CREATION OF A GEODATABASE ON GEOMORPHOMETRIC PARAMETERS OF THE RELIEF BY THE EXAMPLE OF THE NOVOSIBIRSK REGION USING SATELLITE IMAGES

Abstract

Traditional methods of qualitative description of relief used in the process of landscape analysis, zoning, forecasting of exogenous processes, assessment of agricultural land do not allow to distinguish objectively forms of relief, to reveal statistically reliable relations between relief indicators and components of geosystems. For agro-ecological evaluation and land grouping with the help of modern methods of data classification, a geodatabase with quantitative indicators of relief is needed. Space images of high and medium spatial resolution are necessary for formation of geodata base of geomorphometric indicators during regional researches.
 Background. Quantitative (geomorphometric) indicators of relief are of great practical importance when identifying and describing landforms and relief elements. One or two indicators are often used for agricultural land assessment. In foreign literature various combined topographic indices are widely used for land classification. The scientific significance of the research is associated with modern methods of geomorphometric analysis of the relief based on remote sensing data and their application for automated mapping of relief types.
 Purpose. Creation of a geodatabase (GDB) of geomorphometric parameters of the Novosibirsk region on the basis of satellite data.
 Materials and methods. Digital modeling and geomorphometry methods were used in the work.
 Results. The database of geomorphometric parameters of the relief with the use of middle and high spatial resolution space images ALOP PALSAR (12.5 m/pixel) and ALOS DSM (30 m/pixel), topographic maps of M 1:25000 were developed. A geodatabase of geomorphometric parameters of the relief by the example of the Novosibirsk Region was developed including GIS information layers with geomorphometric parameters of the relief and attribute tables. The LGB has been used to map the types of relief in the Novosibirsk Region agrolandscapes. Automatic classification of relief types using multispectral images of K-Nearest Neighbor (KNN or k-NN, nearest neighbor) teacherless algorithms and the iterative self-organizing algorithm ISODATA (Self-Organizing Data Analysis) were performed. The input raster data for the terrain type classification were terrain parameters and vegetation indices NDVI, EVI, OSAVI, TSAVI, determined on the basis of multispectral satellite images Sentinel-2 A. By the example of Priobskiy central forest-steppe agrolandscape mapping of relief types was carried out: watershed drained undrained flat-leveled, watershed drained weakly drained, interfluvial undrained flatwaterlogged. The main relief-forming processes in the Priobskoye central forest-steppe agro-landscape are planar washout and waterlogging.
 Conclusion. Modern methods of geomorphometry and geoinformatics allow you to create a spatial database of geomorphometric parameters for a comprehensive assessment of the land on the basis of remote information. The use of a geomorphometric database of geomorphometric parameters of relief and vegetation indices made it possible to carry out mapping of relief types for the Novosibirsk region agrolandscapes.