Advances in Soil Moisture Mapping Techniques and Disaggregation Algorithms for Environmental and Agricultural Applications

Abstract. Soil moisture (SM) is a critical variable for understanding the water cycle, climate change, and agricultural management. This paper reviews advanced remote sensing techniques and disaggregation algorithms for high-resolution SM mapping, focusing on environmental and agricultural applications in North Africa and the UAE. Remote sensing methods, including optical sensors (e.g., Apparent Thermal Inertia, Temperature Vegetation Dryness Index), active microwave sensors (scatterometers, Synthetic Aperture Radar), and passive microwave radiometers (SMOS, SMAP), are evaluated for their ability to map SM at varying spatial and temporal scales. Despite advancements, coarse resolution remains a challenge for regional applications. To address this, two innovative downscaling algorithms are presented: the SMOS Semi-Empirical Method, which fuses SMOS data with ECMWF skin temperature and MODIS/Sentinel-3 NDVI to achieve 1 km and 300 m resolutions, and the Artificial Neural Network (ANN) Method, leveraging multi-sensor data to produce 60 m resolution SM maps. These algorithms have been validated across diverse environments, demonstrating RMSE values of 0.04–0.10 cm³/cm³. The case studies presented highlight their operational utility in flash flood monitoring (Algeria, Tunisia, UAE), ecosystem dynamics (Chott el Djerid, Tunisia), and precision agriculture (East Oweinat, Egypt). Future work includes the integration of multi-sensor data, to enhance machine learning models, and the improvement of SM measurements at deeper soil layers to support applications in arid regions.