Rainfall frequency data—They really do matter

Since the early 1980's, the City of Fort Collins, Colorado, has been aggressive in developing and implementing a comprehensive stormwater management program which includes design criteria, drainage master planning, capital projects, maintenance, water quality protection and emergency response. This program is funded by utility fees paid by the citizens of the City of Fort Collins. In 1997, Fort Collins was nationally recognized by the Federal Emergency Management Administration for their accomplishments and high standards. However, in July 1997, the heaviest rain ever to hit an urbanized area in Colorado dropped 10.5 inches of rainfall in 5 hours with a 30-hour total of 14.5 inches over a portion of the city. The resulting flooding killed five residents, caused at least $200 million in damages and exceeded estimated 500-year flood flows in some urbanized basins.

Following this extreme event, the Fort Collins Utilities convened a technical task force to re-evaluate the city's storm water design criteria. The value that had previously been utilized for modeling and design purposes was the 100-year 2-hour rainfall (2.9 inches) as determined from NOAA Atlas 2, Precipitation Frequency Atlas for the Western United States, Vol. III (Colorado) published in 1973.

This paper describes the process that culminated in a 1999 City Council vote to raise the city's storm water design criteria to a significantly higher level while still considering that value to be a 100-year storm. The problems and questions encountered along the way will be listed. For example, is the NOAA Atlas 2 too out of date to be useful? Is it appropriate for a local community to independently evaluate and update climate information normally provided by NOAA? Were the extreme value statistical distributions used to determine rainfall frequencies appropriate for this area? Should point data or regionally composited data be used? Was the data from the National Weather Service Cooperative Weather station on the campus of Colorado State University representative, and should a single value be used for storm water design, or should values change as a funciton of proximity to topographic features (foothills)? In conclusion, the importance of up-to-date rainfall frequency statistics will be addressed along with the need for collecting and archiving accurate and representative short duration rainfall data for communities all across the U.S.