Summary


Since 2001, water-management goals in the upper Klamath Basin, Oregon and California, have shifted to protect aquatic habitat. As a result, groundwater has become an important supplemental water supply for irrigators in the Bureau of Reclamation Klamath Irrigation Project (Project), particularly during periods of drought. The potential effects of increased groundwater development on groundwater and surface-water resources have caused concern among stakeholders in the basin. This report describes the development and application of a groundwater-management model for the Project that was designed to simultaneously consider the groundwater demands of the Project and the adverse effects of concern to resource-management agencies.


The overall goal of the groundwater-management model was to determine the volumes and patterns of groundwater pumping that, to the extent possible, meet the supplemental groundwater needs of the Project. The objective of the model was to maximize irrigation-season pumping at 117 wells that historically have been used to provide supplemental groundwater to the Project. Total withdrawals were limited by the groundwater demand foreseen under the Klamath Basin Restoration Agreement (KBRA) with a set of constraints defined to limit the adverse effects of supplemental groundwater pumping. These constraints were formulated to (1) limit reductions in groundwater discharge to streams and lakes that provide critical aquatic habitat, (2) control drawdowns throughout the Project and adjacent areas, and (3) limit reductions in groundwater discharge to the Project agricultural drain system.


Six sets of applications of the groundwater-management model were completed to determine the limits of groundwater development for various groundwater-management scenarios. Baseline conditions were defined using the period 1970–2004 and supplemental groundwater demand by the Project during this period was defined as the part of estimated irrigation demand that would not be satisfied by the surface-water diversion allowed under the Klamath Basin Restoration Agreement. The results of the groundwater-management modeling analyses are summarized in the following table (shading shows constraint limits that are changed from base-case values).


Groundwater- management strategy	Constraint limits												Optimized annual withdrawal (acre-feet)			Years groundwater demand is met, out of 22 years
	Seasonal drawdown (feet)			Year-to-year drawdown (feet)			Decadal drawdown (feet)			Change in groundwater discharge (percent baseline)
	Oregon	California		Oregon	California		Oregon	California		Drains	KBRA		Maximum annual withdrawal	Average annual withdrawal in years with unmet demand	Maximum annual groundwater deficit
Base case	25	25		3	3		25	25		20	6		60,458	54,409	44,833	14
DD_1	25	37.5		3	4.5		25	25		20	6		78,401	69,687	29,718	17
DD_2	25	50		3	6		25	25		20	6		84,635	75,565	28,627	18
DD_3	20	20		2	2		20	20		20	6		42,721	32,727	68,510	9
Drain_10	25	25		3	3		25	25		10	6		54,838	41,775	67,378	13
Drain_05	25	25		3	3		25	25		5	6		36,062	24,466	81,930	7
Add pumping capacity	25	25		3	3		25	25		20	6		83,951	68,056	26,593	17
Remove background pumping	25	25		3	3		25	25		20	6		69,996	65,030	32,469	14

The first application of the groundwater-management model limited the reductions in groundwater discharge to critical aquatic habitat to be no more than 6 percent, as defined in the KBRA; applied drawdown limits of 25 feet (ft) for seasonal drawdown, 3 ft for year-to-year drawdown, and 25 ft for 10-year drawdown, as required by the Oregon Water Resources Department; and limited the reductions in groundwater discharge to the Project’s drain system to be no more than 20 percent. Results indicated that the supplemental irrigation wells in the Project would be able to meet groundwater demand in 14 of the 22 years studied with supplemental groundwater needs, with a maximum calculated groundwater withdrawal of about 60,000 acre-feet (acre-ft). Significant irrigation shortages would remain in the years with unmet groundwater demand; average annual withdrawals for those years was about 54,000 acre-ft and groundwater-pumping deficits ranged from 11,000 to 45,000 acre-ft. The patterns of groundwater withdrawal calculated by the groundwater-management model indicated that supplemental groundwater pumping should be concentrated in the Tule Lake and Klamath Valley regions of the Project. The highest annual withdrawals occurred in the southern Tule Lake wells, which provided the highest percentage of pumping in most years and supplied about 22,000–28,000 acre-ft in the years with unmet groundwater demand. Calculated withdrawals from the Klamath Valley wells totaled about 10,000–20,000 acre-ft in the years with unmet groundwater demand, and the withdrawals from the northern Tule Lake wells totaled about 8,700–14,000 acre-ft in years with unmet demand.


Subsequent analyses evaluated the sensitivity of the optimization results to several factors. Sensitivity analyses demonstrated that substantial increases in total groundwater withdrawal could be achieved by increasing the seasonal and year-to-year drawdown limits in the California part of the Project. When the seasonal and year-to-year drawdown limits in the California part of the model were increased to 37.5 and 4.5 ft, respectively, the maximum annual withdrawal was about 78,000 acre-ft, the average annual withdrawal in years with unmet groundwater demand was about 70,000 acre-ft, and the maximum groundwater deficit was about 30,000 acre-ft. Alternatively, when the seasonal and year-to-year drawdown limits were increased to 50 and 6 ft, respectively, the maximum annual withdrawal was about 85,000 acre-ft, the average annual withdrawal in years with unmet groundwater demand was about 76,000 acre-ft, and the maximum amount of unmet groundwater demand was about 29,000 acre-ft. In both cases, the increased withdrawals occurred primarily in the southern Tule Lake well group, which is entirely in the California part of the model.


A second sensitivity analysis evaluated the effects of imposing more restrictive drawdown limits across the entire Project. The groundwater management model calculated substantial reductions in annual withdrawals when the seasonal, year-to-year, and 10-year drawdown limits were reduced to 20, 2, and 20 ft, respectively. The maximum annual withdrawal was calculated to be about 43,000 acre-ft and the average annual withdrawal during years with unmet groundwater demand was about 33,000 acre-ft. In comparison to the base-case analysis, total annual withdrawals calculated for the years with unmet groundwater demand were reduced by as much as 27,000 acre-ft, with the largest reductions in the Klamath Valley and southern Tule Lake well groups. With the more restrictive drawdown limits, expected groundwater needs were met in only 9 of the 22 years studied and the maximum amount of unmet groundwater demand was about 69,000 acre-ft.


A third set of model applications evaluated the influence of uncertainty associated with the constraints controlling groundwater discharge to the Project agricultural drain system that collects irrigation return flows and routs them to downslope irrigators and wildlife refuges. A two-part sensitivity analysis determined the change in pumping volume and distribution associated with a more restrictive constraint on the allowable reduction in groundwater discharge to the Project drain system. The groundwater-management model calculated substantial reductions in annual withdrawals when the drains-constraint limit was reduced below the base-case limit of 20 percent. When the drains-constraint limit was set to 10 percent, the maximum annual withdrawal is about 55,000 acre-ft, the average annual withdrawal in years with unmet groundwater demand was about 42,000 acre-ft, and the maximum groundwater deficit was about 67,000 acre-ft. With the constraint limit set to 5 percent, the maximum annual withdrawal was about 36,000 acre-ft, the average withdrawal in years with unmet groundwater demand was about 25,000 acre-ft, and the maximum amount of unmet groundwater demand was about 82,000 acre-ft. The patterns of groundwater withdrawal calculated by the groundwater-management model indicated that the largest reductions in pumping occur in the southern Tule Lake well group, which generally have the highest annual pumping volumes in the base-case model.


Additional applications of the groundwater-management model evaluated the trade-offs associated with increasing the pumping capacity of the network of wells serving the Project and reducing background pumping at wells in the vicinity of the Project. A sensitivity analysis assessed the potential value of additional pumping capacity at existing decision wells by increasing the well-capacity constraint limits and re-solving the groundwater-management model. The optimization model identified the potential for increased pumping at a small set of wells in the Klamath Valley well group in Oregon. The maximum annual withdrawal for this analysis was about 84,000 acre-ft and the average annual withdrawal in years with unmet groundwater demand is about 70,000 acre-ft. Similarly, the model identified the potential to increase annual pumping amounts by removing off-Project background pumping. Simulated background pumping was removed from areas that were likely to impact groundwater levels at locations with binding drawdown constraints. The optimization model was then re-solved to determine the change in pumping volumes at on-Project wells. The results of the updated groundwater-management model indicated that increased withdrawals, primarily from the northern Tule Lake region, would result from the removal of background pumping. The maximum annual withdrawal was about 70,000 acre-ft and the average annual withdrawal in years with unmet groundwater demand was about 65,000 acre-ft.


The results of the groundwater-management model presented in this report demonstrate the value of a coupled groundwater-simulation and optimization model for evaluating groundwater-management issues in the upper Klamath Basin. The groundwater-management model provides an improved understanding of the capacity for, and limitations of, increased supplemental groundwater pumping. The results of the groundwater-management model demonstrate that groundwater can serve as an important supplemental irrigation supply for the Bureau of Reclamation Klamath Irrigation Project and that supplemental groundwater use by the Project can be managed to avoid adverse effects to groundwater discharge that supports aquatic habitat. Groundwater-management results presented in this report may need to be modified to account for model uncertainty and climate conditions not represented in this study. As a result, next steps in the application of a groundwater-management strategy in the upper Klamath Basin should follow an adaptive management approach that includes a groundwater monitoring program designed to track the effects of on-Project supplemental groundwater pumping.


First posted April 23, 2014