Application of combined traditional and MODE verification methods in spring precipitation forecast evaluation of CPEFS Model

This study employs a combined approach of traditional verification and MODE (Method for Object-based Diagnostic Evaluation) to systematically evaluate the spring precipitation forecast products of the Cloud Precipitation Explicit Forecast System (CPEFS) model developed by the China Meteorological Administration Weather Modification Centre from 2021 to 2023.The evaluation focuses on two typical weather processes: the westerly trough-cold vortex type (WT-CV) and the westerly trough-shear line type (WT-SL),conducting in-depth analysis from three core aspects: forecast lead time,precipitation period,and magnitude threshold.The results indicate that the temporal evolution patterns of forecast performance for both weather processes exhibit high consistency.Specifically,the forecast performance is optimal during the 14:00-20:00;the forecast scores for precipitation ≥1.0 mm are significantly better than those for ≥10.0 mm;and the model forecast capability shows a pronounced decreasing trend with increasing forecast lead time.However,the causes of forecast errors for the two types differ fundamentally.For WT-CV type processes,the core error is "intensity overestimation" which originates from the systematic wet bias of the model.For WT-SL type processes,the key issue for is "position deviation," characterized by significant spatial displacement.Traditional verification enables objective quantitative evaluation of forecasts,while MODE verification can accurately diagnose the sources of spatial structure errors.Therefore,it is recommended to adopt a combined approach of "traditional verification+MODE verification" in operational verification to achieve multidimensional comprehensive evaluation of model performance,and to provide more comprehensive technical support for precipitation forecast operations.