An evaluation of the effect of groundwater pumpage on the infiltration rate of a semipervious streambed

A quantitative evaluation of the effect of ground water pumpage on the Arkansas River in southeastern Colorado using both groundwater and surface‐water data has shown that, in at least one reach of the valley, pumping has lowered the water table below the level of the apparently pervious streambed, thereby breaking the hydraulic connection between the stream and the water table. The river is gaining except in reaches where withdrawals of water from the valley fill are large. However, a large cone of depression has formed directly beneath the stream because of continuous withdrawal of about 10 cfs of groundwater at Lamar. The cone extends under the losing reach of the stream without appreciable distortion and has been as much as 12 feet below the streambed. In the 1‐mile reach immediately upstream from the wells, the stream lost about 2 cfs during each of three sets of observations. Once the water table is lowered below the streambed, hydraulic connection is broken, and changes in depth to water table have no measurable effect on the rate of depletion of streamflow. The major control on infiltration loss, where hydraulic connection is broken, is probably the permeability of the least permeable layer of the streambed. Studies of stream reaches that have beds with wide ranges of grain size and sorting suggest that the range in maximum infiltration rates is small if the streambeds are undistributed and depth of water in the stream is less than about 1 ft. A maximum infiltration rate of 20 gal/ft 2 /day is suggested as a reasonable first approximation in the absence of observed rates.