Assessing operational usability of short-and medium-range hydro-meteorological forecast products in lower Jinsha River Basin

The evaluation of hydro-meteorological forecast products currently relies heavily on qualitative comparison of performance, with a notable lack of quantitative assessment regarding their practical usability. Furthermore, existing evaluation indices often fail to incorporate the inherent uncertainty in forecasts across different lead times. Consequently, the development of a robust, quantifiable availability evaluation indicator has become an urgent priority. In this study, we proposed a forecast product usability evaluation index, integrating conventional accuracy indices—including the Threat Score (TS), mean absolute relative error (MARE), and Nash-Sutcliffe efficiency coefficient (NSE)—while explicitly taking account of forecast uncertainty across varying lead times. Taking the lower Jinsha River Basin as a case study, we applied this index to analyze the accuracy and assess the usability of several hydro-meteorological forecast products. The results indicate the following usability ranking: cumulative inflow forecast products for the Wudongde Reservoir > inflow forecast products > areal precipitation forecast products for the lower Jinsha River Basin > flood process forecast products for the Wudongde Reservoir. Specifically, for the 1~5 d lead times, areal precipitation forecast products for the lower Jinsha River Basin, inflow forecast products, cumulative inflow forecast products, and flood process forecast products for the 1~2 d lead times for the Wudongde Reservoir exhibit high usability (usability index≥0.7). For the 3 d lead time, flood process forecast products demonstrate moderate usability (usability index≥0.6). These findings provide a technical foundation to support the joint operation and scheduling of reservoir groups in the lower Jinsha River Basin.