Verification Techniques Appropriate for Cloud-resolving NWP Models By

Verification is a critical component of the development and use of forecasting systems. Ideally, verification should play a role in monitoring the quality of forecasts, provide feedback to developers and forecasters to help improve forecasts, and provide meaningful information to forecast users to apply in their decision-making processes. In addition, as noted by Mahoney et al. (2002) forecast verification can help to identify differences among forecasts. Finally, because forecast quality is intimately related to forecast value, albeit through relationships that are sometimes quite complex, verification has an important role to play in assessments of the value of particular types of forecasts (Murphy 1993). In recent years forecasting approaches have become more complex and have been applied on finer scales. Unfortunately, traditional approaches for the verification of spatial forecasts, including QPFs and convective forecasts, are inadequate to meet current needs. Typically, verification techniques have been based on simple grid overlays in which the forecast grid is matched to an observation grid or set of observation points. From these overlays, counts of forecast/observation (Yes/No) pairs are computed, to complete the standard 2x2 contingency table. The counts in this table can be used to compute a variety of verification measures and skill scores