Prokopyeva K.O. THE USE OF MULTI-TEMPORAL HIGH-RESOLUTION SATELLITE IMAGES TO SOIL SALINITY ASSESSMENT IN THE SOLONETZIC COMPLEX (REPUBLIC OF KALMYKIA) // Arid Ecosystems. 2022. Vol. 28. № 4 (93). P. 61-74. | PDF
Remote assessment of soil salinity of natural solonetzic complexes, which are characterized by subsurface soil salinity, is a challenging task. Nevertheless, research in this area is promising, as salinity is a striking limiting factor affecting vegetation growth, and thus manifests itself on the spectral characteristics of the land surface. In this paper, an analysis of multi-temporal high-resolution satellite images compared with detailed ground data on soil salinity was carried out using the principal component method and multiple linear regression. Images from QuickBird (2007) and SuperView-1 (2021) satellites were used as remote sensing data with a spatial resolution of 2 m. Salinity was estimated by specific electrical conductivity (EC) in a water suspension of 1:5. Ground surveys were carried out in 2011 and 2021. It was found out that there were no significant changes in soil salinity in the key area during the 10-year period, but there were changes in vegetation conditions, which are reflected in the NDVI vegetation index maps. The principal components were calculated from multi-temporal high-resolution satellite images and it was concluded that the first three components explain almost 97% of the total image variability. Models based on multiple linear regression analysis describe well the soil salinity (R2 of the model is 0.68, 0.77, 0.83 for layers 0-30, 0-50, 0-100 cm, respectively). The constructed models based on remote sensing data have shown good agreement when tested with the control sample (R2 between predicted and measured EC values is 0.70, 0.87, 0.83 for layers 0-30, 0-50, 0-100 cm, respectively). The proposed models will be useful for estimation of salinity of soils of solonetzic complex of dry steppe from high resolution satellite images.
Keywords: QuickBird, SuperView-1, assessment of soil salinity, solonetzic complexes, Principal Components Analysis (PCA), NDVI, Caspian Lowland.
Acknowledgement. The author expresses gratitude for the help in the work of M.V. Konyushkova, I.P. Minkeev, A.F. Novikova, M.B. Shadrina, S.S. Ulanova, I.N. Semenkov, A.A. Kontoboitseva, Isinufen, N.M. Tserenov, U.Y. Ulumdjiev.