Upper Colorado River Commission ; Third Annual Report

page-41

It is not difficult to account for the large increase in the actual discharge of this stream above the estimate. No satisfactory method of predicting summer precipitation has yet been found and therefore when the pattern of summer rainfall differs materially from the average, the difference between forecast and reality will be greater than the limits indicated. Last summer in April, before the irrigation season was more than started, the rainfall at Glen-wood Springs was more than 40% above normal for that month while at Aspen it was nearly 20 % above normal. Again in August when irrigation demands are relatively small, precipitation in these areas was even more above normal than in April (60% and 30% respectively) . This illustrates the uncertainties that are always present in making predictions of this sort. If April runoff and rainfall had been taken into account and the forecast were made early in May, it is probable that this discrepancy would have been materially reduced. The engineering department of the Commission seeks to derive similar forecasting formulas using only 5 months of winter runoff and precipitation records with snow conditions as of March 1st for the first annual forecasts and Later, 6 and 7 months runoff and precipitation with snow as of April 1st and May 1st for more dependable forecasts. When this has been done for the main stem of the Colorado River and the principal tributaries in each state, so that forecasts of stream flow within and at the border of each state and at Lee Ferry can be made, this problem will be considered completed for some time, except for refinements as more data become available. -41-