The study area covers more than 8,000 square miles in Oregon and California. The area is in the rain shadow of the Oregon Cascades and the Siskiyou Mountains. Average annual precipitation at Klamath Falls is less than 14 inches per year. Most of the precipitation falls as snow in the winter. There are 2 distinct drainage systems in the study area, the Klamath and Lost Rivers. A number of significant changes have been made to the natural drainage system over time. The level of Upper Klamath Lake was raised in 1921 by the construction of a dam at the outlet. The Lost River originates at Clear Lake Reservoir then flows north, then west, then southeast to Tulelake. Canals connect the Klamath and Lost Rivers and water can be moved in either direction depending upon need. Large areas near Upper Klamath Lake, Lower Klamath Lake, and Tulelake have been diked and drained to provide land for agricultural uses. Movement of surface water through these areas is highly complex and is controlled by several agencies.

Water is used for several purposes within the basin. Major withdrawals are used for irrigation of agricultural lands. Preliminary estimates indicate that as much as 1,000,000 acre-feet of water may be used for irrigation of agricultural crops per year. Other uses in the area are for public supply, industrial, commercial, stock, and domestic. Nonconsumptive uses of water would include generation of electricity and recreation. Estimates of water use in the basin can be subdivided into the two major sources, surface and groundwater. Due to the interaction of these two sources of water, estimates of movement and use of both sources will be made.

PUBLIC SUPPLIED WATER

Water supplied for public drinking and other associated uses will be estimated primarily from data provided by the Oregon Water Resources Department (OWRD) water-use reporting system. OWRD requires the reporting of monthly totals of water withdrawn by individual points of diversion (POD’s) for each water year (October 1 to the next September 30). This information is reported by all publicly owned systems, whether the water is for public supply, irrigation or improvement districts, and other similar agencies. This information has been collected since 1989 and provides a valuable resource to determine current and historic rates of withdrawal. Data requested from OWRD for the Water-Use Reporting System provided partial records for Bly, Chiloquin, Klamath Falls (includes geothermal uses), Malin, and Merril. No public supply data has been found for California communities. Other information on quasi-municipal and group domestic systems may be compiled from state water rights and state health division records. Preliminary estimates indicate there may be more than 100 wells that supply water to municipalities, private systems and other systems larger than single family domestic systems. Municipalities generally collect detailed information on withdrawals, water levels, and other significant items. Most, if not all, municipal and larger quasi-municipal/group domestic wells will be field located and water use information verified. This information may be useful for estimation of withdrawals for other water uses.

COMMERCIAL AND INDUSTRIAL WATER

Commercial and industrial withdrawal of water will be estimated from several sources. Information on water supplied for public drinking, such as used in restaurants and motels/hotels, is available from a drinking water database maintained by the Drinking Water Section of the Oregon Health Department. Many of the commercial establishments are located within the areas served by municipalities in the area. Those businesses outside the municipal service areas are required to have water rights to withdraw water in any significant amounts. Information on water rights is available from the OWRD and California Water Resources Control Board (CWRCB). For larger industrial users, water usage will be estimated from wastewater discharge permits filed with state and federal agencies. Selected sites should be visited to collect water-use data. There may be slightly more than 100 wells in the study area that are used for commercial/industrial purposes. Data for about 30 Oregon water rights for commercial/industrial uses were selected from the statewide water-rights database. Similar information for California is not currently available. There may be a reasonably significant reduction in industrial water use due to the closure of several large sawmills in the area.

STOCK AND DOMESTIC WATER

Stock and domestic withdrawals are believed to be small for the basin. In most areas, little consumptive use of ground water occurs. No estimates of water use are planned at this time.

IRRIGATION WATER

Irrigation withdrawals account for a large proportion of the total withdrawals but are the most complex to estimate. Surface and ground water may be used in close proximity to each other, or may be applied to the same lands if used in a primary and supplemental application mode. No current information is available on ground-water withdrawals within the study area. There may be as many as 1000 irrigation wells within the study area. Oregon water rights indicate slightly more than 600 POD records within the study area. There are no data to indicate how many of these water rights are currently active.

Estimating irrigation water use over a large area is difficult. To do this in a reasonably accurate manner Current information on land cover within the agricultural areas is needed. Image-processing techniques using Landsat imagery can be used to determine land cover and use within the basin (see section image processing). Using local information on crops grown and timing of planting and harvest of these crops, satellite imagery will be acquired to optimize the identification of important agricultural crops and natural cover (see section on ground truthing). Field crews will collect "ground truth" information on the temporal and spatial distribution of cropland and natural cover. This information will be recorded in the field on USGS orthophoto quads and later digitized into GIS coverages. Pertinent supporting information on conditions at the time will be recorded and transferred to a GIS compatible database.Image processing techniques provides new land cover data of the area, but will not realiably provide information on the source of water for irrigation. Ground truth information is used to verify the crops or crop groups grown and differentiate between natural cover (wetlands) and areas of high water application such as flood irrigated pastures. One possible alternative to the complex and extensive task of creating a land cover map from raw satellite imagery would be to research the availability of completed image processing work done by other agencies. Some likely sources would be universities, other land use agencies such as Dept. of Agriculture, Dept. of Interior, and several federal cooperative groups such as MRLC (Multi-Resolution Land Characteristics project involving US EPA, USFS, USGS-BRD, NOAA, USGS-NAWQA, EROS Data Center), and several others. This data would only be useful if the imagery is recent and has sufficient detail (cover classes) in the needed areas of agricultural activity. Some of these cooperative studies have a specific focus that exclude their utility to the project such as a focus on wildlife habitat. Further checking is needed on this subject.

A second, more viable alternative would be to use Digital Orthophoto Quadrangles (DOQ’s. The DOQ’s are available from USGS-NMD and local sources. The USFS has full quad coverages of all National Forests in Oregon and Washington. The BLM has coverage of areas administered by their agency. Within the project area, most of the lowland area south and east of Klamath Falls is not covered by the USFS and BLM. The Oregon State GIS Service Center is reported to be the repository for copies of the coverages from USFS and BLM. There is some uncertainty about the long term viability of the group and other sources may need to addressed. In any case, the lowland areas would need to be purchased from USGS-NMD in ¼ quad segments and spliced together. This black and white (1-meter resolution) coverage is of very good quality and may provide an adequate background to overlay ground truth and other pertinent coverage’s to create a form of land cover map. Due to the large areas of wetland within the study area, the National Wetland Inventory coverages may be quite useful to eliminate some wetland areas from consideration. The DOQ’s can be used to print paper copies of the quads, used interactively in the field with appropriate software to outline land covers of interest, or possibly linked to a GPS unit to create real-time georeferenced land cover digital maps.

Water rights information will be used to identify Point of Diversion (POD’s) and Place of Use (POU’s) for irrigation rights where available. The OWRD Water Rights Information System (WRIS) recently created GIS coverages of POD’s and POU’s for the Oregon side of the basin. These coverages will be very useful to help identify the source of water applied to selected plots of land. Information layers identifying the known POD’s and POU’s will be used to estimate the amount of irrigation water applied to specific crops by overlaying the land cover maps. These POD and POU files will need quality control checks for spatial and temporal accuracy. Water rights information will also be utilized for the California side.

Data from local agricultural agencies will be used to determine the optimum amount of water needed to grow a specific crop during a specific growing season. Crop water requirements are often estimated from climatic measurements of evapotranspiration (ET) for a reference crop such as alfalfa. Coefficients are used to adjust these water requirements for other crops grown in the area. Estimates of ET are currently available for the Tulelake area (Tulelake Field Station), Alturas (these first 2 are from the CA CIMIS network), Lakeview, Madras, and Powel Butte (the last 3 are US Bureau of Reclamation Agrimet stations. An Agrimet station is scheduled to be installed in 1999 at the OSU Experimental Station south of Klamath Falls. Data from this site should be very useful to correlate with nearby climatic stations.

To refine the crop water estimates, surface- and ground-water sites will be selected throughout the basin (index sites) to monitor the actual water used to grow crops over a growing season (see section on index sites). These index sites will be representative of the types and amount of crops grown in the basin. Field crews will collect data such as; pump type and horsepower, pumping rate, power consumed, type of irrigation equipment used, and types of crops grown at the site. Noninvasive acoustic flowmeters will be used to check pumping rates at sites periodically during the season. Information from these index sites will be used to extrapolate the data to the remaining irrigation water users in the basin (see Sandberg method).

The selection of irrigation wells for water-use purposes will be integrated with other well selection criteria such as areal distribution, geology, hydrology, etc. Those wells that produce large quantities of water and supply this water to a typical use would be of the greatest interest. Wells selected as index sites would be a subset of this category. The distribution of irrigation water-use wells will be indicative of the overall distribution of irrigation wells within the study area. Water rights information can be used to determine areal densities of wells by water use. Suggested areas for the selection of irrigation wells would be: Upper Sprague River area near Bly, Middle Sprague River area between Beaty and Sprague River, the area north and west of Klamath Marsh to Beaver Marsh, Fort Klamath to Agency Lake area, the various valleys of the Lost River Basin (Langell, Poe, Yonna, Swan Lake), the area near Merrill and Malin, the Lower Klamath Lake area, and the parts of California in Butte Valley and surrounding Tulelake.

Irrigation—Location and data collection

Standard techniques will be used when field locating water-use irrigation wells. Accurate location of the well using quad maps, GPS, and detailed sketches of the immediate surroundings of the well will be made. Data collection of information such as elevation, water levels, etc will be performed in much the same manner as other sites. However, other data is needed to assist in the estimation of water used (pumped) at the site. Data on the pump type (turbine, submersible, booster pumps, etc), pump horsepower, electric meter data, irrigation equipment in use (hand lines, wheel lines, center pivots, etc) and if the system is in use, line pressure at the pump, and a pumping water level would be useful. Much of the above data can be recorded on the "Collins field form". Information on the crops being grown in the surrounding fields can be very useful.

Consideration will be given to the measurement of pumping rate (gpm, etc) at as many sites as possible. In-line flowmeters may be installed at some wells within the basin. Detailed information should be collected at these sites if possible. The pump test program administered by OWRD may contain information that can be associated with some of the selected sites. Estimation methods range from counting the number of sprinkler heads in operation, noting the nozzle size, whether these are single or double head sprinklers, and the working line pressure of the system to recorded pumping rates/volumes from flowmeters. Often the local irrigator can supply estimates of the amount of water being applied. It is useful to know how long the sprinklers are left in a location before being moved and how often the crop is irrigated to help with estimates of seasonal water use. Timing of when irrigation begins in the growing season (climate and crop dependent) and how long the crop is irrigated, when it is harvested are useful data. Some systems pump open discharge to a ditch or canal. An estimate of flow can be made from the "yardstick" method by measuring horizontally and vertically from the end of a horizontal, full pipe to a 12-inch drop in the water surface of the discharge. Other more advanced methods use some form of acoustic velocity meters such as owned by OWRD or the Marsh-McBirney meter owned by the USGS. The second meter can be used in submerged discharge pipes of medium to large diameter. Power consumed while the pumping flow test is being conducted is needed. This will allow later computation of power consumed vs. water pumped factors for the area.

Other information needed to estimate irrigation water use for the area would include county and/or basin wide crop acreage statistics. For Oregon, the OSU Ag Extension Service collects annual data on crop acreages at the county level for most irrigated crops. Little to no information is available for irrigated pasture and native hay production. Crop acreage data is useful in determining water application rates for groups of irrigated crops. Information on crop growth cycles is needed to include, average planting date, stages of growth, average dates of harvest, and water needs during the growth cycle. See Cuenca for data on monthly consumptive water needs for various crops throughout the state. Double cropping may be practiced in the area, but due to the average elevation and the presumed short growing season, this may not be an important factor.

Seasonal (irrigation) totals of power consumed for a number of sites are needed around the basin, index sites will provide some of this data. Contact with the local power company(s) to check feasibility of obtaining seasonal total power records for a small number of sites may be persued. Historically, individual owner permission was required before data was released, however, OWRD may be able to request this data from a regulatory agency position. The Sandberg method of estimating power for sites lacking full season power records can be used to supplement power records from full season sites.

In the middle to lower reaches of the Lost River Basin, return flow from irrigation is reported. Although this area is irrigated primarily by surface water, the overall water budget of the area is affected. The Lost River is reported to be completely diverted at Lost River Diversion Dam to the Klamath River. Large tracts of land downstream of this point are irrigated from canals originating from Lake Ewauna. Water accumulates in the Tulelake Sump and is pumped from the area through a tunnel through High Rim toward the Klamath Straits drain. Water deliveries from the upstream reservoirs are reported to be less than estimates of water used in the area served. Return flows may be an important component of the water budget in this area. Surface water quality is presumed to be related to reuse of irrigation water within the system.

Another factor that affects the basin water budget is the drainage of low lying lands adjacent to Agency and Upper Klamath Lake. An extensive network of drainage ditches and canals collect water from the surrounding lands and large pumps lift the water to the adjacent stream or lake. Data were collected by Snyder and Morace for the area. Changes in land use in the area suggests that updated information should be collected for this area. Other locations should be checked for similar conditions.

Ground-truth

This work consists of a field crew (2 persons) traveling about the basin and mapping typical ground cover (agricultural crops, natural cover, urban areas, suburban area, forested areas, etc) onto USGS orthophoto quads. Two-person teams is the most efficient and safe way to collect this information. The areal distribution and extent of mapped areas will be related to desired accuracy of detection of a given cover from the expected results of the image processing work. The timing of collection of ground truth is important. For those land covers that show little change over a growing season (coniferous forests, urban and suburban areas and some water bodies), mapping can be done on a time available basis. For most agricultural crops, especially short season crops and areas with multiple crops grown per year, mapping should be done as closely to the time of the satellite image acquisition as possible. To identify most agricultural crops, at least two satellite images will be purchased ( middle and late growing season) to help differentiate between crops of similar appearance. Knowledge of local farming techniques is essential to planning this fieldwork. Consultation with the local county agricultural extension agents is suggested. Some of this work can be integrated with setting up and collecting data for the water-use index sites (see water-use index sites). There are large areas of natural vegetation (marshes & riparian areas) that may be difficult to distinguish from certain agricultural crops. Mapping effort will be distributed accordingly. Information from other agencies can by utilized to offset some of the USGS mapping work. Suggested contacts are:U.S. Forest Service, Bureau of Land Management, U.S. Fish & Wildlife, Bureau of Reclamation, Natural Resource Conservation Service (SCS), etc. The mapped ground truth will be digitized from the orthophoto maps and notes from fieldwork (condition of crops, type of irrigation being used, etc) entered into a compatible database such as Info or a spreadsheet. There could be as many as 100 quads if all were used, however, since most mapping will be in the agricultural areas, the actual number might be closer to 50 quads.

Index sites

This task involves identification of a number of pumping sites distributed about the basin. These sites should be representative of the uses of water in the basin and could include, ditch pumps used from canals and ditches (both for irrigation and drainage), pumps used to pump ground water, and other types of water application being used in the basin. The sites should be picked to show how water is used to irrigate typical crops and using standard farming practices, cooperative landowners, and accessibility. Some of these sites will probably be used for observation wells also. Typical data collected at the site would include; pump type and horsepower, observations about the irrigation system (hand lines, wheel lines, center pivots, etc) and whether in use at the time, information from the power meter, what crops are growing in adjacent fields and their condition (maturity, vigor, degree of canopy closure, harvesting information), flowmeter readings if available, and sources of water being used (or drained). The site should be visited at least once during the pumping season to measure the actual discharge from the pump while pumping. This can be achieved by using the noninvasive acoustic velocity meter currently owned by OWRD. By measuring discharge and power consumption simultaneously, data can be developed to define a relationship between the two. These relationships will be dependent upon such things as total lift, pump horsepower, condition of the pump and motor, depth of well, and aquifer characteristics. Due to the need for water use estimates for the water budget, a selected number of surface water systems will be inventoried. Some data may be available from the Bureau of Reclamation in the area that they serve. Another source would be the leased lands on the wildlife refuges in the area (see U.S. Fish & Wildlife/Bureau of Reclamation).

Areas to consider for index sites would be the Sprague River valley, Sycan Marsh area, middle Williamson River (above Klamath Marsh), Klamath Marsh, Wood River to Agency Lake area, Williamson River delta, Swan Lake, Yonna valley, Lost River from Clear Lake Reservoir to sump area near Tule Lake, Lower Klamath Lake area, Butte Valley, and other identified larger water-using areas. An initial estimate of index sites would about 30 sites over the 2-state area. This task could be performed by the person assigned to make water level measurements, but should be done monthly during the irrigation season when ground truth and Landsat imagery is acquired. A minimum of one irrigation (or use) season is needed to develop data for power vs. pumpage. An additional 10-15 minutes per site per visit would be needed to collect the necessary data. Additional time would be required when flowmeter measurements were made. An initial estimate of time for a flowmeter measurement would be about half a day, allowing for travel time, setup, data collection, and equipment removal. Total time required for this would depend upon how many sites were selected, the frequency of visitation, and the length of data collection period. About 2-4 weeks should be allowed for data entry and analysis. Most of the analysis would be after the irrigation season of the year the Landsat data was acquired. Additional power consumption data can be easily acquired from the initial contact (field location) of the site and again when the site is visited during the synoptic measurement effort.