Assessing the Effectiveness of the Three-Cornered Hat Data Fusion Technique for Satellite Precipitation Data and Its Impact on Runoff Forecasting

Abstract

Runoff forecasting remains a critical challenge in complex basins around the world, where data scarcity and detection problems limit the forecasting performance. Data fusion offers a promising alternative, with the potential to generate enhanced satellite precipitation datasets for use in data-driven runoff forecasting models. This study investigates the effectiveness of the Three-Cornered Hat (TCH) method for fusing satellite precipitation datasets and its impact on runoff forecasting. The TCH method was applied to three satellite precipitation products (SPP), creating a fused dataset that was compared to two benchmark products: IMERG-ER and MSWEP. While the TCH method demonstrated suitability for forecasting in data-scarce

regions, it did not outperform the benchmark products, showing comparable results to IMERG- ER. Interestingly, MSWEP exhibited superior performance across different lead times,

suggesting that the fusion of multiple precipitation data sources may enhance the performance of runoff forecasting models. The study highlights a key limitation of the TCH method at fine temporal scales. In certain conditions, data may be considered statistically dependent, which could have an adverse effect on the effectiveness of the method. Future research should focus on resolving these limitations and exploring the potential of integrating diverse data sources to unlock further improvements in runoff forecasting.