Verification of QPF from NWP Models and WPC Associated with Tropical Cyclones (340)

Mark Klein, National Weather Service/Weather Prediction Center, College Park, MD

 

Abstract:

One of the most significant impacts of tropical cyclones affecting the continental US (CONUS) is the excessive rainfall that they can produce approximately one-quarter of all storm-related fatalities in the past 50 years were a result of rainfall-induced floods or mudslides. Therefore, providing accurate rainfall forecasts is critical to saving lives and property.The Weather Prediction Center (WPC) has produced quantitative precipitation forecasts (QPF) for the CONUS for nearly 60 years, and for the past several hurricane seasons, storm-centric WPC QPF graphics have been an integral part of the National Hurricane Center's (NHC) public forecast package. Given this increased profile of WPC QPF, it's of particular importance that objective verification is available to assess WPC's skill, as well as the skill of the numerical weather prediction (NWP) guidance that is utilized by WPC forecasters. While WPC has developed and evolved a robust verification program over the past several decades, much of this effort has focused on a CONUS-wide scale, with little work toward objective metrics for high impact events (such as tropical cyclones). Event-focused verification would prove useful not only to improving operational QPFs, but also can be shared with model developers to inform future NWP implementations.This poster will examine performance of WPC and NWP QPFs for tropical cyclones impacting the CONUS from 2012-2017. It will verify Day 1-3 24-hour QPF, subsetting the CONUS domain to include only regions where the observed precipitation was associated with the tropical system. Determination of the regions, which can vary in spatial extent daily, will be based on River Forecast Center (RFC) Stage IV quantitative precipitation estimates (QPE) and the observed track of the cyclone. Stage IV QPE will also serve as ground truth for all statistics presented.