Regional Regression Analysis for Small Watersheds inSouthwestern Utah

摘要

Determining discharge in a stream is important to the design of culverts, bridges, and other structures pertaining to transportation systems. Currently in Utah regression equations exist to estimate recurrence flood year discharges for rural watersheds greater than 30 mi~2, and the rational method is used for areas smaller than 0.5 mi~2, however, there are not good methods available to estimate discharges for watersheds that fall between the two approaches. To solve this issue, a flood frequency analysis was conducted on small rural watersheds within the southwest portion of the state of Utah to develop regression equations estimating flood flows for mid-size watersheds. The watersheds selected range from 0.5 mi~2 to 30 mi~2, and have at least 10 years of peak discharges recorded by the USGS. A flood frequency analysis was performed on the peak discharges for each USGS stream gage according to Bulletin 17B guidelines, using the PeakFQ program. Flood flows were regressed against multiple parameters (watershed geometries, soil characteristics, precipitation data, land use data, etc.) to estimate different recurrence flood years (i.e. 2-, 5-, 10-, 25-, 50-, 100-, 200-, 500-year). Regression equations were developed in the format of the rational method (Q = CiA), where the runoff coefficient was regressed against appropriate determined predictor variables, including parameters such as basin slope, basin length, basin average elevation, percent of basin flowing south, composite curve number, saturated hydraulic conductivity, and mean annual precipitation.