Simulated and AVHRR-derived SST

Simulated and AVHRR-derived SST

SEASONAL VERSUS SYNOPTIC V ARIABILITY IN PLANKTONIC PRODUCTION IN A IDGH-LATITUDE MARGINAL

SEA: THE GULF OF ST. LAWRENCE (CANADA)

ABSTRACT

The Gulf of St. Lawrence (Canada) is a subarctic marginal sea characterized by highly variable hydrodynamic conditions that generate a spatial heterogeneity in the marine production. A better understanding of physical-biologicallinkages is needed to improve our ability to evaluate the effects of climate variability and change on the gulf’ s planktonicproduction. We develop a three-dimensionnal (3-D) eddy-permitting resolution physicalbiological coupled model of plankton dynamics in the Gulf of St. Lawrence. The planktonic ecosystem model accounts for the competition between simplified herbivorous and microbial food webs that characterize bloom and post-bloom conditions, respectively, as generally observed in temperate and subarctic coastal waters. It is driven by a fully prognostic 3-D sea ice-ocean model with realistic tidaI, atmospheric and hydrological forcing. The simulation shows a consistent seasonal primary production cycle, and
highlights the importance of local sea ice dynamics for the timing of the vernal bloom and the strong influence of the mesoscale circulation on planktonic production patterns at subregional scales.

INTRODUCTION

General circulation models generally predict that global climate change associated with increased greenhouse gas concentrations in the atmosphere will lead to an amplified warming in the Arctic and its adjacent seas over the next century (5-8°C in 2070; e.g., Hol/and and Bitz, 2003). Among those, the Gulf of St. Lawrence (GSL) is a large semienclosed sea of 226000 km2 that connects the Great Lakes and the St. Lawrence River with the North Atlantic Ocean [e.g., Koutitonsky and Budgen, 1991]. Runoff from the St. Lawrence watershed is the second most important source of freshwater from North America into the North Atlantic Ocean [e.g., Bourgault and Koutitonsky, 1999]. The GSL exhibits a subarctic climate with a seasonal sea ice cover present between J anuary andApril, and sheds the southernmost extent of sea ice in the northern hemisphere. Freshwater runoff, large to moderate tides, and highly synoptic winds drive the gulfs circulation. These physical forcing, coupled with the relatively large dimensions of the gulf (several internal Rossby deformation radii) and an average depth of 150 m, generate a complex hydrodynamics with eddies, coastal upwellings, and fronts superimposed on a mean estuarine-like circulation [e.g., Koutitonsky and Bugden, 1991; Saucier et al., 2003]. These hydrodynamic conditions have been shown to have a marked effect on summer primaryproduction in the northwestern Gulf [Levasseur et al., 1992; Fuentes-Yaco et al., 1995, 1996, 1997ab; Tremblay et al., 1997], and are thought to generate a spatial heterogeneity inthe marine production of the GSL [e.g., de Lafontaine et al., 1991]. Savenkoff et al. [2001] also suggest that the GSL can be subdivided into distinct subregions on the basis of specific  hydrodynamic regimes that affect the nutrient transport and the resulting planktonic production. Recent observations confmn that the high interannual variability in plankton biomass in the Lower Estuary [Starr and Harvey, 2000; Starr, 2001], the recruitment of fish stocks in the southem gulf [Runge et al., 1999], the aggregation of krill and whales at the head of the Laurentian Channel [Simard and Lavoie, 1999; Lavoie et al., 2000], and the water masses properties of the GSL [Saucier et al., 2003] are strongly linked to the influence of climate and freshwater inputs on the mixing and circulation processes. However, it has not yet been possible to quantify together the detailed circulation and the response of the planktonic ecosystem. Prior to any attempt to predict the effects of global climate variability and changes on the GSL system, we must first acquire a better knowledge of the links between the physicalenvironment and the short-term to interannual variations in planktonic production. In order to improve our capability to predict these responses, we need to develop models that reproduce the spatio-temporal variability of the primary and secondary production cycles. Modelling of planktonic production in the St. Lawrence marine system has been limited to one-dimensionnal (l-D) models of the carbon cycle in the northwestem [Tian et al., 2000] and northeastem [Tian et al., 2001] GSL, to a 2-D modelling study of the phytoplankton production in the Lower Estuary [Zakardjian et al., 2000], and to 3-D modelling of copepods population dynamics [Zakardjian et al., 2003]. This paper aims at describing and quantifying the circulation-planktonic production coupling in the GSL using a detailed 3-D physical-biological model. The coupled model includes both simplified herbivorous and microbial food webs typical of bloom and post-bloom conditions respectively, as generally  observed in temperate and subarctic coastal waters. It is driven by a 3-D high resolution primitive equations ocean-sea ice regional model [Saucier et al., 2003] with realistic tidal, atmospheric and hydrologic forcing. In the present paper, we focus on the ecological robustness of the coupled model performances al the regional scale and the subregional variability of the seasonal plankton cycle in response to varied hydrodynamic conditions. These first results demonstrate that the coupled model predicts realistic levels of biomass and a seasonal cycle of planktonic production dominated by the spring phytoplankton bloom, as observed in the GSL. In addition, the model generates a large synoptic and spatial variability in planktonic production in response to the buoyancy-driven circulation, tidal mixing, and wind events. As a consequence, primary production can locally be as important as during the spring bloom.

MODEL FORMULATION

The 3-D regional circulation model

A detailed description of the deterministic sea ice-ocean coupled model is presented in Saucier et al. [2003], and the characteristics are briefly reviewed here. The ocean model is govemed by the shallow water equations solved by a finite difference scheme. It incorporates a level2.5 turbulent kinetic energy equation [Mellor and Yamada, 1974, 1982] and diagnostic master length scales. A thermodynamic and dynamic sea ice model [Semtner, 1976; Flato, 1993] is coupled with the ocean mode!. Bulk aerodynamic exchange formulas govem the heat and momentum fluxes between the ocean, sea ice and atmosphere. The model domain covers the Estuary and the Gulf of St. Lawrence and is delimited by three open boundaries at the Cabot Strait, the Strait of Belle-Isle, and the upper limit of the tidal influence near Montreal (Figure ll-l). The grid resolution is 5 km on the horizontal and ranges from 5 to 20 m in the vertical, with free surface and bottom layers adjusted to
topography. The model is fully deterministic and tracer conserving [e.g., Saucier et al., 2003], driven by a detailed atmospheric forcing (three-hourly winds, light, precipitation), daily river ronoff data from the St. Lawrence river and the 28 most important tributaries, hourly water levels (co-oscillating tides) and monthly mean temperature and salinity profiles at the Strait of Belle-Isle and Cabot Strait. The model accounts for the variations of sea ice, tides, momentum, heat and salt fluxes, and river discharges with a time step of 300 s and reproduces the high frequency to interannual variations of the circulation, water mass properties, and sea ice coyer. Simulations for 1996-1997 [Saucier et al. , 2003] and recently
domain. Boxes indicate the studied subregions: Lower Estuary of St. Lawrence (LSLE), northwestem Gulf of St. Lawrence (NWG) , Unguedo Strait (USt), Magdalen Shallow (MS), southem Laurentian Channel (SLC), northeastem Gulf of St. Lawrence (NEG) and Jacques Cartier Strait (JCS).

Table des matières

REMERCIEMENTS
RÉsUMÉ
LISTE DES FIGURES
LISTE DES TABLEAUX
1. INTRODUCTION GÉNÉRALE 
n. SEASONAL VERSUS SYNOPTIC V ARIABILITY IN PLANKTONIC PRODUCTION IN A
HIGH-LATITUDE MARGINAL SEA: THE GULF OF ST. LAWRENCE (CANADA)
ABSTRACT
INTRODUCTION 
MODEL FORMULATION
J. The 3-D regional circulation mode/
2. The planktonic ecosystem mode/
3. Coupling with the 3-D regional circulation model
RESULTS
J. Mean seasonal biomass and production cycle
2. Sub-regional differences in planktonic production
DISCUSSION CONCLUSIONS
APPENDIX. PLANKTONIC ECOSYSTEM MODEL DESCRIPTION
m. APPLICATION OF REMOTELY SENSED SEA COLOR AND TEMPERATURE DATA
FOR MES OS CALE AND REGIONAL VALIDATION OF A 3-D HIGH-RESOLUTION
BIOLOGICAL-PHYSICAL COUPLED MODEL OF THE GULF OF ST. LAWRENCE
(CANADA): THE CHALLENGE OF INLAND W A TERS
ABSTRACT
INTRODUCTION 
METHODS
J. The 3-D physical-planktonic ecosystem model
2. Remotely sensed and in situ data
RESULTS AND DISCUSSION
J. Simulated and AVHRR-derived SST
2. Turbidity, simulated and SeaWIFS-derived surface Chi a ..
3. The CDOM hypothesis
4. Tracking the estuarine circulation and the associated mesoscale activity
CONCLUSIONS
N. THE IMPACI’ OF FRESHWATER-ASSOCIATED TURBIDITY ON PHYTOPLANKTON
IN A HIGHL Y DYNAMIC SHELF SEA: A MODELLING STUDY IN THE GULF OF ST. LAWRENCE (CANADA)
ABSTRACI’
INTRODUCI’ION
THE COUPLED MODEL
RESULTS
DISCUSSION
CONCLUSIONS
V. CONCLUSION GÉNÉRALE
VI. RÉFÉRENCES

Télécharger le rapport complet