Document Type


Publication Date

January 1969


This investigation has determined the quantities of water stored as groundwater in the three watersheds in Utah, Logan River upstream from State Dam, South Fork Ogden River upstream from Huntsville, and Weber River upstream from Oakley. The proportion of the total streamflow contributed from this groundwater storage has been determined from past streamflow records, and the knowledge obtained from the analysis of groundwater contribution to streamflow has been utilized in developing water supply forecasting techniques and procedures. The first phase of the study dealt with theory and methods from separating the groundwater component from the total streamflow hydrograph. The method used for this separation is based upon the justified assumption that the effluent from groundwater is proportional to the volume of groundwater in storage at anytime and is a modification of the procedure applied by Troxell and others (1954) in a study of Mill Creek watershed in California. An equation giving the component of groundwater flow with parameters whose magnitudes are determined from past flow records results from this separation technique. By utilizing this equation to forecast the groundwater contribution on a continuous basis and for forecasting the surface runoff and interflow components by regression equations similar in form to those presently adopted by The Water Supply Forecast Division, Weather Bureau, ESSA, a forecasting procedure studied by examining how good past forecasts might have been and how the accuracy of these forecasts compares with those obtained by the currently used procedure. A considerable improvement in the accuracy of past forecasts was made possible by this technique. Consequently a third forecasting procedure (referred to in the discussions as Method 3) was developed by using the equation for the groundwater contribution and regression equations for the remaining portion of the streamflow. The latter regression equations were based on data from high watershed snow courses instead of the valley stations used in the presently adopted ESSA method. By separating the groundwater component of flow from the remaining flow in a forecasting procedure, approximately an additional 30 percent of the unexplained squared deviations between observed and forecasted volume of streamflow can be accounted for.