wave calculation for computational economy (Blumber g and Mellor, 1987). The horizontal diffusion is provided by the Smagorinsky (1963) formulation. The grid for Tampa Bay consists of 70 by 100 cells in the horizo ntal and 11 layers in the vertical and has a minimum depth of 1.3 m MLLW. Open boundary conditions at the mouth are provided by the measured temperature, salinity (provided by the Environmenta l Protection Commission of Hillsborough County), and sea surface elevation col lected by instrumented towers at the mouth of Tampa Bay on Egmont Key and Anna Maria Isl and. The model has free surface boundary conditions which include wind stre ss and mass flux. The mass flux and salinity are also included for rivers discharges th at are located around the border of the bay (Vincent et al., 1997). Vincent (2001) compare d model water levels, salinities, and currents to observed data. He found that the model does an excellent job of reproducing water levels all around the bay with the mean absol ute error being less than 0.025 meters. The mean error (a measure of the overall bias of a prediction) for salinity at all levels was less than or equal to 0.23 ppt and the model result s and salinity data had a positive correlation of 0.93. Specific details and the meth ods and analyses of these results can be found in Vincent, 2001. Meyers et al. (2007) upda ted the model to years beginning in 2001. Model output for 2004 is used in this study. The main limitation of the model relevant to the e xtreme events (defined when winds are three standard deviations above the mean) is the lack of wetting-drying capabilities. The strong winds during these extrem e events can cause the model water

PAGE 27

15 depth to go negative in shallow areas of the bay. To solve this problem the northwest lobe (Old Tampa Bay) of Tampa Bay was given a minim um bathymetric depth of 1.8 m MLLW for all three extreme events. During Hurrican e Jeanne the winds were also limited to 20 m s-1. Hurricane Frances Hurricane Frances developed from a tropical wave th at moved westward from the coast of Africa on August 21, 2004. On August 26th it turned west-northwestward and continued in intensification until August 28th when it reached a wind speed of 115 knots (59 m s-1) and became a category 4 hurricane. The hurricane turned westward on August 29th and tropical storm warnings and hurricane watches were posted for the northeastern Caribbean. The hurricane weakened due to upper lev el shear, but remained a category 3 hurricane through the 30th. On August 30th Frances began to re-intensify and the winds attained 125 knots (64 m s-1) on August 31st and regained category 4 status. At this time the National Hurricane Center was predicting that t he hurricane would make landfall near Vero Beach over the Labor Day weekend. The hurrica ne moved northwestward on September 1-2 with maximum winds remaining at 120-1 25 knots (62-64 m s-1). During the next two days, Frances weakened due to a modera te westerly vertical shear that developed later on September 2nd. It became a category 3 hurricane with winds of 1 00110 knots (51-57 m s-1) over the central Bahamas on September 2-3, and a category 2 hurricane with winds of 85-90 knots (44-46 m s-1) over northwestern Bahamas on September 3-4. Wind gusts of hurricane strength we re felt in the Bahamas for more than 12 hours. Dry air wrapped into the system which cr eated an extremely large eye almost

PAGE 28

16 70 miles across. On September 5th Frances made landfall near Vero Beach as a categor y 2 hurricane. It continued to move slowly west nort hwestward across central Florida near Bartow by 2 pm EDT and northeast of Tampa by 8 pm E DT, when it was downgraded to a tropical storm. The storm entered the Gulf of Me xico before midnight and continued to move northwestward. It made landfall for a second time on September 6th in the Florida Big Bend near Tallahassee as a tropical storm (Beve n, 2004; Alshiemer, 2004). Its full path can be seen in figure 3. Frances caused an enormous amount of flooding, inc luding freshwater overland and rivers and tidal storm surge. Wind damage was confined to numerous downed large limbs and power lines in West Central Florida. Mos t wind damage was noted during feeder bands during the afternoon and evening of Se ptember 4th. Some bands produced wind gusts of 61 knots (31 m s-1) near the Pinellas and Pasco county shoreline, oth ers produced small tornadoes in Polk County. Once Frances entered the Gulf of Mexico near New P ort Richey on September 5th, winds shifted to the southwest and blew at a stea dy pace of 22-30 knots (11-15 m s-1) with frequent gusts of 35-43 knots (18-22 m s-1). The winds lasted approximately 12 hours causing above normal water levels along south -facing shorelines and within Tampa Bay. Surge values of 1.22 to 1.83 meters were repo rted from Cedar Key to Ozello and 0.61 to 1.22 meters farther south from the Pinellas coast and Tampa and St. Petersburg shorelines through Hernando County. On September 6th the daily flow rates according to USGS at the Alaf ia, Hillsborough, Little Manatee, and Manatee rivers we re 133.37 m3 s-1, 45.31 m3 s-1, 18.04 m3 s-1, and 64.85 m3 s-1, all large increases from the average 2001-2003 mo del flows

PAGE 29

17 Figure 3. Complete path of Hurricane Frances.

PAGE 30

18 (Alafia: 12.76 m3 s-1, Hillsborough: 8.89 m3 s-1, Little Manatee: 7.95 m3 s-1, and Manatee: 0.21 m3 s-1) according to Meyers (2007). As of September 9th almost all rivers in west central and southwest Florida were flooded. Seven waterways reached major flood status; three were moderate with one nearing major status, and two were minor but had the potential to become moderate status (SRH, 2004a ). Rainfall ranged from 5 to 10 cm over southwest Flor ida, 10 to 15 cm in Manatee and southern Pinellas County, 15 to 20 cm in Hillsb orough, Hardee and western Polk counties, and 20 to 30 cm in Pasco County. The hea vy rains and poor drainage caused flooding in several low lying areas around and nort h of Tampa Bay (SRH, 2004a). Hurricane Jeanne The path taken by Hurricane Jeanne was similar to t he path of Hurricane Frances and was the fourth hurricane to hit Florida during the 2004 hurricane season. On September 7th Jeanne formed from a tropical wave that moved from Africa to the eastern tropical Atlantic Ocean. A tropical depression for med from the wave on September 13th as it approached the Leeward Islands. From Septemb er 13th through the 18th, Jeanne moved slowly west-northwest with speeds of 5-10 kno ts (2.5-5 m s-1). It strengthened to a tropical storm on September 14th while it moved over the Leeward Islands. On September 15th it slowly moved over the Virgin Islands and the ce nter moved inland over southeastern Puerto Rico with maximum surface winds reaching 60 knots (30 m s-1). As Jeanne passed through the Mona Passage and made lan dfall on the Dominican Republic it became a hurricane with winds of 70 knots (35 m s-1). The slow movement of Jeanne

PAGE 31

19 across the Caribbean caused extensive amounts of ra infall which resulted in freshwater flooding and mudslides killing thousands in Haiti. While Jeanne was causing lots of rainfall over the Caribbean, Hurricane Ivan moved over the Gulf of Mexico and inland on the sou theastern United States. Jeanne took a northerly course across the Turks and Caicos Islands on September 18th then high pressure at mid levels strengthened to the north ca using Jeanne to make a slow clockwise loop and regain hurricane strength. By September 2 3rd Jeanne completed the loop and had strengthened to a hurricane with winds up to 85 knots (40 m s-1). On the 24th Jeanne moved over its own previous track and encountered c ooler waters caused by upwelling from the hurricane, causing its winds to decrease f rom 85 knots (40 m s-1) to 70 knots (33 m s-1). As the hurricane moved westward away from the u pwelled water, the winds increased to 100 knots (51 m s-1) as the center moved over Abaco Island and then Gr and Bahama Island. On September 26th Jeanne made landfall on the east coast of Florida with the center of the eye crossing the coast at th e southern end of Hutchinson Island. Maximum winds at landfall were estimated at 105 kno ts (54 m s-1). As Jeanne moved across central Florida it weakened to a tropical st orm while north of Tampa and then weakened to a tropical depression about 24 hours la ter while moving across central Georgia (Lawrence and Cobb, 2005; Paxton, 2004). H urricane Jeanne’s full path can be seen in figure 4. The primary difference between the paths of Hurrica ne Frances and Jeanne was that Jeanne moved at a steady 10 knots (5 m s-1) across central Florida, limiting the bad weather to 24 hours.

PAGE 32

20 Figure 4. Complete path of Hurricane Jeanne.

PAGE 33

21 The strong winds caused minor to moderate damage to structures that were poorly constructed and mobile/manufactured homes. Similar damage to Frances were downed tree limbs and power lines, uprooted shallow-roote d trees, and damage to residences and businesses, which included lost roof shingles, stri pped siding, and scattered damage to carports and roofs within mobile home parks. In we st central and southwest Florida, gusts were at their highest in Polk and Highlands C ounties, but also along the Gulf coast from Pinellas to Citrus County. Gusts were recorde d in excess of hurricane force in Polk County, eastern Hillsborough County, and along the Pinellas and Pasco County coastline. USGS daily flow rates for September 27th were 95.71 m3 s-1 for the Alafia river, 22.17 m3 s-1 for the Hillsborough river, 11.81 m3 s-1 for the Little Manatee river, and 45.31 m3 s-1 for the Manatee river. These daily flow rates wer e not as high as those seen in Frances, but are still much larger than the aver age 2001-2003 model flows. Coastal flooding was less of a problem with Jeanne than with Frances. The duration of the southwest winds produced more wides pread problems from the Tampa Bay area northward, which induced significant beach erosion on the Suncoast’s barrier islands. Minor flooding was seen at Cedar Key wher e tides switched from more than 1.22 meters below normal to more than 0.91 meters a bove normal hours after the winds had shifted, and along the northern Pasco County co ast, where water levels were up to 3 feet above normal as well (SRH, 2004b). December 26, 2004 Extratropical Storm The night after Christmas, Tampa Bay residents were reminded of Hurricane Frances and Jeanne. The same storm system that hit South Texas twenty-four hours

PAGE 34

22 earlier made its way towards Tampa Bay. Temperatur es around the area dropped tremendously. In Levy County temperatures decrease d to the 40’s (Farenheight, 4 C) on Christmas Eve and for most of Christmas day, in Tam pa it was in the low 50s (10 C). Christmas day consisted of steady rains, with heavi er bands producing 5 to 10 cm of rain. Gale force winds with gusts above storm force (48 knots, ~24 m s-1) were observed in the eastern Gulf by mid evening on Chri stmas night 2004. It was reported that surface pressure was just above 1000 mb, while pressures at nearby buoys were substantially higher. By midnight, the chilly weat her gave way to steadily rising temperatures. Between 07:00 and 08:00 UTC winds fr om the south rose to 30 knots (15 m s-1) near the coast and tides built up to 0.61 m above normal. Soon after came the low center of the storm, which crossed from St. Petersb urg through south Tampa downtown, then towards eastern Hillsborough County. Winds ca lmed for a brief moment in a small area near the center of the storm just before burst ing from the west with values between 26-39 knots (13-20 m s-1) and gusts as high as 62 knots (32 m s-1). The gusty winds caused water to slam along the coa st between Manatee and Pasco Counties. There was overwash, minor coastal flooding, and surges of 0.91 to 1.83 meters. The storm surge was minor due to the tides going out of the bay during this time. Most of the damage included power outages, roofs bl own off of mobile homes, boats were shoved off their moorings, marinas sustained s ome damage along the Manatee County shoreline, and trees were snapped and few we re uprooted. Most of the damage occurred near shorelines or in areas exposed to hig her wind gusts (SRH, 2004c).

PAGE 35

23 Chapter Three Data Collection and Methods Data Collection Data for Tampa Bay is gathered by two systems: Tam pa Bay Physical Oceanographic Real Time System (PORTS) and the Coas tal Ocean Monitoring and Predication System (COMPS). PORTS was installed between 1990 and 1992 by the N ational Ocean Service (NOS) and has been operational since June of 1992. It has been housed and maintained by the University of South Florida (USF) since 1993 through cooperative agreement with the NOAA National Ocean Service. PORTS is a real-t ime data acquisition and dissemination system that consists of tide stations with anemometers (wind sensors) distributed around the bay and current meter statio ns along the main ship channel. The data are archived by USF and NOS and are used here for evaluation of model fields and as meteorological inputs to the model (Burwell et a l., 1999). Detailed information on the Tampa Bay PORTS system can be found on the NOAA/NOS /CO-OPS website at http://tidesandcurrents.noaa.gov/tbports/tbports.sh tml?port=tb. PORTS consists of three Acoustic Doppler Current Pr ofilers (ADCPs), four water level gages, eight wind sensors, an atmospheric temperatu re and barometric pressure sensor, packet radio transmission equipment, a data acquisi tion system, and an information

PAGE 36

24 dissemination system (Figure 5). The data are upda ted every six minutes by line-of-sight radio telemetry. COMPS is also housed and maintained at USF and cons ists of several near realtime stations with various meteorological and ocean ographic sensors along the West Florida coast (Figure 6). It provides additional d ata needed for a variety of management issues and consists of an array of instrumentation both along the coast and offshore, combined with numerical circulation models. Instru mentation consists of several buoys with current meters and meteorological packages tha t are located on the West Florida shelf offshore and coastal towers with water level, temperature and salinity, and meteorological sensors located near shore along the west coast of Florida. Two COMPS coastal stations are located near the mouth of Tamp a Bay and provide real-time input for the open boundary to the Hindcast model that runs a t USF. They provide the open boundary conditions for all simulations and test ru ns of ECOM-3D. The Tampa Bay COMPS stations report to USF via radio every 6 minu tes (Burwell, 2001). All data is automatically transferred to the National Data Buoy Center, where they are put into the NOAA operational data stream with all the standard NOAA quality assurance/quality control procedures. The ADCP used in this study is the only one of thre e ADCPs that is currently active and used in PORTS. This ADCP was the only o ne to have data during all three extreme events. The ADCP located at Old Port Tampa recorded currents during Hurricanes Frances and Jeanne but was not working d uring the December extratropical storm. The ADCP is a Teledyne/RD Instruments Work Horse 1200 kHz system. It is located in the main ship channel beneath the center span of the Sunshine Skyway Bridge

PAGE 37

25 Figure 5. Map of Tampa Bay PORTS stations.

PAGE 38

26 Figure 6. Map of COMPS stations.

PAGE 39

27 at 27 37.22’ N and 82 39.35’ W. The instrument sits in approximately 17 m of water and has 18 bins (12 clean bins), each approximately 1 m in height. It has a blanking distance of about 44 cm above the 52 cm high instru ment, making the first bin about 1 m off the bottom of the bay. The data are telemetere d to USF every six minutes and are continuously quality controlled by watch standers a t the NOAA National Ocean Service through dedicated network connections. All the dat a are made available in real time to the maritime community and to the general public at http://tidesandcurrents.noaa.gov/tbports/tbports.sh tml?port=tb. The axial (along the main axis of the bay) current is calculated by projectin g the measured velocity vector onto the local angle of the ship channel. The local angle o f the bay at the site of measurement is 62 degrees from true north. The ADCP was not opera tional from September 7 through September 24 and November 8 through December 2. To accomplish the overall baseline circulation of T ampa Bay for the years 20012004, the model had to be updated through 2004. Da ta used for the boundary conditions of the model for the year 2004 were gathered from s ome of the PORTS/COMPS sites and others from different groups and services described below. Daily precipitation rates from three different stations (Tampa, St. Petersburg, an d Sarasota airports) for 2004 are obtained from the National Weather Service website. Data from each site were obtained, compared, and then daily averages were computed to get a uniform daily precipitation value for the model. Monthly measured salinity dat a from site 93 (just south of Egmont Key) were collected and provided by the Environment al Protection Commission of Hillsborough County (See Boler, 1992, for a discuss ion). Sea surface elevation was obtained from Egmont Key and Anna Maria Island at t he mouth of Tampa Bay and from

PAGE 40

28 St. Petersburg from PORTS stations EGK, ANM, and 87 26520. Winds from the Cut-C Lower Rear Range Marker (CCUT) instrument in the ce nter of the bay (PORTS station m01010) were used uniformly across the model domain The use of a single wind vector from the CCUT tower was justified by comparing wind speed and direction from five National Weather Service sites around Tampa Bay. T he five sites included Albert Whitted Airport, Clearwater International Airport, Tampa Bay International Airport MacDill Airforce Base, and Sarasota/Bradenton Airpo rt. The comparison shows that for the months of September and December 2004 (Figure 7 ) the timing and direction of the winds at all the sites are alike, even during the e xtreme events when wind speed and directions changed drastically over short periods o f time. The wind speeds from the airport locations are slightly smaller than those o bserved from the CCUT site. This is possibly due to the CCUT monitoring site being in t he middle of the bay where there are no buildings or other objects to cause obstruction and differences in surface roughness between land and water. Daily stream flows from USGS are used to provide ri ver inflow. Interpolation of nearby rivers is done to estimate the stream flow o f the un-gauged rivers. For the rivers that have gauges the data is downloaded and is eith er interpolated to fill in small gaps (less than 2 days) or a ratio of surrounding rivers is used to fill in larger gaps (greater than 2 days). Daily discharge from 4 wastewater treatme nt plants in Tampa Bay are also included for a total of 36 freshwater point sources (Figure 8). Locations of each model freshwater point source are shown on table 1.

PAGE 41

29 Figure 7. CCUT wind speeds and directions versus f ive airport sites around Tampa Bay.

PAGE 42

30 Figure 8. Map of model freshwater point sources (i n red).

PAGE 43

31 Table 1. Locations for each model freshwater point source (Meyers et al., 2007). Grid Location Name Flow Area Source Gauge I J (m3s-1) (sq mi) 59 65 Delaney Creek 0.39 23.7 Delaney 59 59 Archie Creek 0.27 16.4 Delaney 61 53 Bullfrog Creek 2.11 40.3 Bullfrog 18 87 Rocky Creek 1.62 35 Rocky 4 81 Allen Creek 0.2 8.4 Sweetwater 8 92 Lake Tarpon 1.41 60.3 Sweetwater 27 81 Sweetwater Creek 0.87 37.3 Sweetwater 8 86 Alligator Creek 0.21 8.9 Sweetwater 43 79 Hillsborough River 8.89 650 Hillsborough 65 56 Alafia River 12.76 418 Alafia 57 39 Little Manatee 7.95 222 Ltl Manatee 44 78 Sulphur Springs 0.76 0 Sulphur Spg 51 72 Ybor City Drain 0.16 9.7 Delaney 5 78 Long Branch 0.22 9.4 Sweetwater 13 90 Double Branch 0.71 30.6 Sweetwater 9 92 Safety Harbor Drain 0.19 7.9 Sweetwater 9 87 Mullet Creek 0.07 3 Sweetwater 31 59 Peninsula 1 0.31 18.8 Delaney 42 61 Peninsula 2 0.31 18.8 Delaney 8 73 Mangrove Bay 0.34 14.7 Sweetwater 16 44 North Bayous 0.58 24.8 Sweetwater 59 48 Big Bend Bayou 0.17 10.1 Delaney 15 36 South Bayous 0.61 26.1 Sweetwater 57 45 Apollo Beach Canal 0.09 5.6 Delaney 54 43 Wolf Branch 0.65 12.5 Bullfrog 47 35 Cockroach Bay 0.78 14.8 Bullfrog 54 26 Terra Ceia Bay 2.62 49.9 Bullfrog 66 24 Ward Lake Outfall 3.66 59.5 Ward Lake 66 26 Lake Manatee Discharge 0.21 0 Lake Manatee 65 75 Tampa Bypass Canal 4.1 0 65 75 Falkenberg Recovery 0.25 0 52 28 Piney Point 0.02 0 52 65 Curren Waste Water TP 2.4 0 6 82 Clearwater WWTP 0.24 0 5 79 Largo Waste Water TP 0.33 0 15 38 River Oaks WWTP 0.22 0

PAGE 44

32 Methods Each event was divided into phases to simplify and organize the analysis. The phases were chosen to capture times of minimums and maximums of elevation during each event. The duration, maximum and minimum wind speed, mean wind direction, and total volume flow of each phase for each event is s hown in table 2. The echo amplitudes for each beam and bin of the AD CP were averaged and plotted as a function of range (Figure 9), which is approximately 1 m for each bin, to find which bins contain good data. As the range increas es the echo amplitude decreases due to the reflecting particles getting farther away fr om the ADCP. At a range between 12 and 14 m the echo amplitude increases rapidly, whic h is caused by the sea-surface reflection due to the discontinuity of the air-sea interface. This increase in echo amplitude indicates reverberation from the sea surf ace and renders the last 6 bins unusable; therefore, only the first 12 bins will be used in this study. Bin 12 will be used to represent near-surface currents at the site; how ever, the ADCP sits in 17 m of water in the shipping channel just north of the Sunshine Sky way Bridge. Evaluation of the model was done a few different wa ys. Due to the model grid being closely aligned to the local axis of the bay, only the axial velocity variable was examined. Velocities for the near surface, mid-dep th, and bottom of the bay from the ACDP and model are compared. ADCP bins 2, 8, and 12 were used. Bin 2 is approximately 2 m off the bottom of the bay, bin 8 starts at about 8 m off the bottom, and bin 12 is near the surface approximately 12 m off t he bottom (at relative depths 0.88, 0.53, and 0.29, respectively). The model depth at the grid cell corresponding to the ADCP site is 10.36 m deep, creating possible mismat ches when comparing ADCP data

PAGE 45

33 Table 2. Time duration, maximum and minimum wind speeds, mea n wind direction, and total volume flow for each phase during all three extreme events.

PAGE 46

34 Figure 9. ADCP echo amplitudes as a function of ra nge for all four beams.

PAGE 47

35 and model results. For the model levels 5, 6, and 8 were used (at relative depths 0.25, 0.5, and 0.875, respectively). A cubic spline interpolation was performed on the A DCP data to get the observed data and model results at grid point (35, 25) on th e same hourly time step. Once the ADCP data were interpolated, the correlation was ca lculated between the instantaneous ADCP data and the model results. The correlation wa s computed from July 1 through September 7 because this time frame had the longest record of continuous ADCP data. For the instantaneous model and ADCP data, r2=0.94 at all depths and for the residual model and ADCP data, r2=0.64. Model velocity from grid point (35, 25) (this grid point represents the location of the ADCP) was used to find the best method to isola te the extreme events. The model velocity is de-meaned, a least square analysis is d one to get the tidal components, and then a low-pass filter is applied. For the least s quare analysis, the mean and eight of the main tidal components of Tampa Bay (M2, K1, O1, S2, P1, N2, Q1, and K2, Table 3) were fitted and subtracted from the data for the se cond half of 2004 (July-December). A 25-hr low-pass filter is selected to remove residua l tidal signals. This method was used on the velocity and elevation model output, however the magnitude of the extreme events decreased significantly after the 25-hr filt er was applied (Figure 10). Since the extreme events did not last longer than 48 hours, a pplying a 25-hr filter caused the magnitudes of the extreme events to be smoothed and decreased dramatically. Therefore a low-pass filter is not used after de-tiding. The mean and tides are subtracted from the model variables at each grid cell to obtain the res idual values throughout Tampa Bay.

PAGE 48

36 Table 3. Tidal constituents gathered from www.tide sandcurrents.noaa.gov, St. Petersburg station 8726520, used in least square an alysis. Constituents Amplitude (m) Epoch () Period (hr) M2 0.175 197 12.4206 S2 0.057 211.7 12.0000 N2 0.03 191.3 12.6583 K1 0.167 49.9 23.9345 O1 0.155 37.7 25.8193 Q1 0.029 26.2 26.8684 P1 0.049 57.6 24.0659 K2 0.025 215 11.9672

PAGE 49

37 Figure 10. Model velocity and elevation (black) ov erplotted with the non-tidal component (blue) and a 25-hr low pass filter (red).

PAGE 50

38 The root mean square (RMS) velocity was calculated at every grid point by squaring the model velocities at every point in tim e. Once the velocities were squared they were averaged over the specific time period fo r each event (Table 2) at every grid point and then the square root of the average was t aken. An overall RMS average for each layer in the model was calculated by averaging the RMS values at every grid cell in a specific layer. The same procedure was used to c alculate RMS averages for the monthly averages of September/December 2001-2003, t he tidal component (predicted flow from the eight main tidal components during th e time period of each event), the total component (instantaneous velocities for each event) and for each phase of the events. Phases for each event are defined as periods with c omplete inflow or outflow. The residual axial currents from the model were used to find the maximum inflows and outflows for each extreme event from hourly time se ries plots. Particle tracking was done using Burwell’s (2001) L agrangian method, where he developed the algorithms used in the model to advec t passive particles according to the three-dimensional model velocity field. The code r ecords the grid cell containing each particle, the relative position of each particle wi thin the cell, and the total number of particles in each cell at any given time. It also takes into account the staggered grid and uses the nearest horizontal neighbors for each velo city component to linearly interpolate to the position of each particle. In this study pa rticle tracking was done by filling every grid cell with 20 particles. The model was run for a few days surrounding each extreme event. Each particle has a specific identification number and its path was tracked through time. Once the particles leave the bay, they are n ot allowed to re-enter. The paths of specific particles initialized at different areas i n the bay indicate water exchange that

PAGE 51

39 occurred during these extreme events. Particle coun ts (total number of particles in each cell through time) for the months of September and December of the years 2002 and 2004 were calculated using the model as well. Sept ember and December of 2002 were used as a comparison for September and December of 2004.

PAGE 52

40 Chapter Four Results Hurricane Frances There were three phases in the passage of Frances n ear Tampa Bay. The first phase began at midnight on September 5, 2004, 18 ho urs before the first wind peak as Frances approached Tampa bay with winds to the S an d SE (Figure 11). During this phase sea level remained below MSL. The second pha se began on September 5 at 19:00 UTC as the eye passed nearby and the winds turned t oward the E and NE, and the second wind peak occurred. During this phase the residual currents were strongly positive raising sea level to 1.2 m above MSL at St. Petersb urg. In the third phase (beginning September 6 at 11:00), after Frances left the area, the positive residual flow into the bay was replaced by a strong negative (seaward) flow as the wind and sea level relaxed toward normal. The following results for Hurricane Frances were f irst discussed by Wilson et al. (2006). Model velocity and the ADCP data for the d ays surrounding Frances correspond well to one another (Figure 12). During phase 1 th e near surface instantaneous axial velocity in the model and ADCP shows an inflow and outflow most likely due to the tides. During phase 2 instantaneous near surface v elocity measured by the ADCP reaches a high of about +0.7 m s-1 and +0.9 m s-1 in the model. The winds decrease over the next

PAGE 53

41 Figure 11. Meteorological data for the month of Se ptember 2004: wind speed (top panel), wind direction (middle panel), and water le vel (bottom panel).

PAGE 54

42 Figure 12. Instantaneous ADCP velocities (dotted l ine) versus model axial current (solid line) during Hurricane Frances. Yearly mean for ea ch level (thin solid line) and (dashed line) are shown as well.

PAGE 55

43 24 hours to 5-10 m s-1 in phase 3 and the near surface axial ADCP velocit ies reverse to a maximum speed near -0.7 m s-1 out of the bay and -0.9 m s-1 for the model. The mid-depth axial velocity during phase 1 and 2 g oes from -0.4 m s-1 and then peaks at +0.6 m s-1 for the ADCP and +0.8 m s-1 for the model. During phase 3 as the winds relax the velocity reverses and reaches a max imum of about -0.65 m s-1 for the ADCP and -0.7 m s-1 in the model. The bottom velocities show the same behavior, however, the magnitude of the velocity decreases wi th depth. The inflow ranges from 0.3 m s-1 during phase 1 to +0.5 m s-1 during phase 2 and then reverses to -0.54 m s-1 during phase 3 in the ADCP. The model shows similar behavior at all three depths. In phase 1, the near surface residual current is we ak in the ADCP and the model (Figure 13), but the model shows significant outflo w during the first wind peak. As the wind turns in phase 2, the residual near surface cu rrent peaks over +0.5 m s-1 in the ADCP and +0.7 m s-1 in the model. This is rapidly followed by the rev ersal to -0.8 m s-1 in phase 3. Both mid-depth and bottom residual currents show a similar reversal, though the magnitudes decrease with depth. The mid-depth resi dual velocity in phases 1 and 2 goes from about -0.2 m s-1 in the ADCP and -0.3 m s-1 in the model, and then peaks at +0.45 m s-1 for the ADCP and +0.6 m s-1 for the model as the wind turns to the NE. During this time the bottom velocity goes from about -0.18 m s-1 to a peak of about +0.45 m s-1 for the ADCP and the model. The bottom also shows a la rge reversal to -0.5 m s-1 in both the ADCP and the model. When Frances passed through the bay area, it was in phase with the tides (Figure 14). During phase 1 the total model velocity was i n the same direction as the tidal

PAGE 56

44 Figure 13. Residual ADCP velocities (dotted line) overplotted with model residual axial current (solid line) during Hurricane Frances. Yea rly mean for each level (thin solid line) and (dashed line) are shown as well.

PAGE 57

45 Figure 14. Model velocity (black) overplotted with tidal fit (blue) and the difference between the model and tidal fit (red).

PAGE 58

46 component, leading to greatly increased volume tran sport into the bay. During phase 2 the model velocity has the same shape as the tidal component, however, the magnitude of the velocity has doubled due to the increased winds blowing in the NE direction. At the beginning of phase 3 as the winds began to decrease the model velocity and tidal velocity began to flow out of the bay. The magnitude of the model velocity is still greater than the tidal velocity due to the excess amount of water th at was flushed into the bay during phase 2 is flowing out of the bay. The velocity in tidal component, which consists of the mean and eight main tidal components of Tampa Bay during the three phases of Frances, is weaker than the model velocity from September 2001-2003 (Figure 15). This is possibly due to Frances comin g in during a neap cycle causing the mean of the tida l component to be lower than the averaged September velocities which include a coupl e of spring cycles. The tidal component does not include winds which are included in September 2001-2003 which could affect the mean as well. During phase 2 as t he inflow from Frances came into the bay the RMS velocity of the model doubled compared to the average of September 20012003. At the surface, RMS velocity increased from 0.22 m s-1 to 0.48 m s-1, at mid-depth they went from 0.12 m s-1 to 0.24 m s-1, and at the bottom RMS velocities went from 0.07 m s-1 to 0.15 m s-1. During phase 3 the outflow RMS velocities were n ot as strong as in phase 2, but were still stronger than the September 2001-2003 and baseline velocities. The RMS velocities during phase 3 were 0.28 m s-1 at the surface, 0.17 m s-1 at mid-depth and 0.1 m s-1 at the bottom. Comparisons between the surface, mid-depth, and bot tom of the bay during the time of maximum inflow during Hurricane Frances are shown in figures 16, 17, and 18.

PAGE 59

47 Figure 15. Vertical RMS velocity for all model lay ers during Hurricane Frances. Vertical RMS velocity was calculated for the 3 days during Hurricane Frances (red), the tidal component of the same 3 days (blue), for aver aged Septembers 2001-2003 (green), phase 3 (orange), phase 2 (yellow), and phase 1 (br own).

PAGE 60

48 Figure 16. Maximum surface inflow during phase 2 o f Hurricane Frances. The tidal component is show o n the left, the total component is in the middle, and the differenc e between total and tidal components is on the righ t.

PAGE 61

49 Figure 17. Maximum mid-depth inflow during phase 2 of Hurricane Frances. The tidal component is show on the left, the total component is in the middle, and the diffe rence between total and tidal components is on the right.

PAGE 62

50 Figure 18. Maximum bottom inflow during phase 2 of Hurricane Frances. The tidal component is show on the left, the total component is in the middle, and the differenc e between total and tidal components is on the righ t.

PAGE 63

51 The tidal flow at the surface and mid-depth show hi gher velocities (30 cm s-1) in through Lower and Middle Tampa Bay. At the surface the hig h velocities are near the mouth of Tampa Bay due to water flowing through the Egmont C hannel, the narrow opening between Fort DeSoto and Egmont Key. In the total f low, current speeds reached a maximum of 120 cm s-1 throughout Middle Tampa Bay. The strongest curren t speed values are seen at the middle of the main channel o f the bay and at the mouth of Old Tampa Bay (~120 cm s-1). Velocities increased from 15 cm s-1 (in the tidal flow) to approximately 60 cm s-1 (in the total flow) in the middle and upper areas of Old Tampa Bay and Hillsborough Bay. The same is seen in the plots at mid-depth (Figure 17) and bottom (Figure 18). In the middle of Old Tampa Bay the tidal current velocities were between 30-45 cm s-1 at mid-depth and between 0-30 cm s-1 at the bottom through most of the bay. The non-tidal (total-tidal) plot shows the effects of mostly the winds and freshwater inflow due to the tides being removed fr om the model output. The inflow of water was throughout the entire width of the bay an d not just through the main ship channel as is seen in the tidal flow. The surface current velocities are a bit weaker in the non-tidal flow than in the total flow, however, the winds still caused surface currents to reach a maximum of about 105 cm s-1 throughout Middle Tampa Bay. The same thing is seen in the mid-depth and bottom plots. During phase 3 all the water that had flowed into t he bay started to flush out and sea levels began to reach normal levels. The tidal (Figure 19) outflow at the surface shows that the water escapes the bay at faster velo cities (~ 15-30 cm s-1) through the main ship channel and the eastern side of the mouth of O ld Tampa Bay. In the total and

PAGE 64

52 Figure 19. Maximum surface outflow during phase 3 of Hurricane Frances. The tidal component is show on the left, the total component is in the middle, and the diffe rence between total and tidal components is on the right.

PAGE 65

53 difference plots the higher velocities occur in the se same areas, with the ship channel being the area of maximum velocities ranging from 6 0 cm s-1 in Middle Tampa Bay to 105 cm s-1 at the mouth of the bay for the total flow and 45 cm s-1 to 75 cm s-1 in the difference plot. The same patterns are visible in the mid-depth and bottom plots (Figure 20 and 21). The lowest elevation (relative to mean sea level) seen in the entire bay at the end of phase 2 is 1.1 m at the mouth of the bay (Figure 22) Elevations rose to 1.8 and 2.0 m in the northern parts of Hillsborough Bay and Old T ampa Bay. At the end of phase 3 (Figure 23) when the winds and elevation relaxed, m ost of the bay was level with an elevation of 0.5 m (with the exception of the east coast of Middle Tampa Bay which was at an elevation of 0.4 m) The high winds during the second phase of Hurricane Frances caused water from the Gulf of Mexico to enter the bay and increased s alinities. Five grid points throughout the bay were chosen (Figure 24) to compare saliniti es in different areas of the bay. The 2004 mean salinities for each of these locations ar e: 32.32 at the mouth of the bay (MT), 31.34 at the Sunshine Skyway Bridge (SS), 29.39 in Middle Tampa Bay (MC), 28.92 in Old Tampa Bay (OTB), and 27.21 in Hillsborough Bay (HB). Salinities reach a high of about 30 ppt at the points near the mouth of the ba y, the Sunshine Skyway Bridge, and Middle Tampa Bay. Salinities in Old Tampa Bay and Hillsborough Bay increased slightly but did not reach the high values seen at the other three locations (Figure 25) Salinities decreased at the Sunshine Skyway Bridge, Middle Tampa Bay, and in Old Tampa Bay as phase 3 began and the winds began to r elax. The higher salinity water that was pushed into the bay began to flow out. Towards the end of phase 3, Hillsborough

PAGE 66

54 .Figure 20. Maximum mid-depth outflow during phase 3 of Hurricane Frances. The tidal component is sh ow on the left, the total component is in the middle, and the differenc e between total and tidal components is on the righ t.

PAGE 67

55 Figure 21. Maximum bottom outflow during phase 3 o f Hurricane Frances. The tidal component is show o n the left, the total component is in the middle, and the diffe rence between total and tidal components is on the right.

PAGE 68

56 Figure 22. Baywide elevation at the end of phase 2 during Hurr icane Frances.

PAGE 69

57 Figure 23. Baywide elevation at the end of phase 3 during Hurr icane Frances.

PAGE 70

58 Figure 24. Grid points used to for salinity data at different locations in the bay.

PAGE 71

59 Figure 25. Surface model salinity at different locations in t he bay. The 2004 mean for each location is shown in the thin black line, 1 are shown by the dashed lines, and the vertical lines indicate the beginning and end of ea ch phase.

PAGE 72

60 Bay had the lowest salinity of about 11 ppt due to the freshwater coming in from the rivers. Surface salinities for the entire bay at the end of phase 2 ranged from between 28 and 30 ppt at the mouth to 22 and 24 ppt at the nor thern parts of Old Tampa Bay and Hillsborough Bay (Figure 26) Water with salinities of up to 32 ppt made it up through most of Middle Tampa Bay but not entirely across th e width of the bay. Salinities in Old Tampa Bay and Hillsborough Bay got up to 26 ppt but were never as high as Middle Tampa Bay. During phase 3 salinities in Middle Tam pa Bay lowered to between 24 and 28 ppt and 26 to 28 ppt across the Sunshine Skyway Bridge (Figure 27) Salinities in Old Tampa Bay ranged between 22 and 26 ppt and Hillsbor ough Bay ranged from 4 ppt in the north to 20 ppt at the mouth. Cross sections at four different areas in the bay ( Sunshine Skyway, Middle Tampa Bay, mouth of Hillsborough Bay, and mouth of Old Ta mpa Bay) were plotted to view the salinity throughout the water column at the end of phase 2 and 3. At the end of phase 2 the high salinities of the Gulf of Mexico flushed t hrough the entire water column with salinities of 32 ppt across the Sunshine Skyway (Fi gure 28). In the middle of Tampa Bay the highest salinities (between 28 and 32 ppt) were found in the center of the cross section between the two ship channels throughout th e entire water column. Salinities on the east coast of Middle Tampa Bay were between 24 and 26 ppt, and between 20 and 26 on the west coast. Salinities across the mouth of Hillsborough Bay ranged from 18 ppt on the west coast to 24 on the east coast. The ent ire water column across the mouth of Old Tampa Bay had a salinity of 24 ppt. At the end of phase 3 (Figure 29), horizontal stratification is seen across the Sunshine Skyway w ith more saline waters (30 ppt) at the bottom and less saline (24-28 ppt) at the surface. Across the Middle Tampa Bay the

PAGE 73

61 Figure 26. Baywide view of surface salinity at the end of phase 2 during Hurricane Frances.

PAGE 74

62 Figure 27. Baywide view of surface salinity at the end of phas e 3 during Hurricane Frances.

PAGE 75

63 Figure 28. Salinity cross sections at the end of p hase 2 during Hurricane Frances. Cross sections are from the Sunshine Skyway, Middle Tampa Bay, the mouth of Hillsborough, and the mouth of Old Tampa Bay.

PAGE 76

64 Figure 29. Salinity cross sections at the end of p hase 3 during Hurricane Frances.

PAGE 77

65 salinity ranges from 24 to 26 ppt across the two sh ip channels, and fresher water (12-22 ppt) is seen on the east coast of Middle Tampa Bay due to the outflow from the Little Manatee River. Horizontal stratification is also s een across the mouth of Hillsborough Bay with salinities off 22 ppt at the bottom and 18 -20 ppt at the surface. Salinities across the mouth of Old Tampa Bay remained at 24 ppt throu ghout the entire water column. The high winds caused water to flush into the bay t hrough the entire water column and not just near the surface. The same five point s that were used to view the salinity (Figure 24) were used to view the velocities throug hout the water column at the time where model velocities reached their maximum point of inflow and outflow. At the time of maximum inflow, velocities were greatest at the Sunshine Skyway, the mouth of Tampa Bay, and in the main channel. The same patte rn is seen in all five locations; velocities are highest near the surface due to the wind stress acting on the surface waters and lowest at the bottom due to bottom friction (Fi gure 30). At the Sunshine Skyway the depth average of the vertical profile of horizontal velocity is 0.863 m s-1, the depth varying is 0.444 m s-1, and the ratio between the depth average and depth varying is 0.51. The values for the depth average, depth varying, an d the ratio of the two for all five locations in the bay are shown in table 4. The ver tical velocity at the point of maximum outflow shows a decrease of velocities as the water was beginning to flush out of the bay (Figure 31). Velocities at the point maximum outfl ow were greatest at the Sunshine Skyway and the main channel. At the Sunshine Skywa y the depth average is -0.804 m s1, the depth varying is -0.123 m s-1, and the ratio is 0.152 (Table 4). Particle trajectories were done to help characteriz e anomalous transport of water parcels during each extreme event. Particles were released at four different locations in

PAGE 78

66 Figure 30. Vertical velocity profiles (thick solid line) at five locations in the bay at the time of maximum inflow during Hurricane Frances. Depth average is shown by the thin solid line.

PAGE 79

67 Table 4. Depth average, depth varying, and ration at five different locations throughout the bay at times of maximum inflow (top) and maximu m outflow (bottom) during Hurricane Frances. Inflow Location (Grid Cell) Model Depth Depth Average Depth Varying Ratio Mouth of Tampa Bay (37,17) 5.46 0.645 0.416 0.644 Sunshine Skyway (35,25) 10.36 0.863 0.444 0.515 Central Tampa Bay (34,41) 5.3 0.415 0.375 0.903 Hillsborough Bay (49,57) 4.01 0.126 0.368 2.907 Old Tampa Bay (20,76) 4.66 0.2 0.31 1.547 Outflow Location Model Depth Depth Average Depth Varying Ratio Mouth of Tampa Bay (37,17) 5.46 -0.233 0.092 -0.396 Sunshine Skyway (35,25) 10.36 -0.805 -0.123 0.152 Central Tampa Bay(34,41) 5.3 -0.362 0.092 -0.254 Hillsborough Bay (49,57) 4.01 -0.195 0.133 -0.682 Old Tampa Bay (20,76) 4.66 -0.094 0.081 -0.857

PAGE 80

68 Figure 31. Vertical velocity profile (thick solid line) at the time of maximum outflow during Hurricane Frances. The depth average is sho wn by the thin solid line.

PAGE 81

69 the bay throughout the entire water column; across the mouth of the bay, in the upper main channel, and the middle of Hillsborough Bay an d Old Tampa Bay. Particles were tracked during the phases of inflow and outflow. T he particles released across the mouth of the bay during phase 2 entered Middle Tampa Bay on the east side and only traveled half way up Middle Tampa Bay (Figure 32). Particles released in the northern ar ea of the main channel split. Some traveled into Hillsboroug h Bay and some into Old Tampa Bay. Those released in Hillsborough Bay made their way t o the north of the bay. Some of the particles that were released in Old Tampa Bay made their way to the north of the bay and some got blocked by the Courtney Campbell Causeway. During phase 3 (Figure 33) all the particles released in Hillsborough Bay, Old Tam pa Bay and Middle Tampa Bay began to flush out towards the mouth of the bay. The high winds and fresh water input that occur dur ing Frances directly impact the circulation of the bay. The changes in circula tion produce a significant flushing of Tampa Bay. The model elevation has higher maximum and minimum values than the observed data (most likely due to the boundary cond itions and model resolution), which will cause higher volumes to be calculated using mo del elevation versus observed elevation. Using model elevation data at grid point (15, 37) n ear the St. Petersburg tide station 8726520, inflow and outflow volumes were ca lculated as well as the total volume (sum of the volumes at every grid point) for the en tire bay. The difference between the minimum (-0.687 m below MSL) and maximum (+1.572 m above MSL) points during phase 2 at grid point (15, 37) was 2.25 m (Figure 3 4). Multiplying by the model bay area (810 54.9 m2) gives a volume increase of 910 15.2 m3. For the same calculations

PAGE 82

70 Figure 32. Particles released during phase 2. Par ticles were released throughout the water column at the mouth of Tampa Bay, Middle Tamp a Bay, and in the center of Hillsborough Bay and Old Tampa Bay.

PAGE 83

71 Figure 33. Particles released during phase 3.

PAGE 84

72 Figure 34. Observed water level from the St. Peter sburg station (blue line) and Port Manatee station (red line) compared to model data a t St. Petersburg (black line) and Port Manatee (green line). The top panel represents wate r levels during Hurricane Frances, the middle panel represents water levels during Hurrica ne Jeanne, and the bottom panel represents the Extratropical Storm. Observed water levels are from www.tideandcurrents.noaa.gov.

PAGE 85

73 during phase 3, the difference in elevation is 1.09 m yielding a volume decrease of 910 04.1 m3. The total bay inflow and outflow volume was calc ulated by summing the volumes at every grid point for both the tidal and total components. For the total component, the difference between the minimum (910 91.2 m3) and maximum (910 94.4 m3) bay volumes gives a volume inflow of 910 03.2 m3 (Figure 35). Dividing the volume change by the mean model volume (910 6.3 m3) results in an increase of 56% in bay volume. The volume differen ce during the time of outflow is 910 02.1 m3, a 28% decrease. The difference of the volume inflow/outflow calcula ted using elevation at a single grid point versus the total volume inflow/outflow o f the entire bay is small: 910 12.0 m3 for the volume inflow and 910 02.0 m3 for the volume outflow. The volume differences of the tidal component were calculated as well to g et an idea of the volume change during this time without an extreme event present. During phase 2 the tidal volume change was 810 5.4 m3 and 810 6.4 m3 during phase 3. The differences between the total and tidal bay volume are 910 58.1 m3 during phase 2 and 810 6.5 m3 during phase 3. Using observed data from the water level observati ons at the St. Petersburg gauge, there is a change in water level from -0.51 m at 17:42 UTC on 9/5/04 to a high of 1.17 m at 14:48 on 9/6/04 leading to a water level change of 1.68 m. This value is then multiplied by the surface area of the bay, 910 031 .1 m2 (Zervas, 1993), and leads to a volume change of 1.73 billion m3. Dividing the volume change by the mean volume of the bay, 910 81.3 m3 (Zervas, 1993), results in an increase of 45% in b ay volume. The same calculations are done to find the decrease in bay volume. Water level reached a

PAGE 86

74 Figure 35. Total model bay volume during Hurricane Frances (top panel), Hurricane Jeanne (middle panel) and the Extratropical Storm ( bottom panel). The thick solid black line represents the total component of the extreme events and the blue line is the tidal component.

PAGE 87

75 minimum of 0.360 m at 2:36 UTC on 9/7/04, leading t o a decrease in bay volume of 836 million m3, or a 22% decrease. The high winds and wave actio n during this time leas to significant vertical mixing. Hurricane Jeanne Hurricane Jeanne passed through the Tampa Bay area in just over 24 hours. Hurricane Jeanne consisted of only two phases and o ne wind peak. The first phase began on September 26 at 11:00 and lasted about 12 hours. During this time the winds peaked at 25 m s-1 and were in the S and SE direction. The residual c urrents were positive during this phase, causing sea level to rise to 0.8 m abov e MSL at St. Petersburg. In the second phase, which began an hour before midnight on Septe mber 26, the winds lowered to 8 m s-1 and stayed relatively constant in the Eastward dir ection as Jeanne left the bay area. The currents turned from a positive to negative res idual flow and sea level lowered to 0.2 m (Figure 10). During phase 1 the near surface instantaneous veloc ity measured by the ADCP reaches a high of about +0.6 m s-1 and +0.8 m s-1 in the model. (Figure 36). As the winds decreased during phase 2, the near surface axial ve locities reversed out of the bay to maximum speeds of -0.5 m s-1 for the ADCP and -0.7 m s-1 in the model. The mid-depth axial velocity during phase 1 reaches a high of +0.6 m s-1 for both the ADCP and the model. During phase 2 the currents are going out of the bay and the velocity decreases to -0.5 m s-1 in the ADCP and -0.4 m s-1 in the model. The bottom axial velocities show similar behavior with smaller magnitudes in both the ADCP and model.

PAGE 88

76 Figure 36. Instantaneous ADCP (dotted line) veloci ties versus model axial current (solid line) during Hurricane Jeanne. Yearly mean for eac h level (thin solid line) and (dashed line) are shown as well.

PAGE 89

77 The near surface residual current during phase 1 wa s relatively weak in the ADCP compared to the model (Figure 37). The ADCP record ed a maximum inflow +0.5 m s-1, whereas the model shows a maximum inflow at +1.1 m s-1. During phase 2, there is a maximum outflow of -0.7 m s-1 for the ADCP and -0.9 m s-1 in the model. At mid-depth and bottom the residual currents show the same reve rsal, however, the magnitudes decrease with depth. The mid-depth residual velocity in phase 1 reaches a maximum inflow of +0.4 m s-1 in the ADCP and +0.9 m s-1 for the model. The outflow during phase 2 was -0. 8 m s-1 in both the ADCP and the model. For the bottom vel ocity, during phase 1, the ADCP peaked at +0.3 m s-1and +0.6 m s-1for the model. The outflow during phase 2 reached -0.55 m s-1 for both the ADCP and the model. When Jeanne passed through Tampa Bay it was in phas e with the tides. As Jeanne passed the predicted elevations show two hig hs and two lows from the beginning of phase 1 to the end of phase 2, but the model vel ocities show only one high and one low (Figure 38) during this time. During phase 1 as the tides begin to flow out of the bay, the model velocities were slowing down but were still p ositive and flowing into the bay. In phase 2, the predicted tides would have reversed tw ice and began to flow into the bay and then out half-way threw the phase they started to f low out again. At the end of phase 2, the tides and model velocity both have the point of maximum outflow at approximately the same time. The RMS velocities for the tidal component and Sept ember 2001-2003 show similar patterns and are very close in magnitude (F igure 39). During phase 1 as the inflow from Jeanne came into the bay the RMS veloci ty was 0.42 m s-1 at the surface,

PAGE 90

78 Figure 37. Residual ADCP (dotted line) overplotted with model residual axial current (solid line) during Hurricane Jeanne. Yearly mean for each level (thin solid line) and (dashed line) are shown as well.

PAGE 91

79 Figure 38. Model velocity (black) overplotted with tidal fit (blue) and the difference between model and tidal (red).

PAGE 92

80 Figure 39. Vertical RMS velocity for all model lay ers during Hurricane Jeanne. Vertical RMS velocity was calculated for the 3 days during H urricane Jeanne (red), the tidal component of the same 3 days (blue), for averaged S eptembers 2001-2003 (green), phase 2 (orange), and phase 1 (yellow).

PAGE 93

81 0.25 m s-1 at mid-depth, and 0.11 m s-1 at the bottom. Phase 2 does not have the high velocities seen in phase 1, but shows similar RMS v elocities as the tidal flow and September 2001-2003, 0.23 m s-1 at the surface, 0.14 m s-1 at mid-depth, and 0.06 m s-1 at the bottom. The tidal flow during phase 1 at all depths shows t he tides coming in through the center of the main channel with velocities between 30-60 cm s-1 at the surface and middepth, and between 0-30 cm s-1 at the bottom (Figures 40, 41, and 42). The high velocities seen at the surface near the mouth of th e bay during Hurricane Frances are seen during Hurricane Jeanne as well. The total model f low shows the maximum inflow current velocities making their way into the bay (N E) through the main shipping channel with velocities between 60-90 cm s-1 at the surface and mid-depth, and between 15-45 cm s-1at the bottom. High velocities are also seen at th e mouths of Old Tampa Bay (3075 cm s-1) and Hillsborough Bay (30-60 cm s-1 ) at the surface. The difference between the total model and the tida l velocities shows the surface current velocities are a bit weaker and reach a max imum of 60 cm s-1 throughout most of the bay. Not only do the velocities get weaker but the direction of the flow changes to the southeastward. Velocities weakened to 30 cm s-1 at mid-depth and 15 cm s-1 at the bottom. As Hurricane Jeanne passed Tampa Bay the inflow tha t occurred in phase 1 was replaced with an outflow during phase 2 (Figure 43, 44, and 45). During phase 2 at the point of maximum outflow the tidal flow shows that the tides were flowing out of the bay between 15-30 cm s-1 at the surface. The total flow for Jeanne at the surface and middepth show the water following the path of the shi p channel with velocities of ~60 cm s-1

PAGE 94

82 Figure 40. Maximum surface inflow during phase 1 o f Hurricane Jeanne. The tidal component is show on the left, the total component is in the middle, and the differenc e between total and tidal components is on the righ t.

PAGE 95

83 Figure 41. Maximum mid-depth inflow during phase 1 of Hurricane Jeanne. The tidal component is show on the left, the total component is in the middle, and the differenc e between total and tidal components is on the righ t.

PAGE 96

84 Figure 42. Maximum bottom inflow during phase 1 of Hurricane Jeanne. The tidal component is show on the left, the total component is in the middle, and the differenc e between total and tidal components is on the righ t.

PAGE 97

85 Figure 43. Maximum surface outflow during phase 2 of Hurricane Jeanne. The tidal component is show o n the left, the total component is in the middle, and the differenc e between total and tidal components is on the righ t.

PAGE 98

86 Figure 44. Maximum mid-depth outflow during phase 2 of Hurricane Jeanne. The tidal component is show on the left, the total component is in the middle, and the difference between total and tidal components is o n the right.

PAGE 99

87 Figure 45. Maximum bottom outflow during phase 2 o f Hurricane Jeanne. The tidal component is show on the left, the total component is in the middle, and the differenc e between total and tidal components is on the righ t.

PAGE 100

88 as it flows out of the bay. Bay wide model velocit ies ranged from 15 cm s-1 to 60 cm s-1 at the surface and mid-depth, and 15 cm s-1 at the bottom. The difference between the total model and tidal flows shows most of the water moving eastward. The surface outflow velocities range from 0-15 cm s-1 in the main ship channel and 30-60 cm s-1 on the east and west coast of Middle Tampa Bay. Veloc ities range between 0-30 cm s-1 at mid-depth and 0-15 cm s-1 at the bottom. At the end of phase 1 the direction of the winds c aused water to be pushed up into the bay in the eastward direction. The lowest elev ations are seen at the mouth of the bay with a value of 0.5 m and in the northern and weste rn areas of Old Tampa Bay with values of 0.2 m and 0.3 m (Figure 46). The highest elevations are in Hillsborough Bay with values ranging from 0.9 m on the west coast to 1.1 m on the east coast. As the winds subsided at the end of phase 2 (Figure 47) el evation throughout the bay began to lower to values between 0.1 m and 0.5 m. In Old Ta mpa Bay values lowered between 0.1-0.3 m, 0.2 m in Hillsborough Bay, and 0.5 m at the mouth of the bay. The Gulf of Mexico water that came in during Hurric ane Jeanne increased salinities at the mouth of the bay and the Sunshine Skyway, however t hese high salinities are not seen in the northern part of Middle Tampa Bay (Figure 48). At the mouth salinity increased to about 29 ppt and 30 ppt at the Sunshine Skyway duri ng phase 1. In Middle Tampa Bay salinity increased to about 24 ppt. As the current s turned and started to flow out of the bay during phase 2 salinities lowered in Middle Tam pa Bay (22 ppt) and the Sunshine Skyway (25 ppt). Salinities in Hillsborough Bay an d Old Tampa Bay stayed relatively constant throughout phase 1 and 2 at 20 ppt, respec tively. At the end of phase 1, salinities throughout the entire bay ranged from be tween 26 and 28 ppt at the mouth,

PAGE 101

89 Figure 46. Baywide elevation at the end of phase 1 during Hurricane Jeanne.

PAGE 102

90 Figure 47. Baywide elevation at the end of phase 2 during Hurricane Jeanne.

PAGE 103

91 Figure 48. Surface salinity at different locations (Figure 57) in the bay. The 2004 mean for each location: the thin black line, and 1: the dashed lines.

PAGE 104

92 20 and 24 ppt at the top of Middle Tampa Bay, and b etween 18 and 24 ppt in Old Tampa Bay and Hillsborough Bay (Figure 49). At the end o f phase 2 salinities began to lower. In Middle Tampa Bay salinities lowered between 18 a nd 22 ppt. Across the Sunshine Skyway Bridge salinities are between 22 and 26 ppt. Salinities in Old Tampa Bay and Hillsborough Bay did not change (Figure 50). The salinity cross sections for Hurricane Jeanne ar e a bit different than those for Hurricane Frances. At the end of phase 1 the winds were at 20 m s-1 and were pointing eastward. At the Sunshine Skyway, salinities go fr om 28-30 ppt on the left side to 24-26 on the right side (Figure 51). The salinity of the water decreases in Middle Tampa Bay, salinities range from 20-26 ppt. The highest salin ity (~24-26 ppt) is found above the right branch of the main ship channel. Across the mouths of Hillsborough Bay and Old Tampa Bay salinity stayed in the range of 20-24 ppt At the end of phase 2, vertical stratification is seen across the Sunshine Skyway w ith more saline waters (26 -30 ppt) at the bottom and less saline (22-26 ppt) at the surfa ce (Figure 52). Salinities range from 20-24 ppt across the surface and mid-depth in Middl e Tampa Bay and 22-26 ppt near the bottom. Hillsborough Bay and Old Tampa Bay stay bet ween 20-22 ppt throughout the entire water column. The vertical velocity profile at the mouth of Tampa Bay stayed in the range of 0.15 to -0.07 m s-1 throughout the entire water column with the higher negative value at the surface (Figure 53). This shows that at the mouth the flow was out of the bay, even at the time of maximum inflow. The vertical velocity profile at the Sunshine Skyway has the largest values out of all five locations, 0.85 m s-1 at the surface and 0.43 m s-1 near the bottom. Middle Tampa Bay was 0.17 m s-1 at the surface and 0.24 m s-1 near the bottom.

PAGE 105

93 Figure 49. Baywide salinity at the end of phase 1 during Hurricane Jeanne.

PAGE 106

94 Figure 50. Baywide salinity at the end of phase 2 during Hurri cane Jeanne.

PAGE 107

95 Figure 51. Salinity cross sections at the end of p hase 1 during Hurricane Jeanne.

PAGE 108

96 Figure 52. Salinity cross sections at the end of p hase 2 during Hurricane Jeanne.

PAGE 109

97 Figure 53. Vertical velocity profile (thick solid line) at the time of maximum inflow during Hurricane Jeanne. The depth average is show n by the thin solid line.

PAGE 110

98 Hillsborough Bay and Old Tampa Bay had the same ove rall shape throughout the water column and only differed by about 0.1 m s-1 at the surface and less than 0.03 m s-1 at the bottom. During the time of maximum inflow the dept h average of the vertical velocity profile at the Sunshine Skyway is 0.707 m s-1, the depth varying is 0.13, making the ratio between the two 0.18 (Table 5). The vertical veloc ities at the point of maximum outflow show a decrease of velocities as the water was begi nning to flow out of the bay (Figure 54). The mouth is the only location that there is a positive inflow at the surface (0.08 m s-1). At the Sunshine Skyway the velocity at the sur face was -0.6 m s-1 and -0.2 m s-1 near the bottom. At the Sunshine Skyway the depth a verage is -0.49 m s-1, the depth varying is -0.121 m s-1, and the ratio is 0.247 (Table 5). Particle trajectories were done for Hurricane Jeann e the same way they were done for Hurricane Frances. During phase 1 water parcel s made their way up Tampa Bay. The particles released across the mouth of the bay entered the main channel on the east side below the Sunshine Skyway Bridge (Figure 55). Particles released in the northern area of Middle Tampa Bay split towards Hillsborough Bay and Old Tampa Bay but did not make it across the mouths of either bay. The p articles released in Hillsborough Bay branched off into different directions but stayed w ithin the bay. Most of the particles released in Old Tampa Bay did not make it under the Courtney Campbell Causeway. During phase 2 all the particles released in Hillsb orough Bay, Old Tampa Bay and Middle Tampa Bay did not travel far from the point of their original release (Figure 56). During Hurricane Jeanne the difference between the minimum (-0.486 m below MSL) and maximum model elevation (+0.753 m above MS L) during phase 1 was 1.24 m (Figure 34), resulting in a volume inflow of 910 18.1 m3. During phase 2, the difference

PAGE 111

99 Table 5. Depth average, depth varying, and ratio a t five different locations throughout the bay at times of maximum inflow (top) and maximu m outflow (bottom) during Hurricane Jeanne. Inflow Location Model Depth Depth Average Depth Varying Ratio Mouth of Tampa Bay (37,17) 5.46 -0.124 -0.017 0.137 Sunshine Skyway (35,25) 10.36 0.707 0.13 0.183 Central Tampa Bay (34,41) 5.3 0.311 -0.14 -0.45 Hillsborough Bay (49,57) 4.01 0.191 -0.224 -1.172 Old Tampa Bay (20,76) 4.66 0.215 -0.256 -1.19 Outflow Location Model Depth Depth Average Depth Varying Ratio Mouth of Tampa Bay (37,17) 5.46 -0.046 0.123 -2.652 Sunshine Skyway (35,25) 10.36 -0.491 -0.121 0.247 Central Tampa Bay (34,41) 5.3 -0.178 0.057 -0.32 Hillsborough Bay (49,57) 4.01 -0.094 0.081 -0.866 Old Tampa Bay (20,76) 4.66 -0.067 0.018 -0.264

PAGE 112

100 Figure 54. Vertical velocity profile (thick solid line) at the time of maximum outflow during Hurricane Jeanne. The depth average is show n by the thin solid line.

PAGE 113

101 Figure 55. Particles released during phase 1. Par ticles were released throughout the water column at the mouth of Tampa Bay, Middle Tamp a Bay, and in the center of Hillsborough Bay and Old Tampa Bay.

PAGE 114

102 Figure 56. Particles released during phase 2.

PAGE 115

103 in elevation is 0.52 m for a volume outflow of 810 93.4 m3. The difference between the minimum (910 09.3 m3) and maximum (910 21.4 m3) total bay volumes gives a total volume inflow of 910 12.1 m3 (Figure 35), an increase of 31% in bay volume. Th e total volume outflow is 810 0.5 m3, a 14% decrease. The tidal bay volume increased by 810 1.4 m3 during phase 1 and decreased by 810 0.5 m3 during phase 2, making the difference between the total and tidal volume chang es: 810 1.7 m3 for phase 1 and 0 m3 for phase 2. Using the observed water levels from the St. Peters burg gauge, there is a change in water level from -0.315 m at 10:16 UTC on 9/26/0 4 to a high of 0.759 m at 20:18 on 9/26/04 leading to a water level change of 1.074 m. Using the same calculations as in Hurricane Frances leads to a volume change of 910 11.1 m3 which is an increase of 30% in bay volume during phase 1. Water level reached a minimum of 0.158 m at 11:18 UTC on 9/27/04, leading to a decrease in bay volume of 910 620 .0 m3, or a 16% decrease. Extratropical Storm The extratropical storm passed through Tampa Bay i n just a matter of hours. There were two phases in the passage of this winter storm. The first phase began at 3:59 on December 26, 2004 and contained the wind peak (2 4 m s-1) seen during the storm. The first phase only lasted a about 8 hours and dur ing this time the winds started in the NW direction and then quickly reversed to the S and SE direction by the end of the phase. Water levels spiked at 0.8 m above MSL. During pha se 2, which began at 12:00 on

PAGE 116

104 December 26, the winds decreased dramatically to ~1 0 m s-1 and stayed in the SE direction. The water level during this time lower ed to 0.3 m below (Figure 57). The instantaneous axial current during the extratr opical storm for the model and ADCP correspond well to one another (Figure 58). A t the near surface, during phase 1, model velocities peaked at +0.45 m s-1 for both the model and ADCP. As the winds turned and water began to flow out of the bay durin g phase 2, velocities peaked at -0.9 m s-1. At mid-depth velocities go from +0.4 m s-1 in phase 1 to -0.9 m s-1 in phase 2 for the model and ADCP. At the bottom the same pattern is seen but the magnitudes have decreased. The ADCP shows a velocity of +0.3 m s-1 during phase 1 and -0.7 m s-1 during phase 2. The model peaks at +0.25 m s-1 during phase 1 and reaches a minimum of -0.6 m s-1 during phase 2. The residual axial current at all depths shows sim ilar behavior in the ADCP and the model (Figure 59). As the water was flowing in to the bay during phase 1, both the model and ADCP have maximums at +1.3 m s-1. During phase 2 as the water began to flow out of the bay, the model has a maximum outflo w of -1.0 m s-1 and the ADCP has a maximum of -0.9 m s-1. The mid-depth and bottom residual currents show similar behavior as the near surface. At mid-depth the res idual velocity goes from +1.0 m s-1 to 0.9 m s-1 in the model, and from +1.0 m s-1 to -1.0 m s-1 in the ADCP. The bottom residual current during phase 1 is +0.6 m s-1 in the model and +0.9 m s-1for the ADCP. During phase 2, the model peaks at -0.6 m s-1 and the ADCP peaks at -0.8 m s-1. As the Extratropical storm made its way through th e bay area, it was not in phase with the tides (Figure 60). At the beginning of th e phase 1 the astronomical tide was

PAGE 117

105 Figure 57. Meteorological data for the month of De cember 2004. The top panel shows winds speed with yearly mean (thin solid horizontal line) and (dashed line), the middle panel shows wind direction, and the bottom p anel shows water level with mean and .

PAGE 118

106 Figure 58. Instantaneous ADCP (dotted line) veloci ties versus model axial current (solid line) during Hurricane Jeanne. Yearly mean for eac h level (thin solid line) and (dashed line) are shown as well.

PAGE 119

107 Figure 59. Residual ADCP (dotted line) overplotted with model residual axial current (solid line) during the Extratropical Storm. Yearl y mean for each level (thin solid line) and (dashed line) are shown as well.

PAGE 120

108 Figure 60. Model velocity (black) overplotted with tidal fit (blue) and the difference between model and tidal (red).

PAGE 121

109 falling as the storm was forcing water into the bay During phase 2, as the astronomical tide was rising, the observed velocities were still negative and flowing out of the bay. The RMS velocities for the tidal flow and average o f December 2001-2003 are very similar (Figure 61). The tidal flow is about 0.01 m s-1 stronger at the surface and 0.01 m s-1 stronger at the surface but is similar to the aver age of December 2001-2003 throughout the rest of the water column. Phase 1 i s weaker than phase 2 everywhere except at the surface where phase 1 is about 0.01m s-1 faster than phase 2. As the relative depth increases, velocity in phase 1 is stronger th an in both the tidal component and December 2001-2003 average. Between the relative d epths of 0 and -0.25, the RMS velocity of phase 1 decreases from 0.33 m s-1 to 0.17 m s-1 whereas phase 2 only decreases from 0.32 m s-1 to 0.23 m s-1. Towards the bottom, the velocity profiles in all five cases have similar shapes but different magnit udes. The residual circulation during the extratropical storm was different from the Hurricane Frances and Jeanne. The tidal flow at th e surface during the time of maximum inflow shows the astronomical tidal velocity flowin g out of the bay at relatively high speeds (Figure 62). The tidal velocity is flowing out through the main ship channel with speeds up to 105 cm s-1 and between 45-75 cm s-1 through the mouth of Old Tampa Bay. The total flow shows velocities going into the bay in the NE direction in the bottom half of Middle Tampa Bay with speeds between 45 and 90 c m s-1. The velocities in the top half of Middle Tampa Bay are flowing in the eastwar d direction, whereas velocities in the northern parts of Hillsborough Bay and Old Tampa Ba y are flowing in the southeastward direction. The difference plot between the total a nd tidal shows the effects due mostly to the winds. The velocities reach maximums of 120 cm s-1 throughout the entire width of

PAGE 122

110 Figure 61. Vertical RMS velocity for all model lay ers during the Extratropical Storm. Vertical RMS velocity was calculated for the 3 days during the Extratropical Storm (red), the tidal component of the same 3 days (blue), for averaged Septembers 2001-2003 (green), phase 2 (orange), and phase 1 (yellow).

PAGE 123

111 Figure 62. Maximum surface inflow during phase 1 o f the Extratropical Storm. The tidal component is show on the left, the total component is in the middle, and the difference between total and tidal components is o n the right.

PAGE 124

112 Middle Tampa Bay. Velocities are extremely high du e to the winds overcoming the outflow of the tides to force flow into the bay. A t mid-depth the velocities reach 75 cm s1 for the tidal flow and 60 cm s-1 for the total flow. The difference plots at mid-d epth show velocities reaching maximums of 120 cm s-1 at the mouth of the bay and the Sunshine Skyway, and 90 cm s-1 through Middle Tampa Bay (Figure 63). At the bott om, velocities are strongest again in the non-tidal vel ocity reaching values up to 60 cm s-1 in lower Tampa Bay (Figure 64). The tidal flow velocity shows the tid es flowing out of the bay at about 30 cm s-1 and the total flow shows an inflow between 0 to 15 cm s-1. At the point of maximum outflow the tidal flow sho ws the tides flowing in and out at slow speeds of 0-15 cm s-1 at all depths (Figures 65, 66, and 67). In the to tal flow velocities range from 45-105 cm s-1 at the surface, 30-105 cm s-1 at mid-depth, and 0-45 cm s-1 at the bottom. The total outflow at the surface i s strongest (105 cm s-1) at the mouth of the bay. At mid-depth and bottom the maxi mums are seen near the mouth of the bay ranging from 60-105 cm s-1 at mid-depth and 15-60 cm s-1 at the bottom. Velocities are between 60-105 cm s-1 at the mouth at the surface and at mid-depth. At the bottom velocities range from 15-60 cm s-1 at the mouth. At the end of phase 1 of the extratropical storm, elevations were highest on the western side of Middle Tampa Bay reaching values of 1.1 m a bove MSL (Figure 68). At this time the tides were making their way out of the bay while the winds were trying to push water into the bay, causing a pile up of water in t his area. Elevations in Old Tampa Bay and at the mouth of the bay were the lowest at this time with values of 0.5 m above MSL. At the end of phase 2 the entire bay elevation had lowered between 0.4-0.5 m below MSL (Figure 69).

PAGE 125

113 Figure 63. Maximum mid-depth inflow during phase 1 of the Extratropical Storm. The tidal component i s show on the left, the total component is in the middle, and the difference between total and tidal components is o n the right.

PAGE 126

114 Figure 64. Maximum bottom inflow during phase 1 of the Extratropical Storm. The tidal component is s how on the left, the total component is in the middle, and the difference between total and tidal components is o n the right.

PAGE 127

115 F igure 65. Maximum surface outflow during phase 2 o f the Extratropical Storm. The tidal component is show on the left, the total component is in the middle, and the difference between total and tidal components is o n the right.

PAGE 128

116 Figure 66. Maximum mid-depth outflow during phase 2 of the Extratropical Storm. The tidal component is show on the left, the total component is in the middle, and the difference between total and tidal components is on the right.

PAGE 129

117 Figure 67. Maximum bottom outflow during phase 2 o f the Extratropical Storm. The tidal component is show on the left, the total component is in the middle, and the difference between total and tidal components is o n the right.

PAGE 130

118 Figure 68. Baywide elevation at the end of phase 1 of the Extratropical Storm.

PAGE 131

119 Figure 69. Baywide elevation at the end of phase 2

PAGE 132

120 During phase 1 salinity stayed constant at the mout h of Tampa Bay (32 ppt) and Hillsborough Bay (27 ppt, Figure 70). Salinity ros e to 33 ppt at the Sunshine Skyway, 29.5 ppt in Middle Tampa Bay, and 25 ppt in Old Tam pa Bay. During phase 2, salinity stayed constant at the mouth of the bay and Hillsbo rough Bay. At the Sunshine Skyway salinity decreased throughout the duration of phase 2 and had a minimum at 29 ppt. In Middle Tampa Bay salinity decreased to 27 ppt and t o 24 ppt in Old Tampa Bay. Salinities at the end of phase 1 were high through most of the bay. The highest salinities are in lower Tampa Bay with values of 32 ppt (Figur e 71). Salinities between 26 and 32 ppt are seen throughout Middle Tampa Bay. In Old T ampa Bay, salinities range from 28 ppt at the mouth to 20 ppt towards the northern are a of the bay. Salinities didn’t change much at the end of phase 2 and was between 28-32 pp t throughout a large portion of the bay (Figure 72). The salinity cross sections at the end of phase 1 s how that the entire cross-section had a salinity of 32 ppt (Figure 73). Across Middl e Tampa Bay more saline water are found on the east side (~28-30 ppt) of the bay than on the west side (~24 ppt). At the mouth of Hillsborough Bay, most of the cross-sectio n has a salinity of 26 ppt and Old Tampa Bay has a salinity of 28 ppt and a small sect ion on the west side with a salinity of 30 ppt. At the end of phase 2, salinity ranges for all four cross-sections stayed relatively constant (Figure 74). Across the Sunshine Skyway s alinities of 28 ppt are seen at the surface on the west side and 30 ppt everywhere else Across Middle Tampa Bay more saline waters (~28 ppt) are on the east side with l ess saline waters (~26 ppt) on the west side. In Hillsborough Bay the higher salinity (28 ppt) water is located in the middle of

PAGE 133

121 Figure 70. Line plots of surface salinity at different locatio ns in the bay for December 2004. The 2004 mean for each location is shown in the thin black line, and 1 are shown by the dashed lines.

PAGE 134

122 Figure 71. Baywide salinity at the end of phase 1 during the E xtratropical Storm.

PAGE 135

123 Figure 72. Baywide salinity at the end of phase 2 during the E xtratropical Storm.

PAGE 136

124 Figure 73. Salinity cross sections at the end of phase 1 durin g the Extratropical Storm.

PAGE 137

125 Figure 74. Salinity cross sections at the end of phase 2 durin g the Extratropical Storm.

PAGE 138

126 the crosssection with lower saline water (26 ppt) on each side. At the mouth of Old Tampa Bay, the cross-section has a salinity of 24 p pt. The high winds during the extratropical storm did n ot last for a long period of time, causing the winds to affect only a small port ion of the bay. The velocities at the time of maximum inflow show that the winds had the largest effects at the surface. The velocity profiles for Middle Tampa Bay and Sunshine Skyway were positive throughout the entire water column (Figure 75). The highest v elocities are seen at the Sunshine Skyway with velocities around 0.6 m s-1 at the surface and 0.1 m s-1 at the bottom. Both Hillsborough Bay and Old Tampa Bay have positive ve locities at the surface and negative velocities at the bottom. At the Sunshine Skyway the depth average of the velocity is 0.298 m s-1, the depth varying is 0.278 m s-1, and the ratio between the depth average and depth varying is 1.868. These values f or the other four locations can be seen in table 6. At the point of maximum outflow the ve locity profiles are negative throughout the entire water column at every location except in Old Tampa Bay (Figure 76). There is a small positive flow at the bottom of Old Tampa Ba y. In Middle Tampa Bay the velocity profile shows a negative velocity (-0.26 m s-1) at the surface and slightly positive (0.02 m s-1) near the bottom. The strongest outflow velocitie s are seen at the Sunshine Skyway and at the mouth of the bay. At the surface velocities reach a high of -1.1 m s-1 at the Sunshine Skyway and -0.65 m s-1 at the mouth. At the bottom the velocities weaken to -0.5 m s-1 at the Sunshine Skyway and -0.2 m s-1 at the mouth of the bay. The depth average, depth varying, and ratio at the Suns hine Skyway during the time of maximum outflow was -0.848 m s-1, -0.26 m s-1, and 0.306 (Table 6).

PAGE 139

127 Figure 75. Vertical velocity profile (thick solid line) at the time of maximum inflow during the Extratropical Storm. The depth average is shown by the thin solid line.

PAGE 140

128 Table 6. Depth average, depth varying, and ratio a t five different locations throughout the bay at times of maximum inflow (top) and maximu m outflow (bottom) during the Extratropical Storm. Inflow Location Model Depth Depth Average Depth Varying Ratio Mouth of Tampa Bay (37,17) 5.46 -0.309 0.122 -0.395 Sunshine Skyway (35,25) 10.36 0.298 0.278 0.934 Central Tampa Bay (34,41) 5.3 0.12 0.209 1.868 Hillsborough Bay (49,57) 4.01 0.062 0.206 3.314 Old Tampa Bay (20,76) 4.66 0.011 0.158 14.73 Outflow Location Model Depth Depth Average Depth Varying Ratio Mouth of Tampa Bay (37,17) 5.46 -0.395 -0.142 0.36 Sunshine Skyway (35,25) 10.36 -0.848 -0.26 0.306 Central Tampa Bay (34,41) 5.3 -0.345 -0.179 0.51 Hillsborough Bay (49,57) 4.01 -0.075 -0.104 1.4 Old Tampa Bay (20,76) 4.66 -0.095 -0.162 1.71

PAGE 141

129 Figure 76. Vertical velocity profile (thick solid line) at the time of maximum outflow during the Extratropical Storm. The depth average is shown by the thin solid line.

PAGE 142

130 Due to phase 1 only being a few hours long during t he extratropical storm the particles released did not travel far up the bay (F igure 77). The particles released at the mouth made it to the Sunshine Skyway Bridge and the particles released in the main channel, Hillsborough Bay, and Old Tampa Bay stayed relatively close to their initial release. During phase 2, all the particles that we re released started to make their way out of the bay (Figure 78). The particles released in the middle of Old Tampa Bay and Hillsborough Bay made it to the mouths of each bay. Particles released in the main channel made it down to the Sunshine Skyway Bridge. The minimum and maximum model elevations during pha se 1 where -0.105 m and 1.02 m, resulting in a difference of 1.13 m and an inflow volume of 910 07.1 m3 (Figure 34). During phase 2 the elevation differen ce is 1.58 m resulting in a volume outflow of 910 51.1 m3. The difference between the minimum (910 4.3 m3) and maximum (910 25.4 m3) total bay volumes gives a total volume inflow of 810 5.8 m3, an increase of 24% in bay volume (Figure 35). The total volume outflow is 810 27.1 m3, a 35% decrease. The tidal bay volume does the oppo site of the total bay volume, it decreases during phase 1 and increases during phase 2. The tidal volume decreases by 810 9.6 m3 and increases by 810 7.3 m3, making the difference between the total and tidal volume changes: 910 54.1 m3 for phase 1 and 910 64.1 m3 for phase 2. Using the observed data from the St. Petersburg gau ge, there is a change in water level from -0.161 m at 22:60 UTC on 12/25/04 to a high of 0.831 m at 10:30 on 12/25/04 leading to a water level change of 0.992 m. Using the same calculations as before; results in a volume change of 1.02 billion m3 and a 27% increase in bay volume. Water level

PAGE 143

131 Figure 77. Particles released during phase 1. Particles were released throughout the water column at the mouth of Tampa Bay, Middle Tamp a Bay, and in the center of Hillsborough Bay and Old Tampa Bay.

PAGE 144

132 Figure 78. Particles released during phase 2.

PAGE 145

133 reached a minimum of -0.343 m at 16:36 UTC on 12/26 /04, leading to a decrease in bay volume of 1.21 billion m3, or a 32% decrease.

PAGE 146

134 Chapter Five Summary & Discussion The focus of this study was to observe how an estu ary (Tampa Bay) responds to extreme events. The large exchange between an estu ary and coastal ocean during an extreme event can reset the bay by increasing flush ing over short periods of time. The number of hurricanes that have affected the Tampa B ay area has increased within the past few years. During the month of September 2004, Tamp a Bay was affected by three hurricanes within one week from each other. A mode l was used to simulate the changes in residual circulation caused by the extreme event s. Observed velocities and water levels during each ev ent were much different than times when there were no extreme events in the bay area. Each extreme event had its own characteristics, however they each affected the bay in a similar matter by flushing a large volume of water into the bay. One of the mo st important differences between each extreme event is the duration of each event: Hurric ane Frances lasted approximately two and a half days, Hurricane Jeanne was in the bay ar ea for about twenty-four hours, and the extratropical storm came and left within 12 hou rs. The time duration of each event is a large factor on the effects each event had on the bay. Observed wind speeds during Hurricane Frances were six standard deviations higher than the 2004 mean (4.1 m s-1) and water levels at the St. Petersburg station we re four standard deviations higher. Observed axial cu rrents in the ship channel under the

PAGE 147

135 Sunshine Skyway Bridge were two standard deviations above the mean during times of inflow and outflow at all depths. Strong currents in the residual flow associated with Frances occurred at all depth levels and were more than three standard deviations above the normal residual current. The model reproduces these observed changes (there are some differences between ADCP and model velocities, but this might be due to the fact that the ADCP and model depths do not match). Sali nities stayed within one to two standard deviation of the yearly mean throughout mo st of the bay. During Hurricane Jeanne wind speeds were seven sta ndard deviations higher than the mean and water levels were two standard deviati ons above the mean. The axial currents speed associated with Hurricane Jeanne wer e two standard deviations during the inflow and a bit higher than one standard deviation during the outflow. The residual axial currents were within three standard deviation s in the ADCP data and six in the model during the inflow period, and both the ADCP a nd model were within six standard deviations during the outflow period. The mismatch between the observed and model velocity is possibly due to the model winds being n o larger than 20 m s-1 (because of the lack of wetting and drying capabilities of the mode l) during Hurricane Jeanne. As Jeanne came and left the bay area it came in phase with th e tides, however skipped half a tidal cycle that occurred between phase 1 and 2 (Figure 3 8). As Jeanne came through the tides were flowing into the bay and the winds were blowin g in the southward and southeastward direction. Salinities stayed within two standard deviation of the yearly mean in the five locations seen on figure24. During the extratropical storm, wind speeds were s even standard deviations higher than the mean and elevation was almost four standard deviations higher in

PAGE 148

136 December, 2004. The axial current inflow stayed wi thin one standard deviation but the outflow was slightly higher than two standard devia tions of the yearly mean at all depths. The residual axial currents seen in the model and i n the ADCP at all depths were about five standard deviations higher than the mean durin g times of inflow and outflow. Salinities through most of the bay stayed within on e standard deviation of the yearly mean. In all three events winds were observed to reach hi ghs between 23 – 25 m s-1. During Hurricane Frances and Jeanne the winds had t he same overall pattern as the hurricanes passed through the bay. As the hurrican es approached winds were towards the S and SE direction and then the winds shifted to th e northward and northeastward during Frances and eastward during Jeanne. During the ext ratropical storm the winds started in the northward direction, quickly shifted to the sou thward, southeastward direction, and stayed in the southeastward direction during the du ration of the storm. Water levels were higher during Hurricane Frances a nd the extratropical storm. Though higher wind speeds were seen during Hurrican e Jeanne, the duration (~8 hours) of strong winds was not as long as seen during Hurr icane Frances (~16 hours). The extratropical storm was not in phase with the tides so as the tides were flowing out of the bay, the winds were pushing water into the bay caus ing a bulge of water in the middle of the bay with higher water levels in the area near t he St. Petersburg water level station. The residual circulation from the model shows the e ffects of the extreme events on Tampa Bay. The inflow current velocities caused mainly by the winds are lowest during Hurricane Jeanne. The effects of the winds during times of maximum inflow can be seen throughout the entire bay during Hurricane Frances and Jeanne, whereas during

PAGE 149

137 the extratropical storm the wind effects are seen p rimarily in the main channel of the bay. As each event came to an end all the water that was pushed into the bay began to flush out of the bay. The high winds that occur simultan eously during these events directly impact the circulation of the bay by increasing win d and density-driven components of the non-tidal circulation. The changes in circula tion produce a significant flushing of Tampa Bay over time scales of the extreme events. The subsequent duration of the changes of these events were not investigated in th is study The highest ratios between the depth average and de pth varying are seen in Hillsborough Bay and Old Tampa Bay, meaning that th e depth varying component is more important for these areas during these extreme events. For all three extreme events at the time of maximum inflow, Hillsborough Bay and Old Tampa Bay have the highest ratios of all five locations and are all positive. During the time of maximum outflow the highest ratios are still seen in Hillsborough Bay a nd Old Tampa Bay, however, during Hurricane Jeanne the ratio is highest at the mouth of the bay. During Hurricane Frances and Jeanne the ratios are mostly negative during th e time of maximum outflow but positive during the Extratropical Storm. The lowes t ratios are seen mostly at the Sunshine Skyway, possibly due to the winds not high ly affecting the flushing out of the tides. Each storm impacted the bay in each locatio n differently as seen by the ratios of the depth average versus the depth varying in table s 4, 5, and 6. The volume of water that was flushed into the bay d uring times of inflow was different for each event. However the excess amoun t of water from each event was still large enough to cause flushing throughout the entir e bay. Each extreme event was different in its on way but all three still had the same overall result on the bay, causing

PAGE 150

138 changes in the circulation, and mixing and overturn ing of the bay due to wind stress and excess amounts of water entering the bay. Due to l arge volumes of water flushing in and out of the bay during time periods of half a day to two days, we can infer that this would also cause residence times to shorten during these extreme events. The comparison between particle counts for Septembe r 2002 and 2004 shows how three hurricanes during September 2004 effected flushing times in the bay. Towards the end of September 2004 (Figure 79) the particle count goes below the e-folding scale line, whereas during September 2002 it does not. T his shows that the hurricanes that past during September 2004 caused water to flush out of the bay at a faster rate then during September 2002 when no extreme events had occurred. The same thing is seen when comparing December 2002 to December 2004 (Figure 80 ). In 2002 the particle count does not reach the e-folding line. In 2004, the pa rticle count reaches this line by the end of the month a few days after the extratropical sto rm passed. The effects of extreme events in the Tampa Bay area is still a fairly new subject and there have not been many studies done. There i s much more work that needs to be done in future studies to help understand these ext reme events. In the future, conditions a few days before and after the storm may be studied to see what the circulation in the bay was just before an extreme event and how it reacts a few days after the storm leaves. Knowing the conditions before a storm can help to s ee how long it takes the bay to return to its prior conditions after a storm has passed. Water quality measurements before and after an extreme event can be used and compared to see if extreme events cause any “bad stuff” (harmful algal blooms or remnants from an oi l or phosphate spill) to be cleaned out of the bay. Each variable examined in this study c an be looked at in much more detail to

PAGE 151

139 Figure 79. Total number of particles in Tampa Bay during September 2002 (top) and September 2004 (bottom).

PAGE 152

140 Figure 80. Total number of particles in Tampa Bay during Decem ber 2002 (top) and December 2004 (bottom).

PAGE 153

141 get more specific results on how a storm impacts ea ch particular variable. Including a wetting-drying capability would better improve mode l output. Also using better boundary conditions and improving model resolution in regions of steep bathymetry would help to narrow the gap between the model outp ut and the observation data. An example would be to improve the model grid so orien tation at the Egmont Channel matches better with the bathymetry of the area, so that it is not a source of error in the model. With higher resolution the bathymetry use d in the model would resemble Tampa Bay better and by improving boundary conditio ns we could improve the results at the mouth of the bay.

PAGE 154

142 References Alshiemer, F. 2004. Hurricane Frances Preliminary Storm Survey I. http://www.srh. noaa.gov/tbw/information/2004/frances/localsummary. htm, NOAA National Weather Service. Beven, J. L. 2004. Tropical Cyclone Report: Hurric ane Frances 25 August-8 September 2004. http://www/nhc.noaa.gov/2004france s.shtml, NOAA National Weather Service. Blumberg, A. 1990. A Primer for ECOM-3D. Hydroqual, Inc. Mahway, New Jersey. 58 pages. Blumberg, A. and G.L. Mellor. 1987. A description o f a three dimensional coastal ocean circulation mode, p. 1-16 In N. Heaps (ed.), Three-dimensional Coastal Ocean Models, Volume 4. American Geophysical Union, Washi ngton, D.C. Boler, R. (Ed.). 1992. Surface water quality 1990-9 1 Hillsborough County, Florida. Environmental Protection Commission of Hillsborough County. Burwell, D. 2001. Modeling Eulerian and Lagrangian Estuarine Residence Times. Dissertation, College of Marine Science, University of South Florida, St. Petersburg, Florida. Burwell, D., M. Vincent, M. Luther and B. Galperin. 1999. Modeling residence times; Eulerian vs. Lagrangian, p. 995-1009. In Proceedings of the 6th International Conference on Estuarine and Coastal Modeling. Ameri can Society of Civil Engineers, New Orleans, Louisiana. Cameron, W.M. and D. W. Pritchard. 1963. Estuaries, p. 306-324. In M. N. Hill (ed.), The Sea: Ideas and Observations, Vol. 2. Wiley-Inte rscience, New York. D’Asaro, E. A. 2002. The ocean boundary layer below Hurricane Dennis. Journal of Physical Oceanography. 33:561-579. Estevez, E. D., R. R. Lewis III, S. K. Mahadevan, a nd J. L. Simon. 1985. The rationale for Tampa BASIS, p. 7-9. In Proceedings, Bay Area Scientific Information Symposium. Tampa, Florida.

PAGE 155

143 Forristall, G.Z. 1980. A two-layer model for hurric ane-driven currents on an irregular grid. Journal of Physical Oceanography. 10:1417-1438. Galperin, B., A. Blumberg, and B. Weisberg. 1992. T he importance of density driven circulation in well mixed estuaries: The Tampa Bay Experience, p. 332-343. In: M. L. Spaulding (ed.), Proceedings of the 2nd International Conference on Estuarine and Coastal Modeling. American Society of Civil Engineers, Tampa, Florida. Goodwin, C. R. 1980. Preliminary simulated tidal fl ow and circulation patterns in Hillsborough Bay, Florida. U. S. Geological Survey Open-File Rep., 80-1021, 25 pp. Goodwin, C. R. 1987. Tidal-flow, circulation, and f lushing changes caused by dredge and fill in Tampa Bay, Florida. U. S. Geological Su rvey Water Supply Paper 2282, 88pp. Goodwin, C. R. 1989. Circulation of Tampa and Saras ota Bays, p. 49-64. In: NOAA Estuary-of-the-Month, No. 11, Proceedings of a Semi nar. Washington, D. C. Hansen, D. V. and M. Rattray, Jr. 1966. New dimens ions in estuary classification. Limnology and Oceanography. 11:319-326. Lawrence, M. B. and H. D. Cobb. 2005. Tropical Cyc lone Report: Hurricane Jeanne 1328 September 2004. http://www.nhc.noaa.gov/2004jea nne.shtml, NOAA National Weather Service. Lewis, R. and R. Whitman, Jr. 1985. A new geograph ic description of the boundaries and subdivisions of Tampa Bay, p. 10-18. In Proceedings, Tampa Bay Area Scientific Information Symposium. Tampa, Florida. Li, Z. 1993. Circulation of Tampa Bay in Relation to Tides, Winds, and Buoyancy Driven Convection. Thesis, College of Marine Scien ce, University of South Florida, St. Petersburg, Florida. Li, M., L. Zhong, W. C. Boicourt, S. Zhang, and D. Zhang. 2006. Hurricane-induced storm surges, currents and destratification in a se mi-enclosed bay. Geophysical Research Letters. 33, L02604. Meyers, S. D., M. E. Luther, M. Wilson, H. E. Holm, A. Linville, and K. Sopkin. 2007. A numerical simulation of residual circulation in T ampa Bay. Part I: Lowfrequency temporal variations. Estuaries, in press.

PAGE 156

144 Paxton, C. 2004. Hurricane Jeanne Track and Intens ity Information. http://www.srh. Noaa.gov/tbw/information/2004/Jeanne/localsummary.h tm, NOAA National Weather Service. Pritchard, D. W., 1956. The dynamic structure of a coastal plain estuary. Journal of Marine Research. 15:33-42. Ross, B. E. 1973. The hydrology and flushing of the bays, estuaries, and nearshore areas of the eastern Gulf of Mexico, p. In: A summary of knowledge of the eastern Gulf of Mexico, St. Petersburg, FL. Smagorinsky, J., 1963. General circulation experime nts with the primitive equations, I. The Basic Experiment. Monthly Weather Review. 91: 99-164. SRH, 2004a. Hurricane Frances Preliminary Storm Su rvey II. http://www.srh.noaa.gov/ tbw/information/2004/frances/localsummary2.htm, NOA A National Weather Service. SRH, 2004b. Jeanne follows in Frances’ footsteps. http://www/srh/noaa/gov/tbw/ information/2004/Jeanne/localsummary2.htm, NOAA Nat ional Weather Service. SRH, 2004c. Stormy night punctuates wet, chilly Ch ristmas weekend. http://www.srh. noaa.gov/tbw/information/history/Christmas_2004.htm l, NOAA National Weather Service. Valle-Levinson, A., K. Wong, and K. T. Bosley. 2002 Response of the lower Chesapeake Bay to forcing from Hurricane Floyd. Continental Shelf Research. 22:1715-1729. Vincent, M. S., 2001. Development, implementation and analysis of the Tampa Bay Coastal Prediction System. Dissertation, College o f Engineering, University of South Florida Tampa, FL. Vincent, M., D. Burwell, M. Luther, and B. Galperin 1997. Real-time data acquisition and modeling in Tampa Bay, p. 427-440. In Proceedings of the 5th International Conference on Estuarine and Coastal Modeling. Ameri can Society of Civil Engineers, Alexandria, Virginia. Vincent, M., D. Burwell, M. Luther. 1999. The Tamp a Bay nowcast-forecast system, p. 765-780. In Proceedings of the 6th International Conference of Estuarine and Coastal Modeling. American Society of Civil Enginee rs, New Orleans, Louisiana. Walker, N. D. 2001. Tropical storm and hurricane wi nd effects on water level, salinity, and sediment transport in the river-inf luenced Atchafalaya-Vermilion Bay system, Louisiana, USA. Estuaries. 24(4):498-508.

PAGE 157

145 Weisberg, R. H. and R. G. Williams, 1992. Initial findings on the circulation of Tampa Bay. In: Proceedings, Tampa Bay Area Scientific Information Symposium 2, Tampa, FL, February 27-March 1, 1991, 77-89. Weisberg, R.H. and L. Zheng. 2005. A simulation of Hurricane Charley storm surge and its breach of North Captiva Island, Florida Scientist, in press. Weisberg, R.H. and L. Zheng. 2006a. Hurricane storm surge simulations for Tampa Bay, Estuaries and Coasts, 29: 899-913. Weisberg, R. H. and L. Zheng, 2006b. The circulati on of Tampa Bay driven by buoyancy, tides, and winds, as simulated using a fi nite volume coastal ocean model. Journal of Geophysical Research, 11,C1,C01005. Wilson, M., S. D. Meyers, and M. E. Luther, 2006. Changes in the circulation of Tampa Bay due to Hurricane Frances as recorded by ADCP me asurements and reproduced with a numerical ocean model. Estuaries and Coasts. 29:914-918. Zervas, C. E. (Editor), 1993. Tampa Bay Oceanograp hy Project: Physical Oceanographic Synthesis. NOAA, National Ocean Serv ice, Office of Ocean and Earth Sciences, Silver Spring, MD. 175 pp.