Soil moisture dynamics are a key parameter in regulating agricultural productivity and ecosystem functioning. The accurate monitoring and quantitative retrieval of soil moisture play a crucial role in optimizing agricultural water resource management. In recent years, the development of multi-source remote sensing technologies—such as high spatiotemporal resolution optical, radar, and thermal infrared sensors—has opened new avenues for efficient soil moisture retrieval. However, the accuracy of soil moisture retrieval decreases significantly when the soil is covered by vegetation. This study proposes a multi-modal remote sensing collaborative retrieval framework that integrates UAV-based multispectral imagery, Sentinel-1 radar data, and in situ ground sampling. By incorporating a vegetation suppression technique, a random-forest-based quantitative soil moisture model was constructed to specifically address the interference caused by dense vegetation during crop growing seasons. The results demonstrate that the retrieval performance of the model was significantly improved across different soil depths (0–5 cm, 5–10 cm, 10–15 cm, 15–20 cm). After vegetation suppression, the coefficient of determination (R²) exceeded 0.8 for all soil layers, while the mean absolute error (MAE) decreased by 35.1% to 49.8%. This research innovatively integrates optical–radar–thermal multi-source data and a physically driven vegetation suppression strategy to achieve high-accuracy, meter-scale dynamic mapping of soil moisture in vegetated areas. The proposed method provides a reliable technical foundation for precision irrigation and drought early warning.

Agricultural nonpoint-source pollution (ANPSP) is a key cause of global environmental problems. However, the estimation of ANPSP, based on agricultural land use type, crop management, and attenuation of pollutants with distance, is lacking. Using the Mun River Basin as an example, this study established quantitative response relationships between subbasin flows and hydrological and water quality parameters. A good matching of the monitored sections and the control area based on flow relationships was achieved. By determining flow paths and flow distances, the overland and in-river transport attenuations of ANPSP were clarified. The overland and in-river transport and attenuation parameters were also quantified. The land use distribution and structure were further refined through crop management, which included crop types and crop rotation (monocropping or double cropping). Based on the above procedures, quantitative relationships among land use pattern, crop management, attenuation of pollutants with distance, and river water quality were established and used to construct six kinds of regression models. Among these models, the best modeling results were obtained when the parameters of water quality, land use structure, crop management, and soil nutrient attenuation were included. The modeling accuracy in the dry season increased from 0.398 to 0.881 when information about attenuation with distance and crop management was included. Similarly, the modeling accuracy in the wet season increased from 0.365 to 0.727. This study’s findings indicate that the constructed water quality model is effective and has significance for the quantitative determination of ANPSP.