The Williamson River Basin, located in south-central Oregon, has a drainage area of approximately 3,000 square miles. The Sprague River, which flows into the Williamson River Basin, has a drainage area of 1,580 square miles. Together, the Williamson and Sprague Rivers supply about one-half of the inflow to Upper Klamath Lake. Various statistical techniques, which included trend tests, double-mass curves, and two-sample tests, were used to detect significant changes in the precipitation-runoff relation for the Williamson and Sprague River Basins. Flows from these two rivers were compared with the precipitation and air temperature records collected at Klamath Falls to assess the effect of climate on flow variations.

Most of the double-mass curves showed a major break in the slope of the curve occurring around 1950 and a smaller one near 1990. For the years 1930-50 and 1990-96, February through May flows were relatively lower in the Williamson River than in rivers in nearby basins, by an average of 25,000 acre-feet per year and 36,000 acre-feet per year, respectively, for the 4-month period. From 1950 through 1963, flows were generally higher in the Williamson River compared with the nearby rivers by an average of 38,000 acre-feet for the 4 months. In July through September of 1945-51, 1970-76, and 1992-96, flows were lower in the Williamson River than in the comparison rivers by an average of about 6,000 acre-feet for the 3-month period.

Two-sample statistical tests of the annual flow data sets for the Williamson and Sprague Rivers showed a significant increase in the estimated population mean for the period 1951-96 compared to the estimated population mean for the period 1922-50. However, climate data, which included annual precipitation data from Klamath Falls, Crater Lake, and Medford, and annual air temperature data from Klamath Falls, all showed no significant difference between the two periods.

During the past century, various human land-use activities, such as irrigation, grazing, drainage, and timber harvesting, may have had some impact on the hydrology within the Williamson River Basin. However, relating specific land-use activities to changes in flow is impossible to assess owing to the size and geologic complexity of the basin and to the paucity of historical land- and water-use data for local areas.