USF Libraries

A one-dimensional model of summer oxygen distributions within the Chukchi Sea

MISSING IMAGE

Material Information

Title:
A one-dimensional model of summer oxygen distributions within the Chukchi Sea
Physical Description:
x, 65 leaves : ill. ; 29 cm.
Language:
English
Creator:
Penta, Bradley
Publisher:
University of South Florida
Place of Publication:
Tampa, Florida
Publication Date:

Subjects

Subjects / Keywords:
Oxygen content -- Chukchi Sea   ( lcsh )
Dissertation, Academic -- Marine science -- Masters -- USF   ( fts )

Notes

General Note:
Thesis (M.S.)--University of South Florida, 1993. Includes bibliographical references (leaves 57-61).

Record Information

Source Institution:
University of South Florida
Holding Location:
Universtity of South Florida
Rights Management:
All applicable rights reserved by the source institution and holding location.
Resource Identifier:
aleph - 029656410
oclc - 29811883
usfldc doi - F51-00105
usfldc handle - f51.105
System ID:
SFS0040055:00001


This item is only available as the following downloads:


Full Text

PAGE 1

A ONE-DIMENSIONAL MODEL OF SUMMER OXYGEN DISTRIBUTIONS WITHIN THE CHUKCHI SEA by BRADLEY PENTA A thesis submitted in partial fulfillment of the requirements for the degree of Master of Science Department of Marine Science University of South Florida August 1993 Major Professor: John J Walsh Ph.D.

PAGE 2

Graduate School University of South Florida St. Petersburg, Florida CERTIFICATE OF APPROVAL Master's Thesis This is to certify that the Master's Thesis of BRADLEY PENTA with a major in Marine Science has been approved by the Examining Committee on July 14, 1993 as satisfactory for the thesis requirement for the Master of Science degree Examining Maj9ij John J. Walsh, Ph.D. Gabriel A. Ph.D.

PAGE 3

Bradley Penta 1993 c All Rights Reserved

PAGE 4

ACKNOWLEDGMENTS I wish to express my sincerest gratitude to my wife Heather for her support and encouragement during the preparation of this thesis. I also thank my parents for their help throughout the years in preparing me to get to this point. Thanks also to Dr. John Walsh, who's direction was essential for the completion of this work, and the rest of my committee, Dr. Kent Fanning and Dr. Gabe Vargo. And the others from the ECOS lab Dwight Dieterle, Mark Meyers, and Ray Pribble, for their help over the last few years.

PAGE 5

TABLE OF CONTENTS LIST OF TABLES LIST OF FIGURES LIST OF SYMBOLS ABSTRACT INTRODUCTION MODEL FEATURES SENSITIVITY ANALYSIS RESULTS Grazer Control Light Regulation Nutrient Regeneration Temperature Regulation Nutrient Loading CONCLUSION REFERENCES APPENDIX -Numerical Methods i ii iii v ix 1 15 36 37 39 44 52 54 56 57 62

PAGE 6

LIST OF TABLES Table 1. Initial conditions for the baseline model. 18 Table 2. Zooplankton biomass across Anadyr and Shpanberg Straits in July 1985. 31 Table 3. Resulting maximum of chlorophyll (mg chl m-2 ) and o2 flux, magnitude (103 moles 02 per 60 days), and direction (into or out of the water) from different model parameters. 38 ii

PAGE 7

LIST OF FIGURES Figure 1. The north Bering and Chukchi Seas (from Coachman et al., 1975) 2 Figure 2. Near-bottom temperature (a) and salinity Figure 3. Figure 4. Figure 5. Figure 6. Figure 7. (b) during August 1988 (from Walsh et al., 1989). 4 Surface (a) and near-bottom (b) distribution of silicate l -1), and the surface (c) and near-bottom (d) distribution of phosphate l-1 ) during August 1988 (from Whitledge et al., 1992). 5 Surface (a) and near-bottom (b) distribution of nitrate l-1 ) and the surface (c) and near-bottom (d) distribution of ammonium l-1 ) during August 1988 (from Walsh et al., 1989). 6 The Chukchi Sea. Including August ice data, current meter locations and measured currents, and the location of ice station T-3. 8 Oxygen saturations (%) in the surface water of the Chukchi Sea during (a) the 1968 cruise of the Northwind, and (b) the 1969 cruise of the Staten Island. (Kinney et al., 1970). 10 The monthly o2 saturation (%) within the upper 10-20 m of the water column of the western Chukchi {1922) and East Siberian {1923-24) Seas during the two year drift of the Maud (after Sverdrup, 1929). 11 Figure 8. Dissolved oxygen saturation (%). The Northland and Chelan sampled the eastern Chukchi, Bering Strait, and the eastern Bering Sea. The Northwind and Burton Island explored the Chukchi Sea and the East Siberian Sea (after Codispotti et al., 1991). 12 iii

PAGE 8

Figure 9. The simulated flow field from Shuert and Walsh, 1993 16 Figure 10. The modelled solar radiation incident on the sea surface. 21 Figure 11. Distribution of zooplankton biomass in Chirikov basin and the southern Chukchi Sea (after Kulikov, 1992). 34 Figure 12. The distribution of macrofaunal benthic biomass (from Grebeier et al., 1988). 35 Figure 13. Baseline model results. 40 Figure 14. Distribution of integrated chlorophyll (a), and primary production (b) in the Ch ukchi Sea during August, 1988 (from Korsak, 1992, and Robie et al. 1992). 41 Figure 15. Model results from case with ice edge at 69 N. 43 Figure 16. Modelled contours of phytoplankton chlorophyll in baseline case. 46 Figure 17. Modelled contours of ammonium (a-c) and nitrate (d-f) with each of the kz values used. 47 Figure 18. Oxygen saturation ( % ) in the near-bottom layer d uring midsummer cruises of the Brown Bear during 1960 and the Northwind during 1963 (from Coachman et al., 1975). 48 Figure 19. Figure 20. Figure 21. Figure 22. Model results from the kz = 125. 0 cm2 sec-1 case. Model results from the kz = 5 0 cm2 sec-1 case. Model results from the increased temperature case. Model results from the east Chukchi Sea with initial nutrient conditions equal to the west 50 51 53 Chukchi Sea nutrients of the baseline case. 55 iv

PAGE 9

a B f LIST OF SYMBOLS The time of sunrise Constants in the saturation oxygen concentration equation The biomass of the grazers Constants in the saturation oxygen concentration equation Drag coefficient The specific nitrate uptake rate The specific ammonium uptake rate The difference between the oxygen concentration in the surface water and that at saturation The change of oxygen due to biogenic production The change of oxygen due to gas exchange with the atmosphere The change of oxygen due to mixing The net change of oxygen Phytoplankton growth rate The gas exchange coefficient for oxygen Wind stress Coriolis parameter Net oxygen flux across the air-sea interface The maximum grazing rate of the zooplankton Grazing loss v

PAGE 10

G(z) GPP I s 0 I w 0 i(d) k Grazing flux of zooplankton Gross primary production Maximum incident radiation at the surface each day Mean incident radiation at the summer solstice The hourly change of solar irradiance Mean incident radiation at the winter solstice The saturating light intensity for the phytoplankton Irradiance at depth Photoperiod for each day Photoperiod of the summer solstice Photoperiod of the winter solstice Net extinction coefficient The Michaelis-Menton half saturation constant for ammonium Phytoplankton concentration at which half the maximum grazing occurs Extinction coefficient of ice The Michaelis-Menton half saturation constant for nitrate Extinction coefficient of phytoplankton Extinction coefficient of dissolved components of seawater Extinction coefficient of pure water Horizontal mixing coefficient (x-direction) Horizontal mixing coefficient (y-direction) vi

PAGE 11

lG M NH + 4 w PQ Pa rB rG rm T u v The vertical mixing coefficient The rate of excretion of ammonium by zooplankton The realized growth rate of phytoplankton under light limitation The maximum specific growth rate of the phytoplankton Phytoplankton biomass The grazing threshold Ammonium Nitrate Earth's rotation rate The oxygen concentration at saturation Latitude Empirically determined constant which represents the suppression of nitrate uptake by ammonium stocks Photosynthetic quotient The density of air The benthic regeneration rate Zooplankton respiration Respiration rate of phytoplankton Salinity Absolute temperature The x-component of the velocity field The total specific nitrogen uptake rate vii

PAGE 12

v The y-component of the velocity field The wind velocity The phytoplankton sinking velocity Depth viii

PAGE 13

A ONE-DIMENSIONAL MODEL OF SUMMER OXYGEN DISTRIBUTIONS WITHIN THE CHUKCHI SEA by BRADLEY PENTA An Abstract Of a thesis submitted in partial fulfillment of the requirements for the degree of Master of Science Department of Marine Science University of South Florida August 1993 Major Professor: John J. Walsh Ph.D. ix

PAGE 14

A one-dimensional numerical model was constructed to simulate phytoplankton nutrient uptake, production, and oxygen evolution within the Chukchi Sea. Included were estimates of grazing and respiration by both zooplankton herbivores and benthos. The patterns of surface water oxygen saturations from the 1922-1924 drift of the Maud and 1968-1969 ice-breaker cruises in the Chukchi Sea were simulated using this model. The variation between the surface oxygen saturation in the East Siberian Sea and the surface oxygen saturation of the western Chukchi Sea (measured by the Maud) is found to be a result of light limitation of phytoplankton growth (due to ice cover in the East Siberian Sea). The variation in the 1968-1969 oxygen data between the east and west Chukchi Sea is, instead, due to biogenic nutrient variations between these two regions. Abstract Approved: Major ijohn J. Walsh, Ph.D. Professor, Marine Science Date Approved: -.c X

PAGE 15

1 INTRODUCTION Through the use of mathematical models based on the numerous observations of dissolved oxygen within summer waters of the Chukchi Sea, estimates can be made of the equivalent biological fixation of atmospheric co2 and its possible sequestration within high-latitude marine ecosystems (Walsh, 1989; Zeeman, 1992). Accordingly, a one-dimensional numerical model of nutrient uptake, photosynthesis, and oxygen evolution is used to simulate element cycling by plankton and benthos, along two latitudinal paths through the oligotrophic eastern and eutrophic western halves of the Chukchi Sea. Temperature and light conditions from 1988 are used to constrain the metabolic processes, while independent oxygen observations from 1964-66 and 1968-69 constitute the validation data. The Chukchi Sea extends -700 km from the 85-km wide Bering strait between Cape Prince of Wales and Cape Dezhneva at -66 N, to the edge of the Arctic Ocean at -73 N (Figure 1). The Alaska mainland and the Beaufort Sea border the Chukchi Sea to the East (-165 W), while the Siberian coast and the East Siberian Sea (-180) demarcate the western boundary. The bottom topography of the Chukchi sea is dominated by a wide, shallow continental shelf, with an average depth of -so m This shallow shelf depth allows most

PAGE 16

180" Figure 1. 178' 176' 170" 168' 166' 164' 162' 160" 58' :;;I 40 \ .... .._ '\ ..__ I CHUKCHI I I i I I I r8' I I -66 164' .162' 160' 158" The north Bering and Chukchi Seas (from Coachman et al., 197.5) 2

PAGE 17

3 of the water column to be well mixed by winds, such that a barotropic circulation model can be used to estimate the summer flow fields (Overland and Roach, 1987; Shuert and Walsh, 1992) The flow of water through Bering strait is the only interchange between the North Pacific and the Arctic Oceans (and ultimately the North Atlantic Ocean). This flow is comprised of two distinct water masses: Anadyr Water to the west and Alaskan Coastal Water to the east (Coachman et al., 1975). Anadyr Water, at a mean speed of 15 em sec-1 provides greater than 60% of the summer transport through Bering Strait. It originates from outer shelf water of the Gulf of Anadyr (Figure 1) and is characterized by higher salinity (>32. 5 %o), lower temperature (1-2 C), and higher nutrients, as a consequence of upwelled waters in Anadyr Strait (Walsh et al., 1989) Alaskan Coastal Water flows instead from the nearshore southeastern Bering Sea through Shpanberg Strait at -5 em sec-1 ; it is characterized by lower salinities (<31.8 o/oo), due to seasonal freshwater influxes from the Yukon River, warmer temperatures (>4 C), and lower nutrients, typical of postbloom conditions (Coachman and Shigaev, 1992; Shuert and Walsh, 1992), as shown in Figures (2-4). The zonal nutrient concentrations of the Chukchi Sea thus display distinct east-west gradients, due to the different time histories of plankton dynamics within the distinct water types affecting each region (Figures 3 & 4). For example,

PAGE 18

180" 69" \.. 4--..__" 1 . / (a) 175 170 r:. 33. 0 _.j:\ : (b) no 1ao 175 Figure 2. Near-bottom temperature (a) and salinity (b) during August 1988 (from Walsh et al., 1989).

PAGE 19

5 (b) 180 175 170 1 65 180 175 170 165 CHUKCHI 180 Figure 3. (c) 175 170 165 180 1 75 170 165 Surface (a) and near-bottom (b) distribution of silicate 1-1), and the surface (c) and near-bottom (d) distribution of phosphate 1-1 ) during August 1988 (from Whitledge et al., 1992 )

PAGE 20

6 ... o(a) \""" . 1 "'-'o .,.,. .,. --\ '"'5 (b) JO aoo I , ,,o . c9 1 Q . . (J? oo ,,. ( C ) Figure 4. Surface (a) and near-bottom (b) distribution of nitrate 1-1 ) and the surface (c) and near-bottom (d) distribution of ammonium 1-1 ) during August 1988 (from Walsh et al., 1989).

PAGE 21

7 ten-fold higher nitrate stocks are found at the surface of the western side than in the eastern side, see Figure 4a (Walsh, et al., 1989; Whitledge, et al., 1992), while near-bottom ammonium concentrations show a 400% greater value on the western side (Figure 4c) The biological responses to these different physical and chemical features of the shelf habitat are further modulated by the light field, which is in turn controlled by ice cover, cloud cover, water column turbidity, including self-shading by large phytoplankton blooms, and the seasonal changes of incident radiation. The ice cover over the Chukchi Sea varies from total in winter to partial coverage (Figure 5) over only the northern shelf-break region in summer (Naval Oceanography Command Detachment, Ashville, 1986), while the photoperiod varies from 21.5 hours on August 1 to 10.7 hours on September 30, and the incident PAR varies between 178.89 and 89.26 gcal cm-2 day-1 (see Figure 10). Phytoplankton biomass and primary production are tenfold higher ( >800 mg chl m-2 and >10 g C m-2 day-1 ) on the western side of the Chukchi Sea during August 1988, than on the eastern side (Walsh et al., 1989) Significant differences in oxygen evolution (Figure 6) on the eastern and western sides of the Chukchi Sea may be a consequence of nutrient-limitation of phytoplankton growth, rather than temperature effects on gas saturation. This hypothesis is tested in the simulation analysis by following the

PAGE 22

200 170 --.. ' ........... I ; (:::::::::::::: : \_ . ::: : AUGUST15::: ICE COVER : .. :::::MEAN .... AUGUST 1991 : : : : : CURRENTS : : : : : : : 40 em sec : : : : : 8 Figure 5. The Chukchi Sea. Including August ice data, current meter locations and measured currents, and the location of ice station T-3.

PAGE 23

9 hypothetical trajectory of two water parcels within oligotrophic and eutrophic regions. Historical oxygen, temperature, and salinity data, from the 1922-1924 (Figure 7) cruise of the Maud (Sverdrup, 1929), and 1968-69 (Figure 6) ice breaker cruises (Kinney et al.,1970) were used for validation of the model. Initialization data were taken from a 1988 cruise of the Akademik Korolev (Whitledge, 1988). oxygen saturations of up to 134% for example, were reported by Sverdrup (1929), while values > 160% have been found by the cruises during the 1960's (Figures 6 and 8). The Maud data suggest interannual differences of 95-134% in August oxygen saturation which may reflect light conditions, i.e. ice cover in the East Siberian Sea during 1923 and 1924, with open pack ice in the western Chukchi Sea during 1922. The Maud observations are plotted again in Figure 8 of more recent oxygen measurements from ice-breaker cruises to the Chukchi Sea, i.e. recall Figure 6. The even larger variations of more recent August saturation values, from 55 to 160%, may instead reflect nutrient limitation of primary production and oxygen evolution in the eastern Chukchi Sea (Figures 3-4). Both oxygen and carbon dioxide exchange between the atmosphere and the ocean in gas form, as functions of their partial pressures in air and in seawater. The solubility of each gas in seawater depends upon both physical and chemical

PAGE 24

\107;>!-..r___ o 1}oo.o\ I \ '-/ 109.0\ \.__../ l o I / ..... _.. -I I I / / / -117.5\ ( 121.0\ (a) (b) Figure 6. oxygen saturations (%) in the surface water of the Chukchi Sea during (a) the 1968 cruise of the Northwind, and (b) the 1969 cruise of the staten Island. (Kinney et al. 1970). 1-' 0

PAGE 25

s A 110T 100U R A T I 0 90N 11 1922 1923 1924 J A S 0 N D J F M A M J J A S 0 N D J F M A M J J A .. .... .. .... .... ... .. - Figure 7. The monthly o2 saturation (%) within the upper 10-20 m of the water column of the western Chukchi (1922) and East Siberian (1923-24) Seas during the two year drift of the Maud (after Sverdrup, 1929).

PAGE 26

160 ISO z lio 0 ;:: ... a: "'100 z "' "' >-s 50 40 0 0 0 0 0 0 .a .. t q 8 .: J 0 0 : 0 rf. -j . .... 1:\i" _, 0 0 0 o : . > Ill .. 0 0 0 0 '() -\ .. : [ 0 : :9 .\ tit 0.;. -- 0. m. ._ .... =-00 g, <8> () :B 'ib :'":.. 0 0 0 0 ;. 0 QJ 0 [1:1 8 0 o , 0 0 cB .... _: 0 0 ... . t'" 'lt ... . I .:. LEGEND 0 MAUO, 1'.122-1924 0 CHELAN,I9341 { NORTtfflND ISLAND{= 12 0 0 Figure 8. Dissolved oxygen saturation(%). The Northland and Chelan sampled the eastern Chukchi, Bering Strait, and the eastern Bering Sea. The Northwind and Burton Island explored the Chukchi Sea and the East Siberian Sea (after Codispoti et al., 1991).

PAGE 27

13 factors-temperature, salinity, pressure. The concentrations of these gasses are dependent on the biological factors of production and respiration. The primary production is regulated, in turn, by both temporal and spatial changes of the nutrient and light fields, and by the type of herbivore. The food web is based mainly upon diatoms in the western Chukchi Sea, where copepods might be the dominant herbivore (Walsh et al., 1989). While, in the eastern region, the primary producers are flagellates; here, ciliates may instead crop the smaller stocks (Walsh et al., 1989). Grazing by zooplankton and the benthos also effects the f ratio of "new" to total ("new" plus recycled) production, defined as nitrate based vs. total nitrogen (nitrate plus ammonium) derived protein synthesis, where the f ratio is N03/(N03 + NH4); in this analysis, urea is included as part of the ammonium pool. If zooplankton fecal pellets survive bacterial degradation in the water column, they add to the pool of organic matter of phyto-detritus in the sediments. During bacterial and animal metabolism, water column oxygen is consumed, while ammonium and carbon dioxide are produced, particularly in the winter (Figure 7); nitrification may be important in the southeastern Bering Sea during winter as well (Walsh and Dieterle, 1993), but is ignored during 60 days of summer in the Chukchi Sea. In an attempt to describe these various processes, and to resolve the origins of the surface water oxygen variations found in the data, a depth dependent

PAGE 28

14 simulation model of phytoplankton nutrient uptake, production, and oxygen evolution was constructed.

PAGE 29

15 MODEL FEATURES The physical context of the model is derived from the observed near-bottom flow field, at depths of 35-47m past current meter moorings 1-8 (K. Aagaard, personal communication) during August 1991 in the southern Chukchi Sea (Figure 5). The flow field is similar, in pattern, to the results of a barotropic model for August 1, 1985 (Figure 9), under a total transport of 0.95 sv through Bering strait (Shuert and Walsh, 1993). Once the waters (simulated and observed) exit Bering Strait at a speed of as much as 50 em sec-1 they both slow to ::::; 10 em sec-1 and mainly flow northwestward, along the isobaths, towards Wrangel Island, with weak currents, or southward reversals of direction, along the Alaskan coast. Outside of the prior model domain, waters of Pacific Ocean origin continue to follow the bottom topography of Herald and Barrow Canyons at respective moorings 9 and 11, while a weak influx of East Siberian Sea water occurs in Long Strait at mooring 10, between Wrangel Island and the Siberian mainland (Figure 5). During August 1991, the mean flow at moorings 2-4 and 6-9, along the main axis of northward transport, was 12.6 em sec-1 towards 335T.

PAGE 30

174 I 1 7 2 I 170 I , , .. .. .. ' .. . , 't,"" , .. t ' 16 4 I ' # , ... t , SHPANBERG STRAIT' ..t ' 6230 em/sec VELOCITY FIELD NOME WINDS CASE JULIAN DAY-214 BERING STRAIT TRANSPORT 0 95 Sv Figure 9. The simulated flow field from Shuert and Walsh, 1993 16

PAGE 31

17 In this analysis, water parcels are assumed to transit the Chukchi Sea, from Bering Strait to the shelf-break, at initial speeds of 50 km day-1 for two days and then at 10 km day-1 thereafter, i.e. -12-58 em sec-1 A Lagrangian coordinate system is then defined, fixed relative to the flow field, rather than an Eulerian system fixed in space. In this flow field, the forward advection is orders of magnitude greater than the horizontal diffusion (Coachman et al., 1975); therefore horizontal diffusion is ignored in the model. Accordingly, the results of the model are presented in both time and equivalent distance frameworks, with 1 day = 10 km northward displacement, except for the first two days, over which 100 km is traversed by the phytoplankton populations. The numerical methods used in this analysis are given in the appendix. Two cases of the model are initialized with respect to nitrate conditions of 580 mg-at N03 m-2 and 6 mg-at N03 m-2 and ammonium conditions of 26.5 mg-at NH4 m-2 and 10 mg-at NH4 m-2 on 2 August, reflecting the 1988 chemical habitat on the eastern and western side of Bering Strait (Whitledge, 1988). Initial depth-integrated chlorophyll values of 496.5 mg chl m-2 and 22 mg chl m-2 are taken from the ISHTAR observations as well (Table 1) reflecting separate populations of diatoms and flagellates. The initial conditions were entered into the model in eight 5 m depth boxes. The time history of nutrients and phytoplankton populations are then simulated for a period

PAGE 32

18 Table 1. Initial conditions for the baseline model. west east Nitrate (mg-at. m-2} 580.0 6.0 Ammonium (mg-at. m-2} 26.5 10.0 chlorophyll (mg chl. m-2} 496. 5 22.0 temperature ( oc} 2 4 salinity (%o} 32.8 30.7 oxygen (moles m-2} 3.05 3.05 kz (cm 2 sec-1 } 25.0 25.0 of 60 days, during the northward trajectory of two water parcels from Bering Strait along a nominal 40 m isobath. Similarly, two sets of water temperatures are read into the model initially at Bering Strait (Table 1}. By the time the ice-edge is reached, surface water must cool to the freezing point for seawater at ambient salinity (Riley and Chester, 1971}. During the 60 day time period, a linear relation is used to calculate the temperature between these two endpoints, subjecting the biotic variables of the model to a time-dependent temperature field, while the salinity is held constant at 32.8 psu in the west and 30.7 psu in the east. The maximum incident radiation at the surface is calculated each day from:

PAGE 33

19 (1) where I0 w and I0 9 are the mean incident radiation at the winter and summer solstices and 356 is the Julian day of the winter solstice (Nagle, 1978; Walsh, 1988). The photoperiod for each day of the model run is then calculated (Figure 10) by (Nagle, 1978): (2) where i (d) is the photoperiod of Julian day d, i w is the photoperiod of the winter solstice, i8 is the photoperiod of the summer solstice, and 356 is the Julian date of the winter solstice. Finally, the hourly change of solar irradiance is represented by: l() I (d) 1tx(t-a) Ot = max i(d) (3) where the time of sunrise, a, is calculated by dividing the photoperiod by two and subtracting from twelve (this centers the daylight hours around noon). Below the surface of the model, vertical mixing is induced by the wind stress. The stress is calculated by a quadratic drag law: where Pa is the density of air and W is the wind velocity. c10 is a dimensionless drag coefficient that varies with W,

PAGE 34

20 (4) such that at W 7.0 m sec-1 c10 equals 1.6x1o-3 ; at w 10.0 m sec-1 c10 equals 2.5x1o-3 ; and at 7.0 < W > 10.0 m sec-1 c10 = 0.0003 (W7.0) + 0.0016 (Shuert, 1990). The vertical mixing coefficient is calculated from the wind stress (Csanady, 1976} by: k = z F 200/ (5) where F is again the wind stress and f is the Coriolis parameter in a f-plane approximation: f = 2(a) sinJ> (6) in which w is the Earth's rotation rate and is latitude. The wind velocity was obtained from NOAA climatological data. The average wind velocity during August-September 1985-1987 at Nome, Gambell, and Kotzebe, Alaska was 5.75 m sec-1 resulting in a kz of 25.0 cm2 sec-1 The other physical processes of the model are air-sea exchange of oxygen, initial supply of new nutrients, and vertical resupply of recycled nutrients. At the air-sea interface, the vertical exchange, effected by kz, is replaced with another expression for gas exchange, which ignores wind speed, but does consider ice cover. The net oxygen gas flux, F (moles m-2 day-1 } across the air-sea interface is computed by (Peng et al., 1987}:

PAGE 35

200 -150 "0 E 100 0 0 y 50 I C7l 0 I 0 5 10 15 20 Llqht 1 nc1aem on 25 30 day 35 40 45 50 55 Figure 10. The modelled solar radiation incident on the sea surface. -: -: -: --: 60 f'V ....

PAGE 36

22 (7) where E(02 ) is the gas exchange coefficient for oxygen, Ao2 is the difference between oxygen concentration at saturation and that computed for the surface layer, and (o2>sat is the oxygen concentration at saturation. When sea ice is present in the model, the air-sea gas exchange is reduced to 25% of the value computed without ice (Fanning and Torres, 1991). The saturation oxygen concentration in the underlying sea water is a function of salinity and temperature; it is computed by (Weiss, 1970): ln(OJsat = A1 100 ) +A3ln(_!_) +A4(_!_) +So/oo[B1 +B2(_!_) +B3(_!_)2 ] T 100 100 100 100 (8) where (o2>sat is in units of ml/1 (or mlfkg), Tis absolute temperature, &roo is salinity, and Ai and B1 are constants. For oxygen, these constants are: Al = A2 = A3 = A4 = The -173.4292 249.6339 143.3483 -21.8492 Ao2 value is B 1 = -0.033096 B 2 = 0.014259 B 3 = -0.0017000 equivalent to the apparent oxygen utilization (AOU) value (Peng et al., 1987). The value for the gas exchange coefficient is E(02 ) = 1.047 mol m-2 day-1 and is considered independent of wind and temperature (Peng et al., 1987). A wind-dependent E(C02 ) increased air-sea fluxes of co2 by 15%, compared to a constant value (Taylor et al.,

PAGE 37

23 1991), in a seasonal model of the southeastern Bering Sea (Walsh and Dieterle, 1993). Since only 60 days of August and September are included in this model, the effects of seasonal changes of wind speed on E(02 ) are ignored. The constant gas exchange coefficient is equal to the flux of oxygen into the sea, if the surface concentration of 0 2 were zero. The oxygen concentration in the surface seawater needs to be computed, as a function of time, for use in the net gas transfer of equation (7). The net change of the oxygen is a function of three processes: water mixing, gas exchange with the atmosphere, and biologically affected changes. (9) where the boundary conditions are: at z=O, ao2jat = F(02 ) at z=40 m, kz ao2jat = -rB The biogenic term, (ao2;at)b, is expanded in equation 19. In this Chukchi Sea model, the limiting nutrient of primary production is assumed to be nitrogen. It is supplied by upstream input through Bering Strait, as well as by the process of nutrient recycling; ignored are both in situ nitrification (Walsh and Dieterle, 1993), and supply of nitrate by upwelled Atlantic water at the shelf-break, since the N:P ratio suggests Pacific origin of nutrients in the Chukchi Sea (Codispoti and Richards, 1968).

PAGE 38

24 The tempo-spatial distribution of "new" nitrogen, supplied as initial nitrate conditions at the Bering strait boundary, can be described by (Walsh, 1988}: aNo-3 aNO aNO cPNO cPNO cPNO -= -u 3 -v 3 +k 3 +k 3 +kz 3 -xlM (10) 0x i1y X ax2 y ay2 az2 where the first two terms represent horizontal advection, the third and fourth represent lateral diffusion, and the fifth term represents vertical mixing due to wind stress; upwelling or downwelling is ignored here. The final term of equation (10} represents uptake of nitrate, as a function of phytoplankton biomass, M, and, x1 the specific nitrate uptake rate; x1 is expanded within equation (17} in the usual Michaelis-Menten formulation, with an ammonium-inhibition factor (Wroblewski, 1977). Because the present model is one-dimensional, the lateral advection and diffusion terms drop out; equation (10} then reduces to: cPNO3 = k -x M z: az2 1 (11) The fluxes of recycled nitrogen are represented by a similar equation for ammonium (Walsh, 1988}: aNH + cPNH --4 -= k 4 -x M+lG at z: az2 2 (12) where lG is the rate of excretion of ammonium by zooplankton. For equations (11-12}, there are "no-flux" boundary conditions

PAGE 39

25 at the top of the water column, i.e. nitrogen fixation is ignored. The bottom boundary condition of equation (12) is: kz <3NH4+;az = rB where rB equals the flux of ammonium out of the sediments due to benthic regeneration. The two distinct nutrient regimes, embodied in the initial conditions (Table 1), support different phytoplankton assemblages. In the model, a generic diatom represents the phytoplankton community of the eutrophic western Chukchi Sea, while a generic flagellate represents the oligotropic eastern community. Phytoplankton biomass (g C m-3 ) is represented by the equation (Walsh, 1988): aM cJlM aM = k --+eM-G-w--rM at z az2 s az. (13) where is the phytoplankton growth rate, G is the grazing loss, w9 is the sinking velocity of phytoplankton, and r is the respiration rate. Equation (13) is subject to the boundary conditions: at z=O, at z=40 m, kz <3M/<3z = 0 and w9 <3Mj<3z = w9 M The variable sinking rate (w5 ) of diatoms is computed as a function of their biomass, in units of chlorophyll, to simulate aggregation of diatoms; = 0.011 X Flagellates do not sink in the model. See the appendix for a discussion of the sinking term and its numerical expression.

PAGE 40

26 The maximum specific growth rate of the diatoms and flagellates is a function of temperature (Eppley, 1972). = 0.85 (100. 02751) (14) However, the realized growth rates of both groups of plankton never reach maximal values due to the limiting effects of either light or nutrients. The model considers only a single limiting factor at each time step, rather than a multiplicative interaction between limiting factors (Walsh, 197 5) The underwater light-field experienced by the phytoplankton is described by Beer's law for the irradiance, Iz, at each given depth (Walsh, 1988): I = 0 5 x I e -A4 z 0 (15) where r0 is the incident radiation from equation (3) and 0.5 is the amount of photosynthetically active radiation (PAR). The extinction coefficient (k) was subdivided into extinction, or, more accurately, attenuation, due to pure water = 0.015 m-1 ; Walsh, 1988), due to phytoplankton (kp = 0.03 x chla; Smit:h and Baker, 1982) due to ice (ki = 4. 61 m-1 ; Niebauer and Smith, 1985), and due to other dissolved components of the sea water (k9 = 0.065 m-1 ; Walsh, 1988) such that k = kw + kp + k5 + ki. The light limitation of the growth rate is described by (Parsons et al., 1984):

PAGE 41

27 (16) where Iz is the incident light at depth z, Isat is the saturating light intensity for the phytoplankton (2.5 g-cal cm-2 hr-1 ; Walsh and Dieterle, 1993}, and is obtained from equation (14}. The limiting of phytoplankton growth by nutrients is described next. The uptake of nitrogen is assumed to be the limiting factor of nutrient regulation. When available, ammonium and urea are preferentially taken up over nitrate (McCarthy, et al., 1977; Kristiansen, 1983); the enzyme nitrate reductase, which is required for the uptake of nitrate, is inhibited by the presence of these reduced forms of nitrogen (Packard and Blasco, 1974; Wroblewski, 1977). The total specific nitrogen uptake rate (time-1}, V, is thus described by Wroblewski (1977): (17) where the Michaelis-Menton half saturation constants for nitrate and ammonium (kn and ka} are 1.0 l-1 and 1.5 at 1-1 respectively (Eppely et al., 1969, Walsh, 1975). These are the concentrations of nitrogen which support one-half of the maximum uptake rate, Psi (''l') is an empirically ( t NH4+ l-1 ) 1 (W bl k' determined constant of 1.462 ro ews 1,

PAGE 42

28 1977), which represents the suppression of nitrate uptake by ammonium stocks. Epsilon ( ) of equation ( 13) is now set equal to V or whichever is smaller, at each time step. The phytoplankton evolve oxygen according to the simplified equation for C02 uptake, 02 evolution and organic matter synthesis (Richards, 1965): H3P04 + l6HN03 + 122H20 + 106C02 ... 13802 + (CH20)106 (NH3 ) 16 H3P04 (18) This yields a Photosynthetic Quotient (PQ) for of 1. 30. If the nitrogen source is ammonium, rather than nitrate, the PQ value is 1.0 The expanded expression for of equation (9) is: ( 6 = GPP -rM -rG (19) subjuct to the bottom boundary condition: at z=40 m ao2;at = -rB where GPP is gross primary production, rM is phytoplankton respiration, rG is zooplankton respiration, and rB is benthic respiration. The gross production of o2 comes from the M term of equation ( 13) for synthesis of phytoplankton biomass, with a C/02 conversion ratio of 1.3 or 1.0, depending upon the nitrogen source. If nitrate is taken up, it must first be reduced to ammonium before it can be assimilated; this

PAGE 43

29 produces two molecules of oxygen per atom of nitrogen assimilated (Williams and Robertson, 1991). Phytoplankton respiration, in terms of lost carbon or oxygen, is assumed to be ten percent of gross primary production (Falkowski and Owens, 1978). Zooplankton are assumed to respire 30% of what they injest, with the rest lost to the benthos as fecal pellets. The benthic animals are assumed to respire all food (plants or as fecal pellets) which they ingest. Computation of the grazing stress is discussed first. Across Anadyr and Shpanberg Straits, through which Anadyr and Alaskan Coastal Waters flow before entering the Chukchi Sea, the abundances (number of individuals m-2 ) of dominant copepod species were measured in July 1985 (Walsh et al., 1989). The mean of these data, the approximate dry weight per individual of each species (Vidal and Smith, 1986), and the resultant carbon biomass (using a conversion of 0.45 C 1 dry wt. ) are displayed in Table 2. Calanus marshallae is a copepod species which grazes large phytoplankton This model contains exclusively small flagellates on the eastern side of Bering Strait, therefore the population of C. marshallae measured in Shpanberg Strait is ignored in the estimate of zooplankton grazing on the flagellates. The zooplankton biomasses for the Chukchi Sea obtained from these data are thus -4490 mg c m2 for the west and -200 mg c m-2 for the east.

PAGE 44

30 However, zooplankton abundances in the Bering and Chukchi Seas were also measured (Figure 11) during August 1988 (Kulikov, 1992). Ten-fold less Neocalanus plumchrus, Calanus Acartia RP., and Pseudocalanus 2RP were found in 1988 than 1985, while abundances of Metridia pacifica, Eucalanus bungii, Centropages RP. and Oithona similis were about the same. These differences of herbivore biomass suggest either interannual or seasonal (Walsh et al., 1989) variations of biomass import from the southern Bering Sea. Levels of total biomass in 1988 were found to be a mean of -19 g wet wt. m-2 in the southern Chukchi Sea and northern Bering Sea (Kulikov, 1992); i.e. approximately 1330 mg C m2 or about half the mean in 1985. Around Bering Strait (Figure 11), where the model is initialized, 1988 zooplankton biomass was only -5 g wet wt. m2 or 350 mg c m2 ; both 1985 and 1988 estimates of biomass are used to assess grazing stress on the western diatom population. The zooplankton grazers in the water column are allowed to ingest 10% of their biomass per day (Dagg, et al., 1982), with the caveat that phytoplankton have refuge populations of 4.5 x 10-4 mg c m-3 for diatoms, and 15.0 x 10-4 mg C m-3 for flagellates (Meyers, 1993). The equation used for zooplankton grazing (Walsh, 1975) is:

PAGE 45

Table 2 Z o opla n k ton biomass across Anadyr and Shpanberg strait s i n July 1985. copepod species ll ind .m -2 IN dry wt. ind. -1 mg Q Anadyr Strait Neocalanus cristatus 10 1000 4 5 Neocalanus plumchrus 3600 1000 1620 Eucalanus bungii 2300 1000 1035 Metridia pacifica 3000 80 10.8 Oithona simi lis 25000 1 11.25 Pseudocalanus spp. 18000 12 97 Calanus marshallae 1000 600 270 Centropages abdominal is 1100 1. 75 9 Shpanberg Strait Metridia pacifica 620 80 22.3 Oithona simi lis 5000 1 2 .25 Pseudocalanu s spp. 22000 12 119 Cal anus marshallae 15000 600 4050 Acartia longiremis 10500 12 57 31 m -2

PAGE 46

G(z) = y B M(z)Mu. k 1+(M(z)Mu.) 32 (20) where G(z) is the grazing flux (mg c m-3 day-1 ) at depth z, B is the biomass of the grazers, is their maximum grazing rate ( 10% of B per day) Mth is the grazing threshold (refuge population) and kf is the phytoplankton concentration at which half the maximum grazing occurs (3.17 mg c m-3). The zooplankton are assumed to remineralize 30% of the nitrogen that they ingest (Shuert and Walsh, 1993); the equivalent carbon loss is their respiration. The other 7 0% is incorporated into fecal pellets which sink to the bottom at 100 m day-1 (Walsh and Dieterle, 1993) to be ingested by the benthos. In the eastern Chukchi Sea, macrobenthic biomass i s -5 g c m-2 (Figure 12), compared to as much as so g c m-2 in the western half (Grebmeier, 1992). The benthic maximum grazing rate is also assumed to be 10% of their biomass per day. The benthic grazing in the bottom interface of the model consumes any phytoplankton and fecal pellets which have sunk in to the sediments, i.e. the bottom boundary condition of equation (13) in the model. The benthos can ingest from these two sources exclusively. The oxygen respiration and ammonium regeneration of the two herbivores are then computed from their grazing. In terms of carbon, reproduction and mortality of the herbivores are ignored, such that all o f the ingested phytoplankton carbon is

PAGE 47

33 respired; one mole of oxygen is required to respire one mole of carbohydrate. The respiration of the heterotrophs is 12 moles o2 per c ingested. The zooplankton excretion flux is 30% of their ingestion in units of nitrogen. The benthos meanwhile, are assumed to regenerate ammonium as 80% of their nitrogen ingestion {Shuert, 1990).

PAGE 48

34 1 mm 2 5 glm 2 83 ALASKA 0 0 92 0 100 106 0 0 102 Bering Sea 1 mm 2 5 Ofm 2 0 48 0 0 47 0 45 57 0 ee 0 74 Oo Figure 11. Distribution of zooplankton biomass in Chirikov basin and the southern Chukchi Sea (after Kulikov, 1992)

PAGE 49

Figure 1 2 1 10 20-30 30 40 50 0 0 180 CHUKCHI SEA 0 0 0 0 0 0 0 o : 0 . CHUKCHI PENINSULA 0 0 0 0 175 0 . BER I NG SEA 170 35 65 64 62 61 165 The distribution of macrofauna! benthic biomass (from Grebeier et al., 1988).

PAGE 50

36 SENSITIVITY ANALYSIS Various environmental conditions are explored with the model; among these are the changing physical habitat of ice cover, wind, and temperature (Table 3). Different boundary conditions of nutrients across Bering Strait, as well as different benthic rates of regeneration, different zooplankton biomasses, and the effect of different photosynthetic quotients are also examined (Table 3) Finally, three different percentages of incident light (I0 ) are used to simulate cloud effects (100%, 75%, and 50%) for some of the simulations.

PAGE 51

37 RESULTS Grazer Control The zooplankton biomass in July 1985 prevents the diatom bloom from forming. The diatoms are grazed from their maximum biomass at Bering Strait to their refuge population within nine simulated days; during this time the surface oxygen saturation is <100% dropping below 80% during the heaviest grazing, such that atmospheric oxygen enters the sea (Table 3). If the August 1988 zooplankton data around Bering Strait are used instead, higher chlorophyll stocks and associated oxygen supersaturation occur (Table 3), as observed in 1968-1969 (Figure 6}. The maximum 10-fold difference in zooplankton biomass between the two data sets of the northern Bering Sea may be simply the result of interannual variability, spatial patchiness, or instead, reflect seasonal behavior of the largest copepods. These emigre copepods of the southern Bering Sea are ontogenetic migrators. After feeding in the water column of the outer shelf before July, they may have returned to the deep sea, such that they are no longer abundant during a pre-August drift to Bering strait. The 1988 zooplankton data were collected concurrently with the nutrient

PAGE 52

38 Table 3. Resulting maximum of chlorophyll (mg chl m2 ) and o2 flux, magnitude (103 moles o2 per 60 days), and direction (into or out of the water) from different model parameters. west east [Chl. lmax. 02 flux [Chl. lmax. 02 flux baseline (Table 1) 840 26.7(out) 26 19.0(in) July 1985 zoopl. 465 36.1(in) 17 18.5(in) ice edge at 69N 840 2.7(out) 26 10.9(in) kz = 125 (cm2 sec-1 ) 910 26.9(out) 26 19.6(in) kz = 5 (cm2 sec-1 ) 675 26.1(out) 26 17.2(in) benthic regen. 50% 840 17.1(out) 26 19.2 (in) benthic regen. 15% 840 8.9(out) 26 18.9 (in) N0 3 P.Q. = 1.0 840 16.8(out) 26 19.1(in) N0 3 east = N0 3 west 840 26.7(out) 175 20.9(out) temp +5 c 850 27.6(out) 26 18. 6(in)

PAGE 53

39 and chlorophyll data used to initialize the model {Table 1) and will therefore be used in the remaining simulations. Light Regulation With the mean ice edge at 72 N {Figure 5) in the baseline case {Table 3), the simulated diatom biomass on the western side of the Chukchi Sea reaches a maximum of -840 mg Chl m-2 after 5 days {Figure 13a). From Figure 9, it can be seen that the velocity through Bering Strait may be five times greater than the average velocity for the entire region. If the first two days of phytoplankton transit are assumed to be influenced by a 50 km day-1 flow field, which then slows to the average flow field {10 km day-1 ) for the remaining model time, the diatom chlorophyll peak occurs -140 km north of Bering Strait. This maximum biomass is similar in magnitude and location to observations of chlorophyll stocks in August 1988 {Figure 14a). In this region, the simulated (Figure 13b) and observed {Figure 14b) rates of photosynthesis are both 2-3 g C m-2 day-1 As a consequence of primary production in the west side of the Chukchi Sea, the simulated oxygen saturation in the surface water reaches -130% within the no cloud case {Figure 13c). Similar oxygen supersaturations of 134% were found in August 1922 (Figure 7), and 165% in August 1968 (Figure 6a). After about 30 days, surface saturation of oxygen in the model

PAGE 54

... I (.) 1000 (a) Diat o m Biomass (We s t "' E 400 0 200 'E (.) a 5 1 o 1 5 20 25 30 J5 40 45 50 55 so 6 I 1 doy I 100 150 200 250 JOO J50 400 450 500 550 600 650 km (b) dail 0 5 10 15 20 25 30 J5 4 0 4!> 50 55 6 0 I 1 doy I I 00 150 200 250 JOO J50 400 450 500 550 &00 650 km 90 0 5 1 0 15 20 25 30 J5 40 45 5 0 55 60 J. doy l 100 150 200 250 JOO 350 400 450 500 f>50 600 650 ... 20 I (.) 15 "' E 1 0 5 Biomass (Eas t 0 6 0 6 0 5 1 0 1 5 20 25 J O 3 5 4 0 4 5 50 5 5 6 01 l doy j 100 150 200 250 JOO 350 400 450 500 550 oOO 650 km (e) daily p r imar y produc tion ( ea st) E u 0 .4 c a 2 a o 0 5 1 0 1 5 20 25 JO J5 4 0 45 50 55 6 0 130 120 d o y --' 100 150 200 250 JOO J50 400 450 500 550 600 e.:s ... k m (f) surf a c e oxygen (east) 110 100 l 9 0 f'-"' 0 5 I 0 1 5 l O 2 5 J O 35 4 0 4 5 50 5 5 6 0 L .. .. - d a y I 100 200 :)()() 350 oo 450 soo soo 650 Figure 1 3 Baseline model results ,f:o. 0

PAGE 55

Figure 14. 41 7')"' ............ CHLOROPHYLL,(mqm-2) AUGUST '988 (a) C huk c hi Sea B eri n g S e a (b) Distribution of (a) integrated chlorophyll, mg m-2, and (b) primary production, mg C m-2 day-1 in the Chukchi Sea during August, 1988 (from Korsak, 1992, and Robie et al., 1992).

PAGE 56

42 drops below 100% at -380 km north of Bering Strait (Figure 13c). Changing the percent of total incident radiation, between 100% and 50%, to simulate clouds, had little effect on the model's phytoplankton. In contrast, when ice cover is applied as far south as -69 N (Table 3}, i.e. at the maximum ice edge (Figure 5), the primary production of the diatoms drops to zero under the ice (Figure 15b). Their biomass then declines, as a result of both sinking and grazing (Figure 15a) and the oxygen drops to 95% saturation due to continued respiration (Figure 15c) Note that Sverdrup (1929} found similar values in August 1924 of 95% in the surface water of the East Siberian Sea (Figure 7) During January 1967 at Fletcher's Ice Island, T-3, in the Arctic Ocean (Figure 5), the partial pressure of carbon dioxide was -10-90 greater in the water than in the air (Kelley, 1968}, reflecting continued respiration under the ice as well. In contrast, undersaturated conditions of -200 were found in surface waters of Anadyr Strait during AugustSeptember 1968 (Gordon et al., 1973) Similarly, Zeeman (1992} found surface concentrations of 1609 mg-at m-3 on the western side of the Chukchi Sea in August 1988, compared to 2090 mg-at m-3 on the eastern side. In the simulated eastern side of the Chukchi Sea, surface saturation values of <93% are initially modelled (Figure 13f) and reflect respiration demands which are not met by the

PAGE 57

1 ide _: o o o o o o 30 : (a) Diatom Biomass {west) Flaaellate Biomass (east 25 ::r 1 10 c D 5 0 0 0 10 15 2.0 :1!1 30 40 50 56 10 0 5 1 0 15 20 25 30 35 4 0 45 50 55 60 ._ doy (b) daitx primal)' eroduction (e) doily pdrn_ory_produ0i()n (east) 8[ j 0 o 0 0 o o o o.af J 0.6 4[\1 1 OA 0.2 0 0 5 1 0 15 20 25 30 35 4 0 4 5 50 55 60 0 5 1 0 15 20 25 30 35 40 45 50 55 60 doy day (f) 1.30' t I I 3 1 30 . --.. -. ___ ice -+ 120 1 20 c: 0 :g 110 .. 1110 i 100 0 100 .. -90 90 80 ---... .. ---. o 5 10 15 20 2s JO 35 40 45 so ao 0 5 1 0 1 5 20 25 30 35 40 4 5 50 55 60 doy doy Figure 15. Model results from case with ice edge at 69 N. w

PAGE 58

44 meager primary production of at most 0.7 g c m-2 day-1 (Figure 13e) similar to observed values (Figure 14b) In response to an influx of atmospheric oxygen (Table 3), oxygen saturations then rise to almost 100%, implying that the water of the eastern side is a co2 source to the atmosphere, or a heterotrophic trophic status for this shelf ecosystem. The flagellate phytoplankton, in the eastern Chukchi Sea case of the model, reach a peak biomass of -26 mg chl m-2 on the second day (Figure 13d), which is then steadily grazed down to 4 mg chl m-2 Ciliate biomass in August 1988 was tenfold larger on the eastern side of the Chukchi Sea (Mamaeva, 1992), presumably exploiting the flagellates, but no attempts were made to explicitly incorporate their activities in the present model. Finally, when ice cover is applied, the flagellates of the eastern region also cease growing (Figure 15e) and surface saturation values of -85% are simulated (Figure 15f) since ice cover impedes the influx of atmospheric oxygen. Nutrient Regeneration With the ice cover at 73 N, i.e. the minimum ice edge for August (Figure 5), and no clouds, the diatom bloom falls out of the model water column 5 days after its peak (Figure 16a) Near-bottom ammonium stocks of -4 mg-at m-3 are then simulated (Figure 17b), associated with near-bottom oxygen

PAGE 59

45 undersaturations of 98%, similar to observations of both ammonium (Figure 4d} and oxygen (Figure 18}. Since the flagellates do not sink (Figure 16b}, negligible ammonium was simulated near-bottom on the eastern side of the Chukchi Sea. Cases were run with the vertical mixing coefficient (kz} increased 5-fold to 125.0 cm2 sec-1 (-10 m sec-1 wind speed} and decreased to one-fifth, 5 cm-2 sec-1 (-2. 8 m sec-1 wind speed}, to affect changes in resuspension of phytoplankton biomass and nutrients. The faster mixing caused the diatom bloom to increase to 910 mg chl m-2 (Figure 19a}, with a small increase in the total oxygen efflux over the 60 days (Table 3} as a consequence of a stronger secondary maximum of primary production (Figure 19b vs. Figure 13b}. The near-bottom ammonium pool was depleted to < 1 mg-at NH4+ m-3 (Figure 17a}, which is less than the observations (Figure 4d}. With slower mixing, the diatoms reached a maximum biomass of only 675 mg chl m-2 (Figure 20a}, and outgassed less oxygen than in the other mixing cases (Table 3}. I n this third mixing case, the near-bottom ammonium pool reached concentrations of > 8 mgat NH4 + m-3 (Figure 17c}, similar to mod e l results of the southeastern Bering Sea (Walsh and Dieterle, 1993}. The nitrate field of the model was similarly affected by the changes in kz. In the baseline case, the nitrate is depleted to <0.01 mg-at m-3 by day 13 of the model (Figure 17e}. With the mixing increased, i.e. kz = 125 cm2 sec-1 the

PAGE 60

(a) 5 10 1 5 20 25 (b) 5 10 1 5 20 25 30 day 30 35 3 5 40 40 45 50 5 5 45 50 55 Figure 16. Modelled contours of phytoplankton chlorophyll in baseline case. 60 60 0'1

PAGE 61

, , 0 (d) NO (west) k =125. 0 j 1 0 5 5 !: \ : w 0 5 10 15 20 25 30 .35 40 45 50 55 60 0 20 40 60 (b) NH wes t k =25.0 0 0 1 0 10 .l: .c 120 120 0 5 1 0 15 20 25 30 .35 4 0 45 50 55 60 0 20 40 60 (c) N H .(west)k_=5.0 (f) NO.(west)k=5.0 0 10 .l: Q. 20 4 0 40 0 5 10 15 20 25 30 35 45 50 55 60 0 20 4 0 60 Figure 17. Modelled contours of ammon i um-(a-c) and nitrate (d-f) with each of the kz values used ol:>o -..J

PAGE 62

64 Figure 18. 48 178 176 174 172" 170" 168 168" 114 162" 160 158" 156" I ... HERAI.O I. 70" 178" 176 174 1 72" 170" 168 166 114 162" ISO" 1!18" l!!e oxygen saturation (%) in the near-bottom layer during midsummer cruises of the Brown Bear during 1960 and the Northwind during 1963 (from Coachman et al., 1975).

PAGE 63

49 nitrate becomes <0.01 mg-at m3 by day 7 (Figure 17d). This is due to the increased availability of nitrate to the phytoplankton. Conversely, when the mixing is reduced (kz = 5 cm2 sec-1 ) nitrate is mixed upward slower and is less available to the phytoplankton. In this case the nitrate pool is not depleted in the model (Figure 17f). The effect of the regeneration rate of ammonium by the benthic animals was also explored by using different values of 80%, 50%, and 15% (Table 3). The value of 80% was the only regeneration rate which yielded a near-bottom ammonium pool matching the 1988 observations (Figure 4d) After the diatom bloom had sunk out of the water column, the 15% regeneration rate of ammonium by the benthos made less ammonium available to the phytoplankton than did the 80% regeneration rate. After the nitrate is depleted, this source of nitrogen (ammonium) supports the phytoplankton growth and oxygen evolution for the remainder of the model time. Thus, the regeneration rate had a large effect on the total resultant oxygen flux from the water (Table 3). Table 3 also shows the resulting oxygen flux when the photosynthetic quotient (P.Q.) of nitrate is reduced from 1.3 to 1.0. The total oxygen efflux on the western side of the Chukchi Sea is reduced by approximately one-third by this change; while the influx of oxygen on the eastern side of the Chukchi sea remains approximately the same. This difference is due to the meager stocks of nitrate in the initial

PAGE 64

-' :J: 0 1vto. B iomass-West E' 4oo 200 0 5 1 0 15 20 25 30 35 40 45 50 55 60 ooy (b) 8 00 (c) cl 110 100 90 rod 5 10 15 20 25 JO 35 40 45 50 55 60 day surface 0 0 5 10 15 20 25 30 35 40 45 50 55 60 day 3C'---(d) Phvto. Biomass-East 25 -' 20 :J: 0 0' 15 E H 0 10 5 0 5 10 15 20 25 JO 35 40 45 50 55 60 day (e) 0 8 , ,erill")a!"X 0.6 E 0 0.4 00 0 2 0.0 0 5 10 15 20 25 30 35 iO 45 50 55 60 doy (f) 130 surface 0 120 110 0 ., 10 0 90v 0 5 1 0 15 20 25 30 35 40 45 50 55 60 doy Figure 19. Model results from the kz = cm2 sec-1 case. U1 0

PAGE 65

(d) (a) Phyto. Biomass-West Phvto. Biomass-East 800[ ] 30 I I 0400 0 E 10 200 5 0 0 0 5 10 15 20 25 JO 35 40 45 50 55 60 0 5 10 15 20 25 30 J5 40 45 50 55 60 day day et(b) doi rod. 0 8 (e) doily primary prod. 6 0.6 N . E E u 4 0 0.4 0.2 0 0 5 10 15 20 25 30 35 40 45 50 55 60 0 5 10 15 20 25 30 35 40 45 50 55 60 day day (c) (f) surface 0 surface 0 ' .. ' ] ::r . 1 :: 0 5 10 15 20 25 30 35 40 45 5 0 60 0 5 10 15 20 25 30 35 40 45 50 55 60 doy day Figure 20. Model results from the kz = 5.0 cm2 sec-1 case. 1..11 ....

PAGE 66

52 conditions for the eastern side. Changing the P Q. of nitrate changes the system to exclusively recycled production. This result (Table 3) demonstrates that systems with greater amounts of "new" production i.e. larger f-ratios, produce more oxygen than systems with the equivalent amount of recycled production, i e low f-ratios. However, the oxygen produced from the reduction of nitrate to ammonium by phytoplankton cannot be used in the calculation of the equivalent carbon dioxide uptake. Thus far, the model has considered simulation analysis of the existing structure of nutrient loading and temperature patterns. Two last cases were run; one in which the temperature is increased by 5 C, and one in which the nutrient influx to the eastern side of the Chukchi Sea is the same as the western side. Temperature Regulation On the western side of the Chukchi Sea, a temperature increase of 5 c yields a small increase in peak biomass, to 850 mg chl. m-2 (Table 3), as a result of increased phytoplankton growth -recall equation ( 14) The primary production reaches 8 g C m-2 day-l and the maximum surface oxygen saturation (Figure 21c) reaches 130% more rapidly than in the baseline case (Figure 13c). The flagellates, to the east, do not increase their primary production after the

PAGE 67

_, :I: 0 1000 (a) Phvto Biomass-West e 400 1 M 200 0 5 10 15 20 25 JO 35 40 45 50 55 60 8 "' 90 (b) dai rod. 5 10 15 20 25 30 35 40 45 50 55 60 (C) surface 0 0 5 10 15 20 25 30 35 40 45 50 55 60 30 2 5 _, 20 :I: 0 15 a> e 10 N I 5 0 0 0 8 0 6 E 0 0.4 "' 0 2 (d) Phvto. Biomass-East 5 I 0 15 20 25 30 35 40 45 50 55 60 day (e) doil y primary p_rod. o o c:===::::::::::::::==::::::=::o::::=====d 0 5 10 15 20 25 30 35 40 45 50 55 60 day 130E 1 1 1 surfa c e 0 I I I II 12 0 I I I I '3 (f) ..110 i M 0 5 10 15 20 25 30 35 40 45 50 55 60 day .... Figure 21. Model results from the increased temperature case. Ul w

PAGE 68

54 temperature increase, due to the oligotrophic conditions. But, there is a small change in the influx of oxygen to the eastern Chukchi Sea (Table 3) as a consequence of the reduced solubility of oxygen in warmer water-recall equation (8). This effect is obscured in the western results by the increased primary production. Nutrient Loading An even greater change of oxygen flux occurs when oligotrophic water of the eastern side of the Chukchi Sea is given the same nutrient content as the eutrophic western side. The flagellate biomass now peaks at 175 mg chl m-2 (Figure 22), compared to 26 mg chl m-2 in the baseline case and 840 mg chl m-2 for the diatoms (Table 3). The flagellates do not sink; therefore, self-shading prevents them from reaching the same biomass as the diatoms. The direction of oxygen flux across the air-sea interface now changes, such that the eastern side of the Chukchi Sea also becomes a source of oxygen to the atmosphere. Presumably, the flux of carbon dioxide also reverses, i.e. switching to an autotrophic status as in the southeastern Bering Sea (Walsh and Dieterle, 1993) and the western Chukchi Sea.

PAGE 69

(a) 1 nteqroted chi. 200 1501...J J: u 10010'1 v E 50 0 0 5 10 15 20 25 30 35 40 45 50 55 60 doy (b) doily primary prod. 10 8 ... I 6 E (.) o> "\ 2 0 0 5 1 0 15 20 25 30 35 40 45 50 55 60 doy (c) surface 0 5 1 0 15 20 25 30 :35 40 45 50 55 60 day Figure 22. Model results from the east Chukchi Sea with initial nutrient conditions equal to the west Chukchi Sea nutrients of the baseline case. 55

PAGE 70

56 CONCLUSION The one-dimensional model presented here is useful for the modelling of oxygen in the Chukchi Sea. The differences between the 1922 observations of oxygen supersaturations by Sverdrup and his observations from 1923 and 1924 do appear to be attributable to light regulation by differences in sea ice coverage. Longitudinal variations among data collected by ice-breakers in the 1960's seem instead to be nutrient controlled. The trophic status of the ecosystem depends upon the interaction of primary production and respiration. The populations of zooplankton grazers found around Bering Strait can, depending on the grazing stress they exhibit, prevent a phytoplankton bloom in the Chukchi Sea during AugustSeptember. This causes the trophic status to change from autotrophic to heterotrophic. Also, the trophic status of the ecosystem depends upon the nutrient loading; the trophic status of the eastern Chukchi Sea was reversed by changing the nutrient loading. By including the biological and chemical components of this model in a fully three-dimensional physical model, with introduction of a carbon dioxide component (similar to the oxygen component of this model) a better estimate of the

PAGE 71

56 trophic (autotrophic or heterotrophic) status of this continental shelf ecosystem may be obtained on an annual basis. within Questions of storage of atmospheric carbon dioxide labile pools of DOC (Walsh, 1993) could also be considered in a more complex model. Finally, increased ecological resolution to multiple functional groups of phytoplankton, zooplankton, and bacterioplankton would provide biological realism to future systems analysis of the Chukchi Sea.

PAGE 72

57 REFERENCES Coachman, L.K., K. Aagaard, and R.B. Tripp. (1975) Bering Strait: The Regional Physical Oceanography. University of Washington Press, Seattle. Coachman, L.K. and v.v. Shigaev. (1992) Northern Bering -Chukchi Sea Ecosystem: The Physical Basis. Results of the Third Joint US-USSR Bering and Chukchi Seas Expedition CBERPAC) Summer 1988. P. A. Nagel ( ed.) us Fish and Wildlife Service, Washington, DC. Codispoti, L.A., G.E. Friederich, C.M. Sakamoto, and L.I. Gordon. (1991). Nutrient Cycling and Primary Production in the Marine Systems of the Arctic and Antarctic. Journal of Marine Systems, 2: 359-384. Codispoti, L.A., and F.A. Richards. (1968). Micronutrient Distributions in the East Siberian and Laptev Seas During Summer 1963. Arctic, 21: 67-83. Csanady, G.T. (1976). Mean-circulation in shallow seas. Journal of Geophysical Research, 81: 5389-5399. Dagg, M.J., J. Vidal, T.E. Whitledge, R.L. Iverson, and J.J. Goering. (1982). The Feeding, Respiration, and Excretion of Zooplankton in the Bering Sea During a Spring Bloom. Deep-Sea Research 29: 45-64. Eppley, R.W., J.N. Rogers, and J .J. McCarthy. (1969). Half saturation constants for the uptake of nitrate and ammonium by marine phytoplankton. Limnology and Oceanography, 14: 912-920. Eppley, R.W. (1972). Temperature and phytoplankton growth in the sea. Fisheries Bulletin, 70: 1063-1085. Falkowski, P.G., and T.G. owens. (1978). Effects of light intensity on photosynthesis and dark respiration in six species of marine phytoplankton. Marine Biology {Berlin) 45: 289-295. Fanning, K.A. and L.M. Torres. (1991). 222Rn and 226Ra: Indicators of sea-ice effects on air-sea gas exchange. Polar Research, 10: 51-58.

PAGE 73

58 Gordon, L.I., P.K. Park, J.J. Kelley, and D.W. Hood. (1973). Carbon Dioxide Partial Pressures in North Pacific surface Waters. 2. General Late Summer Distribution. Marine Chemistry. 1: 191-198. Grebmeier, J.M. (1992). Benthic Processes on the Shallow Continental Shelf. Results of the Third Joint US-USSR Bering and Chukchi Seas Expedition (BERPAC} Summer 1988. P.A. Nagel (ed.). US Fish and Wildlife Service, Washington, DC. Kelley, J.J. (1968). Carbon Dioxide in the Surface Waters of the North Atlantic Ocean and the Barents and Kara Seas. Limnology and Oceanography, 15: 80-87. Kinney, P.J., D.C. Burrell, M.E. Arhelger, T.C. Loder, and D.W. Hood. (1970). Chukchi Sea Data Report. Institute of Marine Science, University of Alaska. Korsak, M.N. (1992). Primary Production of Organic Matter. Results of the Third Joint US-USSR Bering and Chukchi Seas Expedition CBERPAC}, Summer 1988. P.A. Nagel (ed.). US Fish and Wildlife Service, Washington, DC. Kristiansen, s. (1983). Urea as a nitrogen source for the phytoplankton in the Oslofjord. Marine Biology, 74. 17-24. Kulikov, A.S. (1992). Characteristics oc Zooplankton Communities. Results of the Third Joint US-USSR Bering and Chukchi Seas Expedition (BERPAC}, Summer 1988. P.A. Nagel (ed.). US Fish and Wildlife Service, Washington, DC. Mamaeva, N.V. (1992). Ciliate Protozoa in Plankton. Results of the Third Joint US-USSR Bering and Chukchi Seas Expedition (BERPAC}, Summer 1988. P.A. Nagel (ed.). US Fish and Wildlife Service, Washington, DC. McCarthy, J.J., W.R. Taylor, and J.L. Taft. (1977). Nitrogenous nutrition of the plankton in the Chesapeake Bay. 1. Nutrient availability and phytoplankton preferences. Limnology and Oceanography, 22: 996-1011. Meyers, M.B. (1993). The Response of Oceanic Phytoplankton to Nitrate Flux in the Eastern Gulf of Mexico: A Simulation Analysis. Doctoral dissertation, University of South Florida.

PAGE 74

59 Nagle, C.M. (1978). Climatology of Brookhaven National Laboratory, 1974 through 1977, BNL-50857 UC-11, Environmental Control Technology Earth Science, T10-4500. Brookhaven National Laboratory, Upton, New York. Naval Oceanography Command Detachment, Asheville. (1986). Sea Ice Climactic Atlas: Volume III Arctic West. Niebauer, H.J. and W.O. smith Jr. (1989). A Numerical Model of Mesoscale Physical-Biological Interactions in the Fram Strait Marginal Ice Zone. Journal of Geophysical Research, 94: 16,151-16,175. O'Brien, J.J. (1986). Advanced Physical Numerical Modelling. Reidel Publishing Company, Dordrecht. Overland, J.E., and A.T. Roach. (1987). Northward Flow in the Bering and Chukchi Seas. Journal of Geophysical Research, 92: 7097-7105. Packard, T.T., and D. Blasco. (1974). Activity in Upwelling Regions. II Dependence. Tethys, 6: 269-280. Nitrate Ammonia Reductase and Light Parsons, T.R., M. Takahashi, and B.T. Hargrave. (1984). Biological Oceanographic Processes. Pergamon, Oxford. Peng, T.H., T. Takahashi, W.S. Broecker, and J. Olafsson. ( 1987) Seasonal variability of carbon dioxide, nutrients, and oxygen in the northern North Atlantic surface water: observations and a model. Tellus, 39 B: 439-458. Richards, F.A. (1965). Dissolved Gasses Other Than Carbon Dioxide. in Chemical Oceanography, Riley and Skirrow, Academic Press, New York 1965. pp 197-322. Riley, J.P. and R. Chester. (1971). Introduction to Marine Chemistry. Academic Press, San Diego. Robie, w.s., C.P. McRoy, and A.M. Springer. (1992). Phytoplankton Biomass Distribution in the Northern Bering Sea and Southern Chukchi Sea. Results of the Third Joint US-USSR Bering and Chukchi Seas Expedition (BERPAC) Summer 1988. P. A. Nagel ( ed.) US Fish and Wildlife service, Washington, DC. Shuert, P.G. (1990). Ecosystem analysis of the Bering/ Chukchi seas using a coupled time-dependent physical/ biological simulation model. Doctoral dissertation. University of South Florida.

PAGE 75

60 Shuert, P.G. and J.J. Walsh. (1992). A Time-dependent Depthintegrated Barotropic Physical Model of the Bering/ Chukchi Seas for use in Ecosystem Analysis. Journal of Marine systems, 3: 141-161. Shuert, P.G. and J .J. Walsh. (1993}. A coupled physicalbiological model of the Bering-Chukchi Seas. Continental Shelf Research, 13: 543-573. Smith, R.C. and K.S. Baker. (1982}. Oceanic Chlorophyll Concentrations as Determined Using Coastal Zone Color Scanner Imagery. Marine Biology (Berlin), 66: 269-279. Smolarkiewicz, P.K. (1983). A Simple Positive Definite Advection Scheme with small Implicit Diffusion. Monthly Weather Review, American Meteorological Society, Vol. 11(March): 479-486. Sverdrup, H.U. (1929}. The Waters on the North-Siberian Shelf. Geofysisk Institute; Bergen. Taylor, A.H., A.J. Watson, M. Ainsworth, J.E. Robertson, and D.R. Turner. (1991). A Modelling Investigation of the Role of Phytoplankton in the Balance of Carbon at the Surface of the North Atlantic. Global Biogeochemical Cycles, 5: 151-171. Vidal, J. and S.L. Smith. (1986). Biomass, Growth, and Development of Populations of Herbivorous Zooplankton in the Southeastern Bering Sea During Summer. Deep-Sea Research, 33: 523-556. Walsh, J.J. (1975). A spatial simulation model of the Peru upwelling ecosystem. Deep-Sea Research, 22: 201-236. Walsh, J.J. (1988}. On the Nature of Continental Shelves. Academic Press. Walsh, J.J. (1989). Arctic Carbon Sinks: present and future. Global Biogeochemical Cycles, 3: 393-411. Walsh, J.J. (1993} Observations and Hypotheses of DOC storage in Polar Seas. Journal of Geophysical Research (submitted). Walsh, J.J., C.P. McRoy, L.K. Coachman, J.J. Goering, J.J. Nihoul, T.E. Whitledge, T.H. Blackburn, P.L. Parker, C.D. Wirick, P.G. Shuert, J.M. Grebmeier, A.M. Springer, R.D. Tripp, D.A. Hansell, s. Djenidi, E. Deleersnijder, K. Henriksen, B.A. Lund, P. Andersen, F.E. Muller-Karger, and K. Dean. (1989). Carbon and nitrogen cycling within the Bering/Chukchi Seas: source regions of organic matter

PAGE 76

61 effecting AOU demands of the Arctic Ocean. Progress in Oceanography, 22: 279-361. Walsh, J.J. and D.A. Dieterle. (1993). co2 cycling in the Coastal Ocean. I A Numerical Analysis of the Southeastern Bering Sea with applications to the Chukchi Sea and the Northern Gulf of Mexico. Progress in Oceanography (submitted). Weiss, R.F. (1970). The solubility of nitrogen, oxygen, and Argon in water and seawater. Deep Sea Reseach, 17: 203-215. Whitledge, T.E. (1988). ISHTAR Cruise Report, Leg 3 R/V Akademik Korolev 26 July -2 September 1988. Whitledge, T.E., M .I. Gorelkin, and S.M. Chernyak. (1992). Biogenic Nutrient Content. Results of the Third Joint USUSSR Bering and Chukchi Seas Expedition (BERPAC), Summer 1988. P.A. Nagel (ed.). us Fish and Wildlife Service, Washington, DC. Williams, P .J.leB. and J.E. Robertson. (1991). overall planktonic oxygen and carbon dioxide metabolisms: the problem of reconciling observations and calculations of photosynthetic quotients. Journal of Plankton Research, 13 suppliment: 153-169. Wroblewski, J. s. ( 1977) A model of phytoplankton plume formation during variable Oregon upwelling. Journal of Marine R esearch, 35: 357-394. Zeeman, S.I. (1992). The Importanc e of Primary Production and co2 Results of the Third Joint US-U SSR Bering and Chukchi Seas Expedition (BERPAC), Summer 1988. P.A. Nagel (ed.). US Fish and Wildlife Service, Washington, DC.

PAGE 77

62 APPENDIX

PAGE 78

63 APPENDIX -Numerical methods The numerical methods, used to obtain solutions for the partial differential equations, are discussed in this section. The diffusion equations in the model, of the form: au= k CPu ot oz2 where k is a positive constant, are solved using the finite difference scheme: where n is the time step and j is the depth inteval. This is an explicit scheme where no terms on the right require knowledge of the n+1 (or any future) time step. This scheme is stable, i.e. the solution remains bounded as n goes to infinity (O'Brien, 1986), if kz At/Az2 1/2 as At and Az go to zero. The sinking term in the phytoplankton equation (equation 13) has the form: .!=-_j_ (uliJ) at az where u represents the sinking rate and is constant. Therefore the equation becomes:

PAGE 79

64 An upstream differencing approach to solving this equation numerically has the form: ,,,n ,1,n-1 U _!!(,1,n-1_ ,1,n-1} 't'k+1 = 't'k+1 .6-z 't'k+1 't'k where k increases with increasing depth. Thus the value of psi {V) in a given depth box at a given time step depends upon the value in that box at the previous time step, plus an amount (u tl.tjtl.z vkn-1 ) that sinks into the box from above, minus an amount (u tl.tjtl.z vk+1n-1 ) that sinks out of the box during the time step. However, this scheme introduces numerical diffusion into the solution. By expanding the finite difference form of the equation in a Taylor series, it can be shown that this scheme is actually equivalent to solving the equation: -u __g__(kat az az J.mpl az The second term on the right (ajaz (kimpl av;az)) represents the numerical diffusion and kimpl = o. 5 (u tl.z !:it u2 ) Smolarkiewicz (1983) defines the diffusive velocity as: and then computes an "anti-diffusive" velocity The starred variables represent values computed by an upstream differencing step. Epsilon () is a small number {lo-15) used

PAGE 80

65 to prevent dividing by zero. The anti-diffusive velocity {u) is then used in a second upstream step in place of u to correct for the numerical diffusion of the first step. As dt and dz go to zero this entire scheme reduces to the simple upstream scheme. The scheme is stable as long as the simple upstream stability condition is met (since both steps are upstream schemes). The stability condition states that u dt/dz must be less than or equal to one. This means that the sinking cannot proceed further than one box during one time step. There also exists numerical diffusion in the second step of the scheme because it too 'is an upstream step. This is corrected by increasing the anti-diffusive velocity by the Smolarkiewicz constant (Sc) which is estimated experimentally and lies between one and one point zero eight: Sc i1 ( 1 Sc 1. 0 8)