Scientific Investigations Report 2006–5036
U.S. GEOLOGICAL SURVEY
Scientific Investigations Report 2006–5036
As part of this assessment, open-water evaporation and evapotranspiration losses from areas of emergent and terrestrial vegetation in the refuges were quantified for the purpose of comparison with measured inflows and outflows to and from the refuges. This comparison allowed the estimation of water use in the refuges by two different methods.
Although more than 90 percent of the land area of the two refuges is supplied in some way by Project water, a few upland units and some small land parcels, which are separate from the main bodies of the refuges, were not included in this assessment because they do not receive Project water.
An estimation of total water use in the refuges was made by summing water use from four land-use categories: grain, seasonal wetland, permanently flooded wetland with emergent vegetation, and open water. Approximate monthly and annual water-use rates for the land-use categories, derived from recent studies by the USFWS and Burt and Freeman (2003), are shown in table 1. These rates are multiplied by total acreage of the land-use categories to derive total water-use volumes.
Approximately 55 percent of Tule Lake and 27 percent of Lower Klamath refuge land areas are dedicated to grain production. The USFWS estimated that 2.5 ft/yr was an approximate value of water use for grain production lands (Tim Mayer, U.S. Fish and Wildlife Service, Portland, Oregon, written commun., 2005). This estimate was based on data in the State of Oregon’s hydrology report for the Klamath Basin (Oregon Water Resources Department, 1971), local irrigation district delivery guidelines, and the USFWS assessment of water delivery needs for marsh vegetation. Burt and Freeman (2003) also made estimates of annual evapotranspiration for 1999, 2000, and 2001 for barley, oats, and wheat that ranged from 2.22 to 2.64 ft for the Tule Lake and Lower Klamath refuge areas. On the basis of these studies, 2.5 ft/yr was selected as an appropriate annual water-use rate for grain production areas. Mean monthly reference evapotranspiration rates estimated for the Klamath Falls BOR AgriMet station (KFLO), located approximately 15–20 mi north of the refuges, were used to partition the annual grain water use into monthly rates, as shown in table 1 (Bureau of Reclamation, 2004).
Roughly half of the Lower Klamath refuge and 5 percent of the Tule Lake refuge contain seasonal wetlands, which are flooded (sometimes referred to as “floodup”) generally during the autumn (September-November). They are dewatered in the late spring and early summer through a combination of drainage and evaporation. Each seasonal-wetland unit has a target water level associated with it. During the autumn floodup, when sufficient water deliveries are sent to the refuges, water levels are manipulated in each unit until the target level is reached (Tim Mayer, U.S. Fish and Wildlife Service, Portland, Oregon, written commun., 2005). Mayer and Thomasson (2004) measured the volume of water needed to fill three representative seasonal-wetland units in the Lower Klamath refuge. In their study, this volume of water was also partitioned into soil saturation requirements, surface-water volume, and evapotranspiration losses. On average, the floodup volume for the units was 2.90 ft from September-November (table 1). Of this amount, about one-half (1.5 ft) is used to saturate the soils, and roughly 0.5 ft is lost through evaporation and evapotranspiration during October–April. In late spring, the remaining 0.9 ft of delivered floodup water, plus approximately 0.9 ft of accumulated precipitation that falls between October and April, is drained from the seasonal units and leaves the refuges.
Permanent (year-round) wetland units encompass approximately 23 percent of the Lower Klamath refuge and 40 percent of the Tule Lake refuge. Two rates were used to estimate water use in these units. The emergent-vegetation water-use rate was applied to the vegetation-dominated portion of each unit. The open-water evaporation rate (described below) was applied to the open-water portions.
The rate of water use by emergent vegetation was estimated using a USFWS model based on the Priestly-Taylor evapotranspiration equation, which uses minimum and maximum daily air temperatures as input (Ronald R. Thomasson, U.S. Fish and Wildlife Service, Portland, Oregon, written commun., 2005). The model was calibrated using measured field data from energy budget studies made at several locations in the Klamath National Wildlife Refuges (W.R. Bidlake and K. Payne, U.S. Geological Survey, written commun., 1998). Using a simulation period of 1961–1990, the model predicted a median April–October total evapotranspiration loss of 2.41 ft. This value and monthly reference evapotranspiration rates estimated for the Klamath Falls BOR AgriMet station (KFLO) were used to estimate evapotranspiration losses for the months of November through March (Bureau of Reclamation, 2004). The annual evapotranspiration rate for the emergent vegetation category was then estimated to be 2.63 ft (table 1).
Pan evaporation data from the Tule Lake Agricultural Station were used to estimate evaporation from open-water portions of the permanent wetlands. The data were from April through October 1956–1981. Average total April–October pan evaporation for the period of record was 4.08 ft. Because the open-water areas on the refuges are small and shallow, a pan evaporation coefficient of 0.9 was used (Dunne and Leopold, 1978). This produced a mean April–October open-water evaporation rate of 3.66 ft, as shown in table 1. Year-round monthly pan evaporation data collected at the Klamath Falls Agricultural Station (located farther away from the refuges than the Tule Lake Agricultural Station) were used to estimate open-water evaporation for the months of November through March. On average, the Tule Lake pan evaporation rates are 95 percent of the Klamath Falls pan evaporation rates due to climatic differences. Therefore, mean November–March Klamath Falls pan evaporation data were simply multiplied by 0.95 to estimate Tule Lake pan evaporation data for those months. Those estimates were then adjusted further by using a 0.9 pan evaporation coefficient to estimate open-water evaporation. The mean annual rate for open-water evaporation was estimated to be 4.07 ft (table 1).
In addition to estimating evapotranspiration losses, it also is necessary to account for the additional water in the refuge units needed to prevent salinity accumulation over time. Specific conductance data collected at irregular intervals by both the BOR and USGS at various inflow and outflow locations near the refuges from the early 1990s to the present show that specific conductance levels have remained relatively constant over time (MacCoy, 1994; Cindy Williams, Bureau of Reclamation, written commun., 2005). The volume of additional water needed to avoid exceeding a maximum salinity level for flows leaving a water body can be estimated by multiplying the net evapotranspiration times a salinity factor. Net evapotranspiration is estimated as evapotranspiration minus precipitation. The salinity factor is estimated as the maximum outflow specific conductance level divided by the difference between outflow specific conductance and measured inflow specific conductance.
The Lost River is a major point of inflow for the Tule Lake refuge. Specific conductance data collected on the Lost River at State Line Road has a mean of approximately 0.5 millisiemens per centimeter. To prevent outflow specific conductance from exceeding 3.0 millisiemens per centimeter, a salinity factor of 1.2 would be needed. For the Lower Klamath refuge, mean inflow specific conductance for the Ady Canal at State Line Road and Pumping Plant D is 0.14 and 0.60 millisiemens per centimeter, respectively. Salinity factors based on these specific conductance levels were computed as 1.16 and 1.25, respectively. On the basis of these calculations, an overall salinity factor of 1.20 (20 percent of the net evapotranspiration) seemed reasonable to apply to both refuges. The U.S. Fish and Wildlife Service also estimated a salinity factor of 1.20 for both refuges in their analyses (Tim Mayer, U.S. Fish and Wildlife Service, Portland, Oregon, written commun., 2005).
The Lower Klamath refuge is divided into approximately 50 land-use units. A breakdown of the cumulative acres by area and land-use category for 2003, 2004, and 2005 is shown in table 2. Acreage data for each land-use category are from the annual USFWS habitat plans for the refuge. However, the total number of acres shown in table 2 does not include all of the Lower Klamath refuge. Some small, noncontiguous units and nonirrigated upland areas are not included because precise data for both refuges was not easily obtainable. However, the table does show totals for essentially all units that typically are irrigated for a particular land use. Approximately 35,000 acres, most of the refuge, is located in California south of State Line Road. With the exception of approximately 5,000–7,000 acres, most of this area is irrigated by Project water. Some units that are not irrigated with Project water are irrigated with natural surface-water runoff from Sheepy, Cottonwood, and Willow Creeks and ground-water discharge. An additional 6,600 acres of the refuge are located north of State Line Road. Known as the Area K Lease Lands, these units are leased to farmers on an annual basis for grain production and are irrigated with Project water. The water for these units comes from the Ady Canal at various diversion points upstream of the Ady Canal flow gage on State Line Road.
The main area of the refuge, south of State Line Road, contains a mix of land-use categories. Most of these units are managed as seasonal wetlands. However, some units are permanent open-water wetlands, and other units are leased for grain production. With the exception of permanent open-water units, many units in the refuge change from one land use to another on an annual basis. Table 3 shows water-use estimates for the Lower Klamath refuge derived by using annual evapotranspiration and evaporation rates for grain, seasonal wetland, emergent vegetation, and open water shown in table 1. For the largest open-water permanent wetland unit (Unit 2), the emergent vegetation evapotranspiration rate was applied to 83 percent of the area, and the open-water evaporation rate was applied to the remaining 17 percent. A mixture of other emergent vegetation and open-water ratios were applied to the other permanent wetland units (Tim Mayer, U.S. Fish and Wildlife Service, Portland, Oregon, written commun., 2005).
The mean annual water use for the entire Lower Klamath refuge for 2003–2005 (including seasonal-wetland water use) was approximately 124,000 acre-ft (table 3). This volume includes water use in all units served by Project water on both sides of State Line Road plus approximately 5,000–7,000 acres of land not served by Project water. However, the 3-year mean annual water use for refuge lands south of State Line Road, which are served by the Project water deliveries, was approximately 89,000 acre-ft.
To estimate needed annual water deliveries for the refuge, estimated annual water use was reduced by annual precipitation and adjusted for salinity flushing. For water years 2003, 2004, and 2005, annual precipitation, measured nearby at the BOR AgriMet station in Worden, Oregon (WRDO), near the northwestern corner of the refuge, was 10.1, 8.4, and 10.6 in., respectively (Bureau of Reclamation, 2004). Needed annual water deliveries for each of the three main refuge lands (Project, Area K, non-Project) are shown in table 3. The needed mean annual water delivery for the entire refuge for the 3-year period was estimated to be 107,000 acre-ft.
The total number of acres by sump and by land-use category for the Tule Lake refuge for 2003, 2004, and 2005 are shown in table 4. Because the cumulative size of each land-use category continually changes, a better representation of recent conditions can be made using the last 3 years of data instead of just 2005. Acreage data for each of the land-use categories are from the annual USFWS habitat plans for the refuge. The total number of acres each year for each sump and the sum of all the sumps are not necessarily the same each year. Differences can be attributed to mapping error or to small units that were included or not included in the habitat plan for a given year. The total number of acres shown also does not include all refuge property. Small, noncontiguous units; the Area J unit; and narrow upland strips located along the borders of Sumps 1A, 1B, and 2 were not included because they do not receive Project water.
Sump 1A was managed as a permanent wetland in all 3 years. Approximately 10 percent of the unit was defined as emergent vegetation and the remaining 90 percent as open water (Tim Mayer, U.S. Fish and Wildlife Service, Portland, Oregon, oral commun., 2005). To estimate water use in the sump, the annual rates of emergent-vegetation evapotranspiration (2.63 ft) and open-water evaporation (4.07 ft) were applied to their associated portions of the sump for each year. The 3-year average water use in the sump was approximately 36,500 acre-ft (table 5).
All of Sump 1B was managed as permanent, open-water wetland in 2003. However, in 2004–2005, the sump was managed as both a seasonal and permanent wetland. Appropriate rates shown in table 1 were used to compute water use for each year. To estimate 2004–2005 water use, the annual emergent-vegetation-evapotranspiration rate (2.63 ft) was applied to 25 percent of the sump and the annual open-water evaporation rate (4.07 ft) was applied to the other 75 percent of the area. The 3-year mean annual water use in Sump 1B was approximately 11,400 acre-ft.
The 3-year mean annual water use in Sumps 2 and 3 was approximately 16,000 and 31,900 acre-ft, respectively. Most of Sumps 2 and 3 are managed for grain. Water use in these units was based on the annual evapotranspiration rate of 2.5 ft. However, Sumps 2 and 3 also contain seasonal and permanent wetlands, which change in area from year to year. Water use for those units was computed using the annual rates shown in table 1. Because the permanent wetland units in these sumps are much shallower than Sump 1A, a 1:1 ratio of emergent-vegetation and open-water rates was applied to these units.
The 3-year mean annual water-use volume for the entire refuge, including seasonal-wetland, open-water, and agricultural water use, was approximately 95,900 acre-ft (table 5). To estimate needed annual water deliveries for the refuge, estimated annual water use was reduced by annual precipitation. For water years 2003, 2004, and 2005, annual precipitation, measured at the nearby Tule Lake Agricultural Station in Tule Lake, California, was 10.4, 7.4, and 13.1 in., respectively (California Department of Water Resources, 2005). After subtracting precipitation from the water-use volumes, 20 percent was added to account for salinity flushing. The estimated needed water deliveries for each of the sumps are shown in table 5. The needed mean annual water delivery for the entire refuge for the 3-year period was estimated as about 82,800 acre-ft.
Various factors contribute uncertainty in estimating water-use and water-delivery needs based on open-water evaporation and evapotranspiration estimates. The accuracy of the evaporation and evapotranspiration estimates described earlier could be improved if pan evaporation data were collected at both refuges. The rates presented herein are based on average conditions. In reality, evaporation and evapotranspiration rates vary from year to year. The rates presented in this report also are based on evaporation and air temperature data that were not collected in recent years nor are local to the refuges. The open-water evaporation rate is based on pan evaporation data collected from 1956 to 1981. Wind conditions also significantly affect evaporation and evapotranspiration rates. As a result, evaporation and evapotranspiration rates can vary among different locations within the two refuges because of geographic variation in wind conditions. Overall accuracy of the evapotranspiration estimates also could be improved with specific evapotranspiration rates for different types of grain vegetation in the refuge. However, acquiring more specific crop information was not possible within the scope and timeline of this project.
Adjusting needed water deliveries to account for salinity flushing over time introduces additional uncertainty. The salinity factor used in the study was based on measured specific conductance data collected at various locations within and surrounding the refuges. However, specific conductance levels vary with location and with seasonality.
For more information about USGS activities in Oregon, visit the USGS Oregon Water Science Center home page.
U.S.
Department of the Interior | U.S. Geological
Survey
Persistent URL: https://pubs.water.usgs.gov/sir20065036
Page Contact Information: Publications
Team
Page Last Modified: Thursday, 01-Dec-2016 19:01:09 EST
