SATELLITE DATA PROCESSING ALGORITHM IN THE PROCESS OF FORMATION OF THE TIME SERIES OF VEGETATION INDEXES

Abstract

The diverse spectral indexes computed from the satellite images are used extensively in the world practice of remote sensing of the Earth from space. This approach proved its validity for the satellite monitoring of the underlying terrain, detection of ongoing changes and trends of their dynamic patters. Accumulated prodigious amount of satellite data, the state-of-the-art methods of thematic interpretation gave rise to creation of services providing free access to both images and to image processing results. Notwithstanding the foregoing, in the furtherance of the local and regional scale it turns out that usage of the end products of thematic processing of space information supplied by the known available services was not efficient on all occasions. Consequently, we may need to generate our own archives of the long-term series of satellite indexes. The volume of files containing the digital index matrices computed based on the MODIS satellite low resolution data subject to the complete coverage of the territory of Kazakhstan surpasses 4 Gb. This often results in the delayed computations, and on frequent occasions in infeasibility of computation of a full matrix when the medium specs computers are employed. This article is focused on the satellite data processing algorithm in the process of formation of the time series of vegetation indexes. As a consequence, the multi-year archive of vegetation indexes (over a period of 2001-2020), which provided a basis for trend analysis of the underlying terrain, determination of their future trends and forecasting of their changes was created within the territory of the Republic