The water resources of the upper Klamath Basin, in southern Oregon and northern California, are managed to achieve various complex and interconnected purposes. Since 2001, irrigators in the Bureau of Reclamation Klamath Irrigation Project (Project) have been required to limit surface-water diversions to protect habitat for endangered freshwater and anadromous fishes. The reductions in irrigation diversions have led to an increased demand for groundwater by Project irrigators, particularly in drought years. The potential effects of sustained pumping on groundwater and surface-water resources have caused concern among Federal and state agencies, Indian tribes, wildlife groups, and groundwater users. To aid in the development of a viable groundwater-management strategy for the Project, the U.S. Geological Survey, in collaboration with the Klamath Water and Power Agency and the Oregon Water Resources Department, developed a groundwater-management model that links groundwater simulation with techniques of constrained optimization.

The overall goal of the groundwater-management model is to determine the patterns of groundwater pumping that, to the extent possible, meet the supplemental groundwater demands of the Project. To ensure that groundwater development does not adversely affect groundwater and surface-water resources, the groundwater-management model includes constraints to (1) limit the effects of groundwater withdrawal on groundwater discharge to streams and lakes that support critical habitat for fish listed under the Endangered Species Act, (2) ensure that drawdowns do not exceed limits allowed by Oregon water law, and (3) ensure that groundwater withdrawal does not adversely affect agricultural drain flows that supply a substantial portion of water for irrigators and wildlife refuges in downslope areas of the Project. Groundwater-management alternatives were tested and designed within the framework of the Klamath Basin Restoration Agreement (currently [2013] awaiting authorizing Federal legislation), which would establish a permanent limit on the amount of surface water that can be diverted annually to the Project. Groundwater-management scenarios were evaluated for the period 1970•2004; supplemental groundwater demand by the Project was estimated as the part of irrigation demand that would not have been satisfied by the surface-water diversion allowed under the Klamath Basin Restoration Agreement. Over the 35-year management period, 22 years have supplemental groundwater demand, which ranges from a few thousand acre-feet (acre-ft) to about 100,000 acre-ft in the driest years.

The results of the groundwater-management model indicate that supplemental groundwater pumping by the Project can be managed to avoid adverse effects to groundwater discharge that supports critical aquatic habitat. The existing configuration of wells in the Project would be able to meet groundwater-pumping goals in 14 of the 22 years with supplemental groundwater demand; however, substantial irrigation shortages can be expected during drought periods when the demand for supplemental groundwater is highest. The maximum irrigation-season withdrawal calculated by the groundwater-management model is about 60,000 acre-ft, the average withdrawal in drought years is about 54,000 acre-ft, and the amount of unmet groundwater demand reaches a maximum of about 45,000 acre-ft. A comparison of optimized groundwater withdrawals by geographic region shows that the highest annual withdrawals are associated with wells in the Tule Lake and Klamath Valley regions of the Project. The patterns of groundwater withdrawal also show that a substantial amount of the available pumping capacity is unused due to the restrictions imposed by drawdown constraints.

Subsequent model applications were used to evaluate the sensitivity of optimization results to various factors. A sensitivity analysis quantified the changes in optimized groundwater withdrawals that result from changes in drawdown-constraint limits. The analysis showed the potential for substantial increases in withdrawals of groundwater with less restrictive drawdown limits at drawdown-control sites in the California part of the model. Systematic variation of the drains-constraint limit yielded a trade-off curve between optimized groundwater withdrawals and the allowable reduction in groundwater discharge to the Project drain system. Additional model applications were used to assess the value of increasing the pumping capacity of the network of wells serving the Project, and the relation between reduced off-Project groundwater pumping and increased pumping by Project irrigators.