Real-time Flash Flood Observational Datasets in Support of the NSSL's 2014 HWT-Hydro Experiment

- Indicates paper has been withdrawn from meeting
- Indicates an Award Winner
Tuesday, 6 January 2015: 4:45 PM
127ABC (Phoenix Convention Center - West and North Buildings)
Brandon R. Smith, CIMMS/Univ. of Oklahoma, Norman, OK; and J. Gourley, L. P. Rothfusz, R. Clark III, J. M. Erlingis, Z. L. Flamig, E. Mintmire, K. L. Ortega, and B. Cosgrove

The National Severe Storms Laboratory's 2014 Hazardous Weather Testbed - Hydrology Experiment (HWT-Hydro) brought National Weather Service (NWS) forecasters together to test experimental flash flood prediction products and forecasting methods. The experiment ran from July 7th August 1st, 2014, featuring up to four NWS forecasters each week who issued experimental flash flood watches and warnings based on the operational and experimental forecast products. In order to provide verification for these experimental watches and warnings, four different verification datasets were utilized: 1.) Severe Hazards Analysis & Verification Experiment (SHAVE) reports, 2.) NWS Local Storm Reports, 3.) mPING citizen science reports, and 4.) USGS streamflow. The combination of these four datasets allowed for the forecasters to perform daily evaluations of the previous days' watch and warning operations. While all four datasets provided validation, the HWT-Hydro experiment worked extensively with the SHAVE project to deliver reports that were of a highly detailed nature both in content and density. This presentation gives an overview of the collection methodologies of flash flooding occurrences during HWT-hydro and evaluates their skill in terms of reporting flood magnitude, delineating spatial extent, and identifying the specificity of the impact.