Areal rainfall verification and analysis of the "multi-statistical optimal integration method" in qingjiang river basin

Based on the verification of the EC ensemble precipitation forecasts, the Multi-statistical Optimal Integration Method (OES) were designed. Six kinds of precipitation forecasts including OES, EC, NCEP, AVE, and MODE were verified over Qingjiang River Basin in 2020 flood season. The results showed that precipitation forecast by OES exhibits better performance in light rain, rainstorm and heavy rainstorm, while EC and NCEP performs better in medium and heavy rain. The predictability of OES was slightly decreased in 48 and 72h.OES work more stable when the low-level vortex or the trough coupled with low-level shear dominates. OES have the advantage in extend lead times. The mean absolute Error of OES in 288-360h is smaller than that of the other kinds of precipitation forecasts in 240h.