Abstract
Soil moisture plays a key role in the formation of surface runoff and floods, especially in conditions of spring snowmelt and extreme precipitation. The relevance of soil moisture monitoring is increasing for areas with high flood risk, such as the East Kazakhstan region. This article provides an overview of modern remote sensing (remote sensing) methods used to assess soil moisture in order to predict floods. Global and regional studies demonstrating the effectiveness of integrating satellite data (SMAP, Sentinel-1/2, SMOS, etc.) into hydrological models are considered. A comparative analysis of ground-based and remote methods of measuring soil moisture is carried out, approaches using radiometers, radars, multispectral sensors, as well as unmanned aerial vehicles (UAVs) are described. Special attention is paid to data processing algorithms, including spectral indexes, machine learning methods, and neural network models. The capabilities of the software (Google Earth Engine, SNAP, ArcGIS, QGIS) in mapping and monitoring tasks are analyzed. The review highlights the potential of remote sensing in improving the accuracy of early flood warnings and laying the foundations for sustainable water risk management.