WSP01635 - Laserfiche WebLink

Home Browse Search

OJn474 I. EXAMPLE OF CORRELATION STUDY BY MACffiNE TABULATION PROCEDURE Precipitation data from Delta, Gunnison and Crested Butte were used in a study to attempt to derive forecasting equations for seasonal runoff (Apl'il - July) for the Gunnison River above Gun- nison Tunnel. In making this study it was recognized that the runoff from the Gunnison River was dependent upon factors other than prf>cipitation alone. No attempt was made to "weight" the precipitation according to elevation or area. 1. Objective The purpose of the study was: a. To attempt to develop forecasting equa- tions for seasonal runoff for the Gunnison River. b. To attempt to develop procedures and techniques to be followed using a "refined" clima- tological precipitation data as developed in this study. c. To deduce certain physical facts regard- ing the mechanisms affecting runoff on the Gunnison River. 2. Procedure The procedure for this study was as follows: The seasonal runoff of the Gunnison River was correlated with precipitation from three stations, Delta. Gunnison, and Crested Butte (stations located in and near the Gunnison River drainage area). The following combinations were used. Combinations of stations: Delta, a low elevation station - L Gunnison, a middle level elevation station - M Crested Butte, a high elevation station - H. All possible combinations of stations, L, M, and H, LM, MH, LH and LMH were used for a total of seven combinations. Five estimates of evapotranspiration were used. This first estimate, evapotranspiration estimate A, was the observed precipitation without any deductions for evapotranspiration. Evapo- transpiration estimate B was the same as given in Table III in this report. Evapotranspiration esti- mate C was obtained by subtracting O. 10 of an inch more per storm than the estimates given in Ta- ble III. Evapotranspiration estimate D was obtained by subtracting O. 10 of an inch less per storm than the amounts shown in Table III. Evapotranspiration estimate E was obtained by subtracting 0.20 of an inch more per storm for the low level station, O. to of an inch more per storm for the middle level station and subtracting the same amount for 21 the high level station as the amounts shown in Table III. A total of 34 precipitation periods were analyzed. Period one was to correlate October precipitation only with the following seasonal run- off. Precipitation period two was to use the sum of October plus November. Precipitation period three was to use October plus November plus December, etc. until we get to period ten which was the summation of October plus November-- plus July correlated with the seasonal runoff. Periods 11 through 19 used November alone for period II. November plus December for period 12, etc. until we get to precipitation period 19, which was the sum of all months. November through July. Precipitation period 20 was December alone, precipitation period 21 was December plus January. etc. until we get to precipitation period 27 which was the sum of December plus January plus all months through July. In a similar manner, precipitation periods 28 through 34 were for January through July. The variables used were five evapotranspira- tion estimates, seven station combinations and 34 precipitation periods. The product of 7 x 5 x 34 = 1190 separate combinations. 3. Results For each of these 1190 separate computations the following information was obtained: Equations of the form Y = BO + BIX1 were obtained for single stations. Equations of the form Y = BO + B 1 Xl + B2X 2 were obtained for two stations. Equations of the form Y::z B1 + B1X1 + BtXz + B3X3 were obtained for three stations. Where Y = seasonal runoff, April through July. XI' X2' X3 = precipitation amounts from the three stations. In addition. the correlation coefficient, the constants BO' B1, B 2' B3' the standard error of estimate of Y , and the standard error of estimate for the individual regression coefficients were obtained. 4. Discussion The details of this study are too lengthy to be included in this report. However, the following