Process-oriented Model Diagnostics for Extended-range Forecasts

Model validation and evaluation is an indispensable part of model improvement efforts. While performance-oriented metrics provide quantitative measures on how well a model does, process-oriented metrics help to reveal model deficiencies and identify pathways to model improvement. A suite of process-oriented, observation-based model diagnostics are developed to evaluate the processes that are critical to forecasting on the synoptic to subseasonal time scales. The suite consists of three levels of diagnostics: i) evaluation of systematic model errors in representing moist convection and cloud processes; ii) evaluation of the sources of predictability relevant on S2S timescales (such as the MJO, NAO and weather regimes); iii) evaluation of high-impact weather systems (such as tropical cyclones, blocking, etc.). The presentation will illustrate examples for each level of diagnostics using the GEFS retrospective forecasts. The diagnostics will be made available to the community via the Model Evaluation Tools (METplus) and the Model Diagnostics Task Force (MDTF) Diagnostic Package.