To the Formation of Ichthyoplankton Assemblages Along the

To the Formation of Ichthyoplankton Assemblages Along the Eastern English Channel French Coast: Numerical Approach Alexei Senchev Konstantin Korotenko UMR 8013 “ELICO” Université du Littoral - Côte d'Opale Wimereux, France alexei. sentchev@mren 2. univlittoral. fr P. P. Shirshov Institute of Oceanology Moscow, Russia kkoroten@hotmail. com 1

Acknowledgement A. Grioche, X. Harlay, P. Koubbi UMR 8013 “ELICO” Ichtyoécologie Marine Université du Littoral - Côte d'Opale Wimereux, France 2

Outlin e: n n n n Motivation Biological and hydrological situations Problem definition Model description Numerical experiments Particle and larvae migration Conclusions 3

Motivation Environmental problems caused by the increase of pollutant discharge into natural waters require complex studies of physical processes of mixing and dilution, biological and mathematical modeling, data acquisition via remote sensing and in situ measurements. In this context, development of a multi-functional hydrodynamic/transport model which allows to perform fully prognostic computations of coastal water circulation and it application to environmental problems is an important task. 4

Region • Strong tidal and storm activity • Complex bottom topography • Important river discharge Zones: • Offshore • Near-shore • Dover Strait • Hydrological front 5

Hydrological situation Surface salinity April 11 -13, 1995 Mai 2 -5, 1995 Temperature profiles (April, 11 -13) Salinity 6

Biological situation Pleuronectes flesus larvae distribution for each development stage and survey (Ind. / 100 m 3. Normalized) (Grioche, Koubbi, Sautour, 1997) Initial conditions: Eggs (April, 11 -13) Larvae stage 3 (two weeks later) Larvae stage 2 (one week later) Larvae stage 3 (tree weeks later: Mai, 3 -6) 7

Biological situation Pleuronectes flesus larval transfer from the spawning grounds to the nurseries 8

Problem How strong is the influence of hydrodynamics on the larvae migration? Can we identify relationships between tidal motions, wind forcing and P. flesus larvae drift? Method Numerical modeling 9

Tracer 10

Approach Combined use of n 2 D finite element tidal model n 3 D Princeton Ocean Model n Particle transport model 11

Particle transport model (Korotenko, JMS, 1999) Input: velocity diffusivity coefficients water density Method: random walk in the horizontal random buoyancy in the vertical Output: 3 D particle displacement 12

Circulation model n Princeton Ocean Model (POM) Blumberg, Melor (1977 -87), Mellor (1996) n n n 3 -D Primitive equation, time-dependent Sigma coordinate Prognostic temperature and salinity fields Free surface k-kl turbulent closure scheme 2 2 km regular grid Real bottom topography Forcing Tidal forcing Wind forcing Fresh water discharge 13

Data assimilation Tidal Model Finite-Element Tidal Model (Le Provost, Poncet, IJNME, 1978) n 2 D n Spectral n Barotropic shallow water equation n Quadratic parameterisation of bottom friction n Depth depended grid ranging from 0. 5 to 5 km n Real bottom topography n Tidal forcing at the open boundaries n Numerical solution for individual tidal constituents Assimilation technique Augmented Lagrangian function method (Sentchev, Yaremchuk, CSR, 1999) Observations: 13 sea level observations 1039 tidal current velocity ellipses Control variables: 65 boundary conditions Results: Optimized boundary conditions for individual tidal constituents 14

Numerical experiments Surface salinity Salinity at 5 m Sea surface elevation and surface velocity after 10 days run 15

Features of Tidal Turbulence 1. Bottom-friction-generated; 2. Spatial and temporal inhomogenuity; 3. Strong horizontal intermittency along the French coast; 4. Suppressed in river discharging areas; 16

Concentration Evolution Spatial distribution of particles as a function of time (0, 1, 2, 3 weeks ) expressed in terms of concentration (nb. of particles in a unit column) The implemented complex numerical approach allowed predicting larvae assemblage in the Eastern English Channel 17

Conclusions I. A use of the coupled flow/transport modeling is a powerful tool to study tracer dispersion. II. Numerical experiments conducted in the eastern English Channel revealed the following features of tracer dynamics: The joint effect of tidal motions and river discharge gives rise to local concentrations of particles in the frontal zone. Particles - initially homogeneously distributed. Turbulent diffusion is the second factor contributing to particle concentration in the vicinity of the front. In areas with low turbulence, suppressed by fresh water discharge, vertical mixing is considerably reduced. Particles continue to move in the upper layer and are blocked by the fresh water discharge in their movement toward the French coast. Effect of the bottom topography on the tracer dynamics has been recognized. III. Experiments with particles representing P. flesus larvae provided the following results: The coupled flow/transport model reproduced the major features of the larvae migration under the influence of different forcing terms. Combined effect of tidal motions (M 2 constituent) and river discharge creates favorable conditions for larvae drift toward the French coast. Introducing of the mean sea level and more tidal constituents generates a steady larvae drift to the North along the 18 coast.