Scientific Investigations Report 2006–5212
U.S. GEOLOGICAL SURVEY
Scientific Investigations Report 2006–5212
A 2005 U.S. Geological Survey (USGS) assessment of the Bureau of Reclamation (Reclamation) Klamath Pilot Water Bank program found inconsistencies in the river and canal flow records from sites along the Klamath River between Klamath Falls and Keno, Oregon, for water years 1961–2004. The sum of flows through eight diversions and returns was subtracted from flows at a downstream streamflow-gaging station to compute a surface-water balance for this reach of the river. Four diversion flows in the equation included Lost River Diversion Channel at Station 48, Lost River Diversion Channel at Miller Hill Pumping Plant, North Canal at U.S. Highway 97, and Ady Canal at U.S. Highway 97. Four return flows in the equation included Link River at Klamath Falls, Lost River Diversion Channel at Lost River Diversion Dam, Lost River Diversion Channel at Miller Hill Pumping Plant, and Klamath Straits Drain at F-FF Pumping Plants. The Klamath River USGS streamflow-gaging station at Keno, Oregon (11509500) was the downstream flow-measurement site for the water-balance equation. The water balance showed that the study reach was losing flow in the 1960s and 1970s, and gaining flow in 1980s and 1990s. For the first and second 22-year periods (water years 1961–82 and 1983–2004), the mean annual net water-balance flows were -68,000 and 35,000 acre-feet (acre-ft), respectively. The absolute difference in flows between the two periods was 103,000 acre-feet per year (acre-ft/yr).
Possible explanations for the water-balance inconsistency between the early and late periods include (1) error in the flow-data records included in the water balance, (2) changes in undocumented surface-water inflows and outflows not included in the water balance, and (3) changes in ground-water exchange with the river.
In this study, the data quality of all flow records used in the water balance was evaluated using USGS criteria for flow accuracy. With the exception of the USGS Klamath River at Keno record, which was rated as “good” or “excellent,” the eight other flow records were rated as “poor.” All eight flow records were from non-USGS sites and had insufficient data-collection documentation. Other than for the Link River flow record, there was no evidence of abrupt changes in data-collection procedures in the flow records during the 44-year period that could have explained the water-balance inconsistency. The Reclamation-PacifiCorp Link River record included both river and westside power canal flows. Because of rating curve biases, the river flows could have been overestimated by 25,000 acre-ft/yr on average during water years 1961–82, and underestimated by 7,000 acre-ft/yr on average during water years 1983–2004. For water years 1984–2004, the canal flows were in error by about 11,000 acre-ft/yr. These errors in the Link River flow record (combined river and canal) explained part, but not all, of the water-balance inconsistency.
Smaller undocumented surface-water diversion and return flows along the Klamath River that were in the study area reach and not included in the water-balance equation were evaluated. These diversion flows included individual withdrawals from the Klamath River for irrigation and industrial use. Return flows included drainage from irrigation, storm runoff, discharge from wood-products processing plants, and effluent from wastewater-treatment plants. There was no evidence of abrupt changes in the flows of these diversions and returns that could have explained the water-balance inconsistency. The cumulative magnitude of the undocumented flows also was too small to explain the inconsistency.
Climate data and ground-water level data from wells in or near the study reach did not show an increasing trend during the 44-year period of record that would have caused an increase in ground-water discharge to the river mean annual water year precipitation at Klamath Falls for 1961–82 and 1983–2004 was 13.3 and 13.8 in., respectively, which is a difference of less than 5 percent.
Trends in the eight diversion and return-flow records used in the water balance were evaluated to determine whether known changes in water-management practices in the Klamath Project and Lower Klamath and Tule Lake National Wildlife Refuges during the 44-year period are evident in the flow data. Although trends related to water management cannot explain water-balance inconsistencies (the same trends would be reflected in the most downstream flow record, Klamath River at Keno), the analysis was included to determine whether management changes were apparent in the flow data and whether possible data error could be detected by noting inconsistencies among measurement sites. Many of the water-management changes were implemented in the early 1980s. Mean annual diversion flows at Station 48, North Canal, and Ady Canal during water years 1983–2004 increased over 1961–82 by 16,000, 8,000, and 21,000 acre-ft/yr, respectively. Conversion from flood to sprinkler irrigation actually increased total irrigation use in some of the lands serviced by these canals by enabling production of crops with a higher consumptive use. Additionally, summer crop production was implemented in some fields that had been used solely for winter wheat production. Some of the increased Station 48 and Ady Canal flows were also used to accommodate increased water use in both the Lower Klamath and Tule Lake refuges. The conversion from flood to sprinkler irrigation also resulted in a significant reduction in return flows in the 1980s and 1990s. Mean annual flows at the Lost River Diversion Channel at Lost River Diversion Dam and the Klamath Straits Drain at F-FF Pumping Plant for water years 1983–2004 decreased by 31,000 and 27,000 acre-ft, respectively, compared with water years 1961–82. Releases through the Link River Dam in the 1990s also were reduced to maintain minimum lake elevations for fish habitat.
The probable cause of the water-balance inconsistency was error in the eight diversion and return-flow data records; however, data-collection documentation during the 44-year period was insufficient to determine which flow record, or records, contained most of the error. Some of the 103,000 acre-ft water-balance inconsistency can be explained by a 32,000 acre-ft error in the river flow portion of the Link River flow record. Another portion can be explained by an 11,000 acre-ft/yr error in the westside power canal flow record, included in the Link River flow record, for water years 1984–2004. The cause of the remaining 60,000 acre-ft of error could not be determined. Some of it could still be in the Reclamation-PacifiCorp flow record for the Link River, and some might be distributed among the seven other flow records. Coincidentally, the error bias in all eight flow records was consistent among the sites and with known water-management changes in the Klamath Project and National Wildlife Refuges.
From the analysis of flow data in this study, it is evident that an improved water measurement network is warranted to better understand both natural and anthropogenic impacts on the hydrologic system. In particular, this analysis described errors in the Link River flow record that can be corrected in Reclamation’s hydrologic database to ensure that any historic analyses of these data and the computation of Upper Klamath Lake monthly net inflows are based on sound hydrologic data.
U.S.
Department of the Interior | U.S. Geological
Survey
Persistent URL: https://pubs.water.usgs.gov/sir20065212
Page Contact Information: Publications Team
Page Last Modified: Thursday, 01-Dec-2016 19:16:47 EST
