多源降水数据的不确定性解析与融合算法研究OA

High-precision,high-resolution precipitation data are critical for simulating river basin hydrological pro-cesses,managing water resources management,and monitoring floods and droughts.Satellite remote sensing and reanalysis systems can provide valuable spatiotemporal coverage;however,their raw estimates often exhibit large uncertainties,making it difficult to meet the application requirements of sectors like water resources.Data quality must be improved through bias correction and data merging,while traditional multi-source precipitation merging approaches,which rely heavily on dense gauge networks,are unsuitable for sparsely observed regions.Statistical uncertainty analysis offers an alternative by jointly evaluating multiple datasets with no much dependence on ground observations,quantifying their error characteristics,and minimizing overall errors to produce optimal merged esti-mates.This review summarizes major statistical uncertainty analysis methods,outlines common error sources and cor-responding correction strategies,and highlights emerging challenges and future directions in multi-source precipita-tion merging.This study aims to provide more reliable precipitation estimates and holds significant importance for improving the quality of precipitation products to support water management and agricultural risk reduction.