Estimation of hydrophysical characteristics of the aquatic environment using satellite images in the context of incomplete information
In the method of assessing the state of individual parts of the aquatic ecosystem based on satellite images (determining the concentration of total suspended and dissolved organic matter in particular), an algorithm for calculating their concentration based on data from space scanners is used for complete information. This is achieved by creating, using cosmic data, a multidimensional linear regression model of the dependence of the characteristics of temperature, salinity, total suspended matter and dissolved organic matter on the influencing factor variables — spectral values 7 bands of Landsat-5. To assess the characteristics of the aquatic environment from satellite images of the sea surface, when the surface of the water is sometimes covered with clouds or there is no image of a part of the studied water area with the required data, it is proposed to use an optimal interpolation (extrapolation) of the spectral brightness values of images in the water areas. The method of optimal interpolation of Kolmogorov spectral brightness of satellite images under incomplete information in the task of improving and testing methods for the remote determination of the hydrophysical characteristics of the marine environment is described. The optimal Kolmogorov interpolation method was tested for the field of the spectral values of the sea surface for the problem of estimating the hydrophysical characteristics of the aquatic environment using the example of the Sentinel-2 satellite channels, which showed for the spatial resolution of 10 and 20 m insignificant errors and a sufficiently high correlation with the brightness values, and for the channels with the spatial discrimination of 60 m is smaller than the value of the correlation coefficient, which is associated with significantly larger distances between the pixels. These results are acceptable for practical use of interpolation (extrapolation) of spectral brightness values of satellite images of the sea surface under incomplete information, which creates prerequisites for creating a multidimensional linear regression model for calculating the hydrophysical characteristics of the aquatic environment according to incomplete information satellite conditions.
Kolmogoroff, A. (1941). Interpolation und Extrapolation von station ren zuf lligen Folgen. Izv. AN SSSR. Seriya matem., no. 5, pp. 3–14. (in Russian). http://www.mathnet.ru/links/29ab5de7233d6a082164a9d83777f31e/im3775.pdf
Dargeyko, L. F., Fedorovskiy, A. D., Lukin, A. Ye., Porushkevich, A. Yu. (2011). Evaluation of oil-gas-bearing capacity of territory sites by the Kolmogorov spatial interpolation method. Dopovidi Nacionaljnoji akademiji nauk Ukrajiny, no. 10, pp. 100–103. (in Russian).
Pukhtiar, L. D., Stаnychnyi, S. V., Tymchеnko, I. Ye (2009). The optimal interpolation of remote sensing data of sea surface. Morskoi hydrofyzychеskyi zhurnаl, no. 4, pp. 34–50. (in Russian). http://xn--c1agq7a.xn--p1ai/index.php/repository?id=459
Fedorovsky, A. D., Porushkevich, A. Yu., Chepyzhenko, A. A., Yakimchuk, V. G. (2013). The regioanals alhorythms with space survey of sea researching on Kerch Strait example. Ekolohichnа bеzpеkа tа pryrodokorystuvаnnia: Zb. nаuk. pr., vol. 12, pp. 33–42. (in Russian). http://dspace.nbuv.gov.ua/handle/123456789/57574