| OCR Text |
Show 1952 spring inflow to Lake Mead are greater than on January 1. As of February 1, the following flow in the Colorado River near Grand Canyon, Arizona, for the period April through July 1952, based on an average October through January precipitation of 8.09 inches at thirteen key stations on the upper watershed, can be anticipated: Maximum 20,200,000 acre-feet Mean 17,000,000 acre-feet Minimum 13,800,000 acre-feet The above mean figure has been obtained by reducing the value of 17.5 million acre-feet, as obtained from the least square line on the attached February 1 forecast chart, by 0.5 million acre-feet, which is the estimated depletion during the April-July 1952 period by transmountain diversions and upstream reservoirs constructed since closure of Hoover Dam (1935). The probability is 9 chances in 10 that the actual flow at Grand Canyon will fall between the above-mentioned minimum and maximum amounts. Actual runoff during the period April-July 1951 was 6,284,000 acre-feet." Forecast of the mean flow for May, 1951 was 6,700,000 acre-feet -an error of only 614%. Since the correlation between the inflow to Lake Mead and the discharge of the Colorado River at Lees Ferry is very good indeed, these forecasts, together with those of the Weather Bureau, could be used by the Commission today if an emergency should suddenly arise. It should be noted, however, that the Bureau of Reclamation's forecasts depend on only 13 stations in the Upper Basin. None of these are located in the basins of either the Green River in Wyoming or the White River in Colorado. Two stations were selected in the Colorado River basin above Grand Junction and two are in the San Juan basin. In contrast to this, four of the 13 stations used are in the basin of the Gunnison, which would seem to give rather undue weight to the discharge of this tributary, although its importance as one of the largest contributors to the outflow of the Upper Colorado River is fully recognized. These and other questions should be fully analyzed in the investigation of the problems of Forecasting Stream Flow. -38- |
