Numerical Experiments to Identify Suspended Matter Flow Rate over the Seabed in a Model of Impurity Transport

The paper deals with variation assimilation of model data on the concentration of suspended matter in the upper layer of the Sea of Azov. Such information is used during practical evaluation of identification algorithms to test the assimilation of concentration values derived from satellite information. Combined use of surface concentration estimates and modeling results based on the transport model is of interest for determining the strength of sources of suspended matter inflow. The test problem has been solved of determination of the required parameter in the sea bottom boundary condition when parameterizing the sediment inflow (agitation) from bottom sediments due to dynamic processes in the sea bottom layer. Two approaches to search for the required constant for the parameterization used in the calculations are implemented. A variational identification algorithm based on adjoint problem solving is used in determining the spatially variable flow of suspended matter on the seabed. The assimilation of measurement data into a model of passive admixture transport allows to determine the spatial structure of such flows at a given time interval. When implementing the variational identification algorithm, gradient methods are used to find optimal estimates by minimizing the quadratic functional of the prognosis quality. The solution of the adjoint problem is used to construct the gradient of the prognosis quality functional. Descent is performed in the direction of this gradient. During realization of variational procedure the main problem, the adjoint problem and the problem in variations, which is necessary to determine an iteration parameter, are solved. The flow fields and turbulent diffusion coefficients used in the calculations were obtained using a dynamic model of the Sea of Azov in sigma coordinates under exposure to intense easterly wind.