Water users rely on surface and ground water to irrigate crops and maintain wildlife refuges in the 2,200-square-mile Mud Lake study area. Water managers need the ability to evaluate the effects of water-use changes on the future supply of surface and ground water. A five-layer, three-dimensional, finite-difference, numerical ground-water flow model, calibrated to assumed 1980 steady-state hydrologic conditions, was used to evaluate potential effects of seven water-use alternatives on ground-water levels and on losses from and gains to streams and lakes. The model was used to simulate steady-state water levels and ground-water flow for average 1980-90 hydrologic conditions and for seven water-use alternatives that represented changes from average 1980-90 conditions. Five alternatives represented reduced withdrawals from five different sets of wells, the sixth represented increased withdrawals in areas that could potentially support additional irrigation development, and the seventh represented reduced recharge in part of the study area where change from subirrigation to sprinkler irrigation is taking place. Simulated results from each alternative were compared with results for average 1980-90 conditions.

Among the five water-use alternatives in which withdrawals from wells were reduced, simulated water levels were 0.1 to 40 feet higher than average 1980-90 conditions. Simulated stream and lake losses were as much as 4,700 acre-feet less and simulated gains were as much as 19,000 acre-feet greater in response to simulated water-level rises. Simulated underflow into the study area was as much as 8,200 acre-feet less and simulated underflow out of the study area was as much as 91,000 acre-feet greater. Simulated water-level declines were as great as 15 feet for the sixth alternative (increased withdrawals) and 10 feet for the seventh (reduced recharge). Simulated stream and lake losses were as much as 5,700 acre-feet greater and simulated gains were as much as 37,000 acre-feet less for stream and lake segments due to simulated water-level declines. Simulated underflow into the study area was as much as 7,200 acre-feet greater and simulated underflow out of the study area was as much as 23,000 acre-feet less.