Skip to main content
ARS Home » Research » Publications at this Location » Publication #129577
Title: CUMULATIVE STORM RAINFALL DISTRIBUTIONS: COMPARISON OF HUFF CURVES

Author
item Bonta, James - Jim
item SHAHALAM, ABULBASHER - KUWAIT UNIVERSITY

Submitted to: Journal of Hydrology
Publication Type: Peer Reviewed Journal
Publication Acceptance Date: 12/18/2002
Publication Date: 1/6/2003
Citation: BONTA, J.V., SHAHALAM, A. CUMULATIVE STORM RAINFALL DISTRIBUTIONS: COMPARISON OF HUFF CURVES. JOURNAL OF HYDROLOGY (NZ). 2003. V. 42(1). P. 65-74.

Interpretive Summary: Storm rainfall data measured minutely are seldom available, and methods to simulate storm intensities are needed. One useful method for describing rainfall intensities uses a set of curves called Huff curves. However, methods to develop curves that are stable are needed. Factors such as number of storms affects how stable the curves are (the more measured storms, the less the curves will change). The question is, what is the minimum number of measured storms needed to develop a set of stable curves. A method is needed to compare sets of curves so that one can use the curves with confidence for a practical application. Results are presented on a study of two different methods of comparing curves. The results will be useful to researchers interested in global change studies and weater simulation for simulation of storms and weather simulation models.

Technical Abstract: Short-time increment precipitation intensity data are often available only at sparsely scattered experimental watersheds, which are intensively instrumented, but poorly distributed. Methods are needed to synthesize short-time increment data for watershed modeling. One promising method for doing this is by use of Huff curves - statistical characterizations of rainfall intensities within storms. These curves are developed from 15-mi or hourly data, and can vary depending upon factors that affect their development, such as sample size, storm size, season of year, etc. Methods to determine when curves are stable (curves that appear unaffected by some of these factors), and methods of statistical detection of differences between sets of curves (e.g., region of country, etc.) are needed to determine objectively when Huff curves are stable. The study presents results of an investigation into two possible methods of comparing curves using data from the USA and New Zealand - a Kolmogorov-Smirnov (KS) method and isopleth-comparison method. Both methods yielded different results when investigating the minimum number of storms needed stable curves, with the isopleth method resulting in more storms than the KS method. The isopleth method better estimates the minimum number of storms needed for stable curves. The results are useful to researchers interested in global change studies and weather simulation.