摘要: |
Regional regression equations were developed to estimate peak-flow magnitudes using Geographic Information systems (GIS). Peak discharges were estimated at return intervals ranging from 2- to 500-years in Nebraska. Flow data from gaging stations located in or within 50 miles of Nebraska were collected. Regional regression analysis, using weighted-least squares (WLS) regression and data from 273 gaging stations, were used to develop equations for seven hydrologic regions. The WLS regression accounted for the differences in record lengths of the annual peak streamflows between sites. Contributing drainage areas ranged from 0.42 to 6,230 mi (sub 2). The equations can be used to estimate peak discharges for selected return periods at sites without flow data. Digital Elevation Models (DEMs) were the primary data used to extract basin characteristics. The DEMs used in this project are based on 30 m by 30 m data spacing intervals with a Universal Transverse Mercator projection, and are commercially available from the USGS. Morphometric basin characteristics were extracted using ArcInfo software. The DEMs reduced processing time and improve the accuracy of the physical basin characteristics. Soil characteristics were used to improve the accuracy of the regression equations while precipitation data were found to be of lower statistical importance than other characteristics. Regression equations were developed for seven hydrologic regions in Nebraska. Two sets of regression equations were developed for seven hydrologic regions in Nebraska. Two sets of regression equations were developed for each region: one representative of basins with drainage areas of less than 10 mi (sub 2) and one for the complete range of drainage areas. The standard error of estimate for the 10- and 25-year frequency equations ranged from 24 to 93 percent for the complete range of drainage areas. The equations for basins with areas of less than 10 mi (sub 2) had a standard error of estimate for the 10- and 25-year return period of 22 to 75 percent. Based on standard error estimates and comparison with other methods, the regression equations worked best for regions located in eastern Nebraska. The equations for western Nebraska regions do not estimate peak flows as accurately because of insufficient peak flow data and high spatial variability of basin attributes. |