Using BOI_Verify to Identify Model Preferences in Large Temperature Change Events

BOI_Verify, created by Tim Barker (WFO Boise, ID (2006)), is an NWS framework for archiving model and human gridded forecasts of temperature, probability of precipitation, and other elements, as well as observed grids of those elements. BOI_Verify therefore provides the capability of verifying any of the gridded forecasts--in particular, maximum and minimum temperatures. This study focuses on those cases where forecast temperature changes were large from one day to the next, i.e., "high-impact" events. The cases are sorted in order of largest overall temperature change within a forecast region. By comparing verification of different model forecasts against each other, it is possible to see which models did best under cases of rapid warming or cooling. BOI_Verify also allows for the identification of past analogs to any given model forecast and the ability to see if the analogs verified in similar ways to the current model forecast. When a given model consistently displays similar spatial forecast errors, those errors can be used to adjust the model for use in real-time gridded forecasts, resulting in much more accurate forecasts than using the current model forecast alone.

The poster will illustrate how BOI_Verify is used to identify large temperature-change days and how various models verified on those days. The analog forecast method will also be briefly described in order to show how a current model's grids can be improved in those cases.