NWP Testing and Evaluation for Specific Needs: The DTC Connection with NOAA Testbeds

Numerical weather prediction (NWP) research at NOAA testbeds and National Weather Service agencies in recent years has increasingly involved phenomena that require non-standard and adaptable verification methods. Examples include forecasts for ‘atmospheric rivers' (ARs) and related extreme precipitation events investigated by the Hydrometeorological Testbed (HMT), and high resolution ensemble-based severe weather forecasts inter-compared during spring exercises of the Hazardous Weather Testbed (HWT). The Developmental Testbed Center (DTC) has collaborated with many of these projects by facilitating verification activities and by developing verification techniques that can provide useful assessments and comparisons. Many of these techniques and utilities have been incorporated into the Model Evaluation Toolkit (MET), including new object-based verification methods in the Method for Object-based Diagnostic Evaluation (MODE) package, and comprehensive display utilities in the METViewer package. In this poster we present some pertinent results that have proceeded from DTC-based verification efforts undertaken in conjunction with other NOAA testbeds. Among those illustrated are ensemble verification routines that have been particularly useful during the HMT's winter exercises in California, spatial verification of reflectivity objects for the HWT, and verification uncertainty estimates for evaluation of hurricane model parameter changes resulting from collaboration between the DTC and the Environmental Modeling Center of NCEP. We also suggest possible future directions for verification technique development.