EIS Arapahoe & Roosevelt National Forest, Pawnee National Grassland

Home Browse Search

. .? SPECIALIST REPORT ON WATER YIELD ? Routt Forest Pian Revision WATER YIELD METHODS BASELINE WATER YIELD Baseline water yield for the Routt National Forest was determined using USGS gage station records from fourteen gages which are on or near the forest. Only gages which represent relatively undisturbed basins, or which had records prior to significant anthropogenic distur- bances were used. All gages had a minimum of ten yeai•s of record. The Quattro Pro spread- sheet program (Borland, 1993) was used for all statistical analysis. The first step was to determine annual water yield in feet per acre for watersheds with USGS stream gages. This was based on the average annual discharge divided by the drainage azea. Results were highly variable suggesting that other factors such as the average annual precipita- tion influenced the quantity of water produced in different basins. To determine the relationship between annual precipitation and baseline water yield, the average annual runoff was regressed against the average annual precipitation for those watersheds with USGS gages. The average annual precipitation was determined using the Colorado Average Annual Precipitation 1951-1980 1:500.000 scale map (Colorado Climate Center, 1984). The isohyetal lines from this map were digitized into the GIS system and then overlayed with the sixth level watersheds. The acres within each precipitation zone were used to calculate a weighted mean annual precipitation for the sixth level watersheds with USGS stream gages. For gages which did not lie on the sixth level watershed boundary, estimates were made of how much of the area fell within different precipitatioii ranges. In this situation, generally all of the acres in the higher precipitation ranges were considered to be part of the watershed, and acres within the lower precipitation ranges were reduced. For gages below the forest boundary, estimates were made of how much of the non-USFS land was in a given precipitation zone using the precipitation map cited above. Generally the non-USFS land was in the lower precipitation ranges. Regressing the mean weighted annual precipitation against baseline water yield developed a regression model with mean annual precipitation as the independent variable, and water yield in feet per acre as the dependent variable. The regression model had an R-square of 0.87, and a standard enor of 0.3 feet. The high R-square indicates that the weighted mean annual precipita- tion explains 87 percent of the variability in water yield from undisturbed basins. For the purposes of the forest plan, baseline water yield was modeled for the fourth level watersheds which include the Litde Snake, North Platte, Upper Colorado, and Yampa river basins. A weighted mean annual precipitation for each of the major river basins was calculated based on the number of acres in each precipitation zone. The resulting weighted mean annual precipitation was then plugged into the regression model developed from the USGS stream gage ?