Determining Precipitation Frequencies Over the Southern United States
A COMET Partners Project
As a result of the modernization of the National Weather Service (NWS) which began a decade ago, forecasters now are
equipped with a variety of new tools and technologies which enable them to provide products and services in ways never
before possible. Although the mission of the NWS has not changed, the capabilities of forecasters have been vastly
enhanced. The end result is that the public will be provided with more accurate forecasts, with higher time and space
resolution, and in a gridded-digital form that will enable them to utilize the products in new ways. In particular,
forecasters now issue forecasts of the probability of precipitation (PoP) for 3-hourly time periods in the form of
high resolution gridded fields. This is a significant increase in resolution over the 12-hourly PoPs that have been
issued by the NWS for over 35 years covering much larger areas--county-sized regions or larger. The new software
that the NWS uses to prepare these forecasts is called the Interactive Forecast Preparation System (IFPS) which runs
in the AWIPS environment. The IFPS software allows forecasters to develop and communicate products such as the high
resolution PoPs in gridded, graphical, and text forms.
A knowledge of the climatological likelihood of an event is an excellent starting point for a forecast.
Climatological frequencies also provide a basis for assessing one measure of forecast skill. Jorgensen (1967)
developed climatological precipitation probabilities for the U.S. for 6-, 12-, and 24-hour periods (using data from
1949-1964), publishing them in the form of maps. Jorgensen's work has served well for decades in providing important
training materials for forecasters. However, frequencies are needed for 3-hour intervals, and using the longer period
of data that now is available (e.g., from 1949-present).
This research is developing climatological guidance that will provide an essential foundation for forecasters as
they initiate the higher resolution PoPs. To our knowledge, precipitation frequencies for 3-hourly periods, for
a dense array of sites, has never been developed. Our objective is to extend Jorgensen's work to 3-hourly
periods, and if resources permit, to re-derive climatological frequencies for the longer time periods using data
for the most recent 30 or 40 years.
The data from which to derive the precipitation frequencies are available on CD-ROMs developed by NOAA's Forecast
Systems Lab and NCDC which contain the NCDC Hourly Precipitation Data. We have developed software to extract the
hourly precipitation data and compile the precipitation frequencies. The frequencies will be displayed in tabular and
graphical forms. Maps showing horizontal distributions of the results also will be prepared.
Once these probabilities are calculated, they will be published in forms most useful to forecasters and the research
community. The results already are on the web. The goal is for forecasters to easily integrate the results into the
operational system which NWS forecasters use to produce their forecasts.