EXPERIENCE IN IMPROVING THE SPATIAL RESOLUTION OF GRACE SATELLITE GRAVIMETRY DATA FOR THE ZHAIYK RIVER BASIN (URAL) USING MACHINE LEARNING

The article presents a study on using machine learning to improve the spatial resolution of GRACE satellite gravimetric data for the Zhaiyk River basin (Ural). The work aims to solve the problem of low spatial resolution of GRACE data. making it difficult to analyze water resources in detail at the regional level. The research methodology uses a statistical method of scaling down using the Random Forest model. Additional datasets were used to improve spatial resolution. including a digital terrain model, vegetation indices (NDVI), and water surface indices (NDWI). The results demonstrated the effectiveness of the proposed methodology: improved data provides a more accurate distribution of the equivalent thickness of the water layer while maintaining general statistical properties. The Pearson correlation coefficient was 0.9424, which means a significant degree of reliability of the results. The practical significance of this study lies in the prospective application of the developed approach to improve the assessment of water resources, forecasting hydrological events. and developing plans for adaptation to climate change in the Zhaiyk River basin.