All physical components Fulvestrant such as velocities, salinity and temperature were calculated in the 3D hydrodynamic model.

The output from this model as an average value for the period 1960–2000 (ECOOP IP WP 10.1.1) at temporal and special vertical scales for three areas (Gdańsk Deep, Bornholm Deep, Gotland Deep) was linearly interpolated at every time and vertical step of the 1D POC model. The 3D model was forced using daily-averaged reanalysis and operational atmospheric data (ERA-40) obtained from the European Centre for Medium-range Weather Forecasts (ECMWF). The 1D POC model is a one-dimensional biogeochemical model. It has a high vertical resolution with a vertical grid of 1 m, which is constant throughout the water column. This means that the check details model calculates the vertical profiles of all its variables and assumes that they are horizontally homogeneous in the sub-basins. In comparison with vertical changes, the dynamic characteristics remain almost unchanged in a horizontal plane. Hence, the magnitudes of the lateral

import/export are lower, and the above assumption can be made. The horizontal velocity components (v, u) obtained in the ECOOP IP project WP 10.1.1 model for the Baltic Sea (ECOOP IP project WP 10.1.1) were averaged and used to calculate hydrodynamic variables such as w, Kz, S and T. In order to include horizontal variations in the southern Baltic (a larger area) it was divided into three sub-basins – 1 – Bornholm Deep (BD), 2 – Gdańsk Deep (GdD) and 3 – Gotland Deep (GtD) – each of which has 64 pixels; 1 pixel = 9 × 9 km2. The main average circulation of the Baltic Sea is called the Baltic haline conveyor belt (BCB, Doos et al. 2004, Meier 2006). If we take BCB into account, the main flow though the sub-basins Amobarbital is assumed to be part of BCB, and other flows can be neglected. The horizontal transport of the variables Nutr, Phyt, Zoop and DetrP between sub-basins is treated as a typical advection process. For each time step the POC concentration is determined as the sum of phytoplankton, zooplankton and pelagic detritus concentrations. The model does not include the inflow

of nutrient compounds from rivers or the atmosphere. Hence, the 1D POC model has zero boundary conditions (from the land and atmosphere). It was assumed that the initial conditions of the numerical simulations were the average winter values from the previous 4 decades and that the final states of one year would be the starting points of the next year. It was further assumed for GdD that since there were few phytoplankton values for January and December, a constant value of Phyt0 = 10 mgC m−3 ( Witek 1995) could be applied. Owing to the long simulation period (from January) preceding the spring bloom (April/May) the model is not sensitive to the initial phytoplankton concentration. The initial zooplankton biomass was calculated on the basis of data from Witek (1995) as Zoop0 = 1 mgC m−3.