Development of New Probabilistic Hazard Information Using High Resolution Quantitative Precipitation Forecasts

Given operational time constraints, it is very difficult for an operational forecaster to thoroughly analyze data from every high resolution model. This is particularly concerning for quantitative precipitation forecasts (QPF) from Convection Allowing Models (CAMs), which can exhibit complex and highly varying spatial structures. Informative, simple and novel displays of QPF data are needed to aid forecasters in the probabilistic forecasts of heavy rain and flooding, among other things. This talk will detail progress made from a subproject of a United States Weather Research Program (USWRP) proposal, which involves creating probabilistic hazard information (PHI) from identifying heavy QPF objects within CAMs.

QPF objects are identified using the Method for Object-Based Diagnostic Evaluation (MODE) tool within the Model Evaluation Tools (METv5.1) developed at the Developmental Testbed Center (DTC). In short, MODE identifies coherent spatial objects that are easily recognized by the human eye, but not always easily displayed with common ensemble verification. Treating forecasted heavy precipitation areas as objects allows the structure and magnitude of QPF to be preserved.

MODE is applied to an ensemble of opportunity consisting of time-lagged runs of the High Resolution Rapid Refresh (HRRR), the National Severe Storms Laboratory (NSSL) ensemble, the High Resolution Window Advanced Research WRF (HIRESW-ARW) and Non-Hydrostatic Mesoscale Model (HIRESW-NMMB). The utility of new potential PHI created from heavy QPF objects will be shown. These techniques are in an early testing phase, and will be further refined before they are used by the Weather Prediction Center (WPC) as a forecasting tool.