92nd American Meteorological Society Annual Meeting (January 22-26, 2012)

Thursday, 26 January 2012
An AMBER Playback of the Tennessee Flash Floods of May 2010: Contributions of Tropical Rainfall
Hall E (New Orleans Convention Center )
Robert S. Davis, NOAA/NWS, Pittsburgh, PA
Manuscript (7.3 MB)

Poster PDF (1.1 MB)

During the widespread flash floods across Tennessee, sixty-one rain gages within 225 km of the Nashville WSR-88D, reported over 8 inches of rain (203 mm) between 06 UTC on 01May2010 and 06 UTC on 03May2010. Thirty-three of those rain gages had 12 inches (305 mm) or more of rain with a maximum of 19.70 inches (500 mm) at the Warner Parks Nature Center. The robust playback capability of the Areal Mean Basin Estimated Rainfall (AMBER) is used to compute both tropical Z/R and standard convective Z/R at the 61 rain gage locations. The radar rainfall estimates for both tropical and standard convective Z/R are compared with the observed rainfall at each gage. The gage/radar comparisons are then used to determine if tropical rainfall rates occurred during the flash flood event. The upper air soundings for Nashville are also examined for parameters associated with tropical Z/R, such as high precipitable water, and a deep warm rain coalescence layer. The Flash Flood Monitoring and Prediction Program (FFMP) is the primary flash flood detection tool of the National Weather Service. The occurrence of tropical Z/R has a critical impact on the capability of FFMP to detect flash floods. FFMP uses the Z/R relationship that is manually set in the WSR-88D precipitation rate algorithm. If tropical rainfall rates are occurring, but the WSR-88D is set to standard convective Z/R, the radar estimates in FFMP will only be about half the observed rainfall. In this scenario, FFMP will grossly underestimate the rainfall and the potential severity of the flash flooding.

Supplementary URL: