As part of an interagency agreement with the Bureau of Reclamation (USBR), the
U.S. Geological Survey (USGS) reviewed several documents pertaining to the activities of the Klamath Basin Rangeland Trust (KBRT). The documents reviewed included:

Attached are summaries of the major comments with regard to evapotranspiration, ground water, surface water, and water quality for each report.

At the request of USBR, a direct comparison of the KBRT 2002 Pilot Project Monitoring Report and the USBR Evaluation of First Year and Average Annual Yield from Forbearance was made. In brief, our review indicates that KBRT and USBR used similar techniques to evaluate the consumptive use of water on KBRT land, with one important difference. The KBRT analysis assumes that "native" vegetation will not use ground water after forbearance. We believe this assumption is incorrect and therefore KBRT overestimates the actual amount of water saved. USBR does a more credible analysis of the amount of water saved. They calculate the water savings by subtracting an estimate of evapotranspiration (ET) from "native" vegetation under forbearance from the expected ET from vegetation prior to forbearance. The USBR analysis accounts for the specific vegetative and hydrologic characteristics of KBRT land after forbearance. The net benefit of forbearance calculated by USBR (due to a decrease in ET) is about 0.9 feet of water. Several USGS reviewers concluded that the estimate of 0.9 feet is reasonable; however, one USGS reviewer believes that the net benefit of forbearance could be as high as 1.3 feet.

Attempts to quantify any benefits of forbearance through an analysis of streamflow records are inconclusive. A lack of data before and after forbearance, the complexity of the interconnected system of canals and drains, the lack of adequately accounting for climatic variability, and the large uncertainty in streamflow measurements make it impossible to reliably quantify any changes in streamflow that could be attributable to forbearance.

An additional volume of water (2,050 acre-feet) is included in the forbearance value by KBRT, which is related to irrigation return flow from ground water. Based on our review of KBRT’s reports, we have determined that their analysis, and their ground-water model used to support this, is inadequate to determine this volume of water or the timing of this return flow. Moreover, this volume of water would return to the lake regardless of KBRT activities, and the assumption that this water is made available at a more advantageous time is not scientifically supportable given current knowledge`. It should also be noted that water entering Upper Klamath Lake after an irrigation season helps to refill Upper Klamath Lake for the next year’s irrigation season, which can be of significant value particularly during drought years.

No substantive comments.

No substantive comments.

• KBRT and USBR used similar methods to estimate average annual evapotranspiration (ET) for irrigated land by applying the modified Blainey-Criddle equation; as a result, the average annual ET for irrigated pasture land estimated by each is similar.
  • KBRT adjusted their estimate for contributions of precipitation to determine average annual consumptive use.
  • USBR ignored precipitation in their estimate (presumably because the precipitation component is similar in both irrigated and forbearance conditions) but included open-water evaporation and reduced ET rates to determine average annual consumptive use for topographic high and low areas, including marsh.
• Regardless of adjustments made to determine consumptive use for irrigated land, the application of the Blainey-Criddle equation is a reasonable approach to estimate ET when air temperature is the only temporal data set available.

• Estimates of consumptive use under forbearance conditions differ significantly between KBRT and USBR and, as a result, estimated yield by KBRT is approximately twice that of USBR.
  • KBRT assumed that forbearance land use would consume some water and that this use should be subtracted from the estimated irrigated pasture consumptive use to calculate yield. However, their estimate of forbearance use was based only on the amount of water from latent soil moisture retained from winter precipitation and effective precipitation during the growing season and, therefore, does not include ground water as a source to support the vegetation growing under forbearance conditions. Water level data, however, suggest that ET from the water table occurs throughout the growing season.
  • In contrast, USBR assumes that ground water supports much of the vegetation likely to grow under forbearance conditions in areas where the water table is near or above the root zone. The occurrence of the water table above, at, or near the root zone is supported by the presence of marsh areas and strongly supports the assumption of a ground-water source for forbearance vegetation. USBR used imagery and water levels to delineate areas of similar water-table depth relative to the root zone. USBR estimated average annual consumptive use for forbearance conditions using ET rates obtained from published reports for each delineated area with similar water-table depths.
  • The approach applied by USBR is a common and accepted method for extrapolating ET over broad areas and is a more reasonable approach in areas of shallow water tables.
  • While in principle the USBR approach is more reasonable, it appears that the use of ground water may be overestimated. Data presented regarding shallow water tables seem to be inconsistent. While many areas were noted as having green growth, which is consistent with a shallow water table, no standing water was encountered in shallow auger holes.
    • One USGS reviewer suggests that the water savings from forbearance (as a result of ET) may be as high as 1.3 feet. This is based partly on the assumption that a deep water table observed in some wells during mid-season is representative of conditions throughout a large area. The degree to which these piezometers are representative of the general area is, however, unclear. The water table in these wells appears to be 1-2 feet below lake level, and starts to rise in late September or early October prior to the rise in lake level. These observations suggest that the water table is being influenced by ET and that the depth to water in these wells may not be representative of the entire pilot project area.
    • It is possible that the capillary fringe extends upward from the water table (which accounts for the observation of moist soils throughout most of the auger hole depth).
    • It is also possible that finer, thinner plant roots extend below the depth observed in the auger holes and utilize water from the capillary fringe or directly from the water table, which could explain the lowered water table observed in this area.

The approach used by the USBR, which includes ground water as a source for supporting forbearance vegetation, is the more reasonable approach in ground-water discharge areas or in areas of a relatively shallow water table. However, a more rigorous analysis of the forbearance vegetation and better estimates of ET rates are needed to refine actual yield.

Quantifying the response of a shallow ground-water system to changes in irrigation practices is an exceedingly difficult task. Many quantities involved cannot be measured directly, but must be inferred from other related but measurable quantities. Because of measurement error and simplifying assumptions, analyses of complex natural systems will always have inherent uncertainty. The authors of this report do a commendable job of developing a conceptual model of the ground-water system (the hypothesized behavior) and of data collection. In the final analysis, however, conclusions from the 2002 monitoring are limited by the short data collection period and lack of preexisting data. There is no doubt that the quantitative understanding of the shallow ground-water system in the Wood River Valley will improve as more data are collected and analyzed.

• Any conclusions regarding the hydrologic behavior of the ground-water system in the pilot project area are severely limited by the short period of data collection and the lack of baseline data collected under irrigated conditions.
• Although the proposed hypothetical behavior of the ground-water system in the pilot project area is reasonable, the volumes of water moving through the system (such as recharge from irrigation and discharge to streams) and timing and magnitudes of storage changes are unknown.
• Observations regarding subirrigation in the pilot project area in the ground-water discussion are ambiguous. Ground-water levels below lake level indicate that ground water is leaving the system through means other than discharge to streams, such as through pumping of drains or wells, or evapotranspiration.
• The observed rise in the water table in late September and early October, prior to the rise in lake level, also suggests a mechanism other than discharge to streams is influencing the water table in the pilot project area.
• The ground-water model analysis presented in the 2002 monitoring report should not be used for quantitative purposes. The model is not constrained by data (it is uncalibrated), has inaccurate boundary conditions, and incorrectly simulates known hydrologic processes. Moreover, the model results are not consistent with field observations. It would be inappropriate to consider results of the model anything but strictly hypothetical.
• The report notes that under irrigated conditions a certain amount of the applied irrigation water goes into ground-water storage and returns to the streams after the irrigation season ends, and that under forbearance this water from irrigation return flow is available in streams during the irrigation season. It should be noted that there is no net increase in water to the streams, only a shift in timing. The magnitude and timing of this affect are unknown. Model results showing that 10% (2,050 acre-feet) of diverted irrigation water returns to the stream system after the irrigation season are without basis and should not be considered meaningful.

The data and analyses presented in this report reflect an impressive effort and tremendous amount of work on the part of the authors in trying to assess hydrologic changes due to irrigation forbearance in the Wood River Valley. Overall, the surface-water data collection methods and analyses presented in the report appear technically sound and appropriate.

• However, it is our opinion that data from the report’s surface-water component do not provide any additional support to the report’s main conclusion that forbearance yielded 7,677 acre-feet of water, for the following reasons:
  • To properly assess the effects of irrigation forbearance using streamflow data it is necessary to have at least 2 to 3 years of continuous daily flow data collected during the irrigation period before and after the start of forbearance. These data do not currently exist. Without these data it is not possible to assess if changes in streamflow are a result of forbearance or climatic variability.
  • It is critical that the flow record years used in the pre- and post-forbearance comparison represent mean climatic conditions for the region. Water year 2002 was extremely dry in the Klamath Basin. This issue is not adequately addressed in the report.
  • With numerous drains and canals, the surface-water drainage network in the Wood River Valley is extremely complex. Either data from additional gaging stations or sufficient synoptic (simultaneous) flow measurements need to be collected.
  • Although the report appendix states that USGS quality-assurance standards were used in flow-data collection, this is not supported in the report or in discussions with the contractors. If proper techniques were not used in flow-data collection, measurement errors could easily be as high as 25 percent. Thus, it is possible that flow measurement errors are greater than any actual changes in flow during the pre- and post-forbearance periods.

The data that are currently being collected, along with the additional data collection that was suggested in this review, should provide the necessary data to evaluate ongoing and future restoration efforts in the study area. However, improvements in water quality may take awhile to materialize, particularly for dissolved nutrients that may enter into the ground-water system. It should be kept in mind that data collected for just 1 year can only be used to formulate hypotheses that can be tested and refined over time.

In a few cases, statements were made that were not adequately supported with data. For example, the report states that a 75% reduction in the phosphorus load to Agency Lake is possible from forbearance and wetland restoration. Although benefits in water quality would be expected from such activities, such claims should be based on data, not assumptions or best professional judgment. At this point, with the available data, it is not possible to make such a statement. In addition, the report states that the most likely source of particulate phosphorus is erosion from cattle grazing. While this may be one source, other potential sources—such as suspended algal material—were not mentioned. At this point, both explanations are hypotheses to be considered. The lack of routine monitoring for algae and macrophytes in the current design represents a significant obstacle to understanding the nutrient dynamics of the system, a limitation that may lead to erroneous conclusions.

Estimation of ET rates through sampling of environmental variables is crucial to determining the consumptive use of crops. The monitoring plan presents a sound and objective approach to estimate evapotranspiration from two selected pasture sites within the Pilot Project Area. Hydrologic Engineering, Inc (HEI) is planning to estimate ET rates within the study area using the Bowen ratio and the eddy correlation energy-budget methods. Both of these approaches are theoretically sound and widely used. However, Bowen ratio systems using two solid-state hygrometers that exchange position every few minutes have proven to be more reliable in the field than Bowen ratio systems that use a chilled-mirror hygrometer for measuring dewpoint temperature.

Because the intent of the evapotranspiration monitoring is to facilitate computation of the evapotranspiration differences between irrigated and nonirrigated pastures for the entire Pilot Project Area, the two sites selected for intensive study should be broadly representative of pastures in the Pilot Project. The plan should state more explicitly than it does how this representativeness will be achieved. The monitoring plan also should contain contingencies should the sites become unsuitable for the Bowen ratio technique at some point during the growing season. Finally, the Bowen ratio technique commonly fails to yield physically plausible evaporative fluxes during periods of each day, and the plan should specify how these fluxes will be estimated for such periods so that complete daily evapotranspiration records can be assembled.

The addition of observation wells in the shallow and intermediate aquifers for water-level monitoring, lithologic characterization, and determination of hydraulic properties is appropriate. These data, in addition to the network of existing observation wells, should help to provide the data needed to assess the effect of forbearance of water rights on ground water and surface water. The acquisition of baseline data for additional areas that may be added to the KBRT forbearance program in the future cannot be overemphasized. Additional comments include:

• The use of a cone penetrometer may not yield results that provide the greatest benefit in terms of information and costs relative to other direct or indirect geotechnical methods.
• The plan to create and monitor irrigation test plots is a potentially worthwhile endeavor that could yield valuable data and insights to the water and nutrient budgets of irrigated and nonirrigated areas. However, this plan may be overly ambitious, could require substantial resources, and must incorporate a careful design with emphasis on measurement of budget components. It may be difficult to locate an area with suitable hydrologic boundaries where all of the inflows and outflows can be accounted for.
• Careful consideration should be made as to the benefit of using a ground-water model with limited data to answer questions relative to the Pilot Project operation. However, if further development of the ground-water model is attempted, then it is essential to acquire baseline data, constrain recharge estimates from irrigation, and calibrate the model to observed conditions.

The goals and objectives for surface-water quantity monitoring seem appropriate. These include: "(1) to determine if reduced irrigation demand results in higher instream flows, (2) to quantify the amount of return (tailwater) flows reaching the main channel network, and (3) to develop streamflow accounting units." However, the document contains no specific information on how these objectives will be achieved. It basically is saying that a lot of surface-water flow data will be collected at various locations throughout the Wood River Valley. However, there is no way to determine if collecting data at these designated locations will be sufficient to meet the objectives. It is possible that the proposed data program collection could be much more than is needed because the data from most sites will be of minimal value for determining changes due to forbearance owing to the lack of historical data.

The data that are currently being collected, along with the additional data collection that was suggested in this review, should provide the necessary data to evaluate ongoing and future restoration efforts in the study area. However, improvements in water quality may take awhile to materialize, particularly for dissolved nutrients that may enter into the ground-water system. It should be kept in mind that data collected for just 1 year should only be used to formulate hypotheses that can be tested and refined over time.

The lack of routine monitoring for attached algae and macrophytes in the current design represents a significant obstacle to understanding the nutrient dynamics of the system; a limitation that may lead to erroneous conclusions.