Use of surface mesonets, radiosondes and profilers to predict basin-specific QPF and flash floods in complex terrain

The prediction of locally heavy rainfall and flash flooding is a very challenging problem in mountainous terrain of Colorado and Arizona. Two significant problems in the provision of basin-specific QPF are the quantitative assessment of changing temperature, moisture and wind flow in the sub-cloud layer and the identification of local terrain-induced convergence zones and wind circulations. Two operational flash flood prediction programs which addressed these concerns were the Flash Flood Prediction Program(F2P2) of Urban Drainage & Flood Control Distrct of Denver, Colorado and the Meteorological Services Pilot Program(MSPP) of the Flood Control District of Maricopa County, Arizona. Both progrmas supported the local emergency response communities and cooperated with the local National Weather service offices.

Both the F2P2 and the MSPP addressed these forecast concerns through the integrated application of surface mesonets, vertical wind profiler observations, and conventional radiosonde obervations to assess the pre-storm atmosphere.

Conventional radiosonde observations from Denver, Colorado and Luke AFB (SWAMP) were made available to provide the basic vertical temperature, moisture and wind profiles. These profiles, in turn, were modified by plotting "surface mesonet" observations from the surface to 7,000ft AGL which were located in the program service areas. At east one automated weather station was located in each basin. The mesonet observations of temperature and dew point were plotted on the Skew-T. Log to monitor hourly variations in the sub-cloud layer moisture and stability fields. Additionally, the PWI could be changed by variations in the vertical moisture structure. These "mesounds" provided very useful and reasonably accurate guidance in preparing predictions of flash flooding and basin-specific QPF.

For each mesonet weather station the "mesound" was analyzed using the stations surface temperature and dew point to run a parcel stability and QPF analysis for that location. The QPF provided an estimate of the peak 30-minute and hourly rain rates if a cloud ingested the air "over" that basin. In some cases, the convective cap inversion could be effectively monitored for explosive release or continued capping effects.

Additionally, surface mesonet maps were used to identify terrain-induced circulation fields which could produce surface convergence fields and assist convective release. The Denver Cyclone and the Phoenix Urban Low were easy to identify and use in predicting storm motion. The F2P2 covers a 1,770 sq mile area and the MSPP covers 5,470 sq miles. The mesonet/meosound approach allowed basin-specific and forecast-zone specific QPF's for areas of 100 square miles to 500 square miles. The QPF's were issued every six hours. Examples of flash flood predictions from the 2000 operational season and verifications from prior years will be presented at the conference. Applications of the techniques can be made in any area were mountainous terrain exists and mesonets have been established.