статья в журнале

En

Sovremennye Problemy Distantsionnogo Zondirovaniya Zemli iz Kosmosa

ISSN:20707401

2023

10.21046/2070-7401-2023-20-1-76-83

NJOTWY

One of the main problems in using satellite data on the spatial structure of impurities in the upper layer of the sea is interference caused by various reasons, for example, by clouds or various aerosols, combustion products, etc. Often such interference can interrupt the receipt of data on the sea surface state. Therefore, the task of determining concentration values for areas with missing information (gap filling) is important and relevant. To solve such a problem, a variational algorithm for assimilation of measurement data over a certain period of time in the transport model of the impurity under study can be applied. On the basis of information for various time points at a given time interval and the model used as a space-time interpolant, it is possible to obtain a solution consistent both with the measurement data and with the model. In this paper, by identifying the initial concentration field while minimizing the quadratic prediction quality functional, we obtain a solution for the entire time interval. To construct the gradient of the functional in the parameter space, the solution of the corresponding conjugate problem is used, and to find the iterative parameter, the problem is solved in variations. When conducting numerical experiments, real data on concentration of chlorophyll a, the information on wind effects over the Sea of Azov (http://dvs.net.ru/mp/data/201507vw.shtml) and clouds (worldview. earthdata.nasa.gov) was used. Using the circulation model of the Sea of Azov in sigma coordinates, the fields of currents, the coefficients of turbulent diffusion under the north and north-west wind, which prevailed in the observed time period, were obtained. The obtained model fields were used as input information for the passive impurity transfer model. As a result of the variational procedure of assimilation of measurement data, the concentration fields found for the time interval used (five days) are consistent with the measurement data and with the transfer model. Such agreement is due to the minimization of the functional under the restrictions used. © 2023 Space Research Institute of the Russian Academy of Sciences. All rights reserved.

Azov Sea, chlorophyll a concentration, satellite data, the adjoint equation, transport model

2023-04-19 15:49:15