ESTIMATION OF PEAK RUNOFF RATES FROM SMALL AGRICULTURAL WATERSHEDS
Abstract
Estimates of peak runoff rates from small agricultural watersheds are necessary for many engineering design applications. Precipitation and runoff data are analyzed in order to develop a method of estimating peak runoff rates. Precipitation data separation techniques and associated spatial and temporal variability are investigated to obtain the critical duration of dry periods between storm events. The effects of critical duration on design storm patterns are also investigated. A comparison of four design storm patterns is made in order to select one for use in a methodology to estimate peak flows. The effects of season of year on peak runoff rate and rainfall characteristics are studied. A methodology to estimate peak runoff rates is then developed. The precipitation analyses show a large difference in critical duration values due to precipitation series separation techniques, but little effect due to critical duration, on Huff curves. Huff curves are shown to be most flexible design storm hyetographs. The growing season is shown to be the season of maximum peak flows. The methodology developed to estimate peak flows uses the CREAMS model (Chemicals, Runoff, and Erosion from Agricultural Management Systems), a weather generator, Huff curves, and random storm depths and durations to find the optimal antecedent soil moisture level at which the probability distribution of measured and simulated peak flows are statistically the same. The CREAMS model is also used to find the probability of occurrence of the optimal antecedent soil moisture. Accurate estimates of peak flow rates and associated return periods can be made with the methodology. Method of storm series separation technique and sampling interval of precipitation data do not affect peak flow return periods, allowing the use of hourly data for watersheds with short times of concentration.
Degree
Ph.D.
Subject Area
Civil engineering|Hydrology
Off-Campus Purdue Users:
To access this dissertation, please log in to our
proxy server.