The hydrology and geology of Williams Lake watershed was studied to evaluate the accuracy of various methods used to determine precipitation and evaporation in lake water-balance studies and to define a lake and ground-water system according to approaches suggested by theoretical modeling studies. Regression analysis between estimated and measured precipitation at the lake showed that the accuracy of regionalization techniques is dependent on the closeness of the data network to the lake. For individual storms, the average-value method was found to be better than either the weighted average or isohyetal methods of determining precipitation, but it was least accurate in estimating 14-day average precipitation. The amount of evaporation calculated by the mass-transfer method ranged from 2 to 7 inches per month from July to October 1978, depending on the method used to determine the mass-transfer coefficient. Test drilling indicated that 30 to 150 feet of sand and gravel overlies till in the Williams Lake watershed. A sand lens about 50 feet thick occurs within the till.
The configuration of the water table and vertical-head gradients measured from July to December 1978 indicate that ground water moves into the lake from the south and east and moves from the lake into the ground-water reservoir to the west. Preliminary numerical models indicate that the sand lens within the till is effectively isolated from the flow system interacting with the lake and that both inseepage and outseepage were about 1.4 inches from mid-July to mid-October 1978. When estimated as a residual in a water balance, ground water showed a net outseepage only of 1.47 inches.
