 |
|
 |
 |
 |
|
 |
 |
 |
 |
 |
 |
 |
 |
Response to Public Comment
Listed below are summaries of comments by Tom Myers, PhD, for the Great Basin Water Network followed by a response from USGS. The summary of each comment is linked to the original document submitted by Mr. Myers.
Comments from Tom Myers, PhD
Complete letter from Tom Myers (pdf)
TM1. Specific Textual and Mathematical Errors
The BCM report mentions in many places that the grid scale is 82.3 feet or 270 meters. The two numbers do not convert to each other.
RESPONSE
A value of 82.3 ft was not found in the BCM report (SIR 2007-5099) but the grid scale conversion from metric to imperial units is incorrect in the report. Two hundred seventy meters is equal to 886 ft rather than the 866 ft as reported in SIR 2007-5099.
TM2. Inconsistency between irrigation area for water use and for ET discharge
A. If correct then explain why the discrepancies do not represent a major error in the GW discharge estimates
RESPONSE
After further review, the authors see no inconsistency among the irrigated cropland acreage reported in the draft summary report (OFR 2007-1156) and supporting publications (DS 273 and SIR 2007-5087). Your reconstructed values of irrigated acreage (table 5) based on the water use and acreage worksheets given in appendix A are, in fact, correct. The noted differences are valid, not errors. The acreage reported in the water use worksheet is defined as the acreage of irrigated cropland for 2005 regardless of whether these lands are located inside or outside of ground-water discharge areas. For example, 9,200 acres are reported for Snake Valley and 4,360 acres for Lake Valley. The acreage reported in acreage worksheet is defined as the composite acreage within ground-water discharge areas for the “recently irrigated cropland-historically mixed phreatophyte” ET unit during the year 2000 and/or 2002, and/or 2005. For example, 9,932 acres are reported for Snake Valley and 0 acres for Lake Valley indicating that in Snake Valley most irrigated acreage is within the ground-water discharge area, and for Lake Valley all irrigated acreage falls outside the ground-water discharge area. Since 9,932 acres was delineated in Snake Valley for the “recently irrigated cropland-historically mixed phreatophyte” ET unit, and only 9,200 acres were identified as actively irrigated cropland in 2005, then there are some fields within the ground-water discharge area that were not actively irrigated in 2005. A table and text was added to the summary report (SIR 2007-5261), and column headings revised in appendix A to help clarify this point.
B. BARCASS states that consumptive use estimate for irrigation in the study area is 2.9 ft/y, but that number is not what was used to calculate the total irrigation consumptive in Appendix A. Values ranged from 2.78 ft/y for Butte Valley to 3.08 ft/y for Little Smoky Valley (Table 4). Snake Valley is 2.99 ft/y while Spring Valley is 2.81 ft/y. The BARCASS report should not state that one value was used when there was actually a range.
RESPONSE
The 2.9 ft/yr is intended to represent the average of the study area but, to avoid any confusion, the summary report was modified to replace the sentence with “Estimates of average crop consumptive use (ETc) for each HA ranging from 2.78 ft/yr to 3.08 ft/yr are in agreement with measured consumptive-use rates for alfalfa and pastureland given in Maurer and others (2006) for a similar climate.”
Plate 4 shows recharge/discharge estimates by basin and by subbasin. However, the values for subbasins sometimes do not add to the total for the basins. For example for Steptoe Valley, the recharge for subbasins is 67,700, 63,300, and 27,000 af/y which sums to 158,000 af/y. The map and Table 6 each reports 154,100 af/y. The ET for the subbasins correctly sums to the reported total. This suggests the error is in the reporting for the subbasins.
RESPONSE
The only inconsistency (not related to rounding) is the reported subbasin recharge in Steptoe Valley. All recharge values were verified, rounded to 1,000’s of acre-feet and corrected for Steptoe Valley. Values that are less than 1,000 acre-ft are now shown as “<1”.
TM4. Need discussion of locations of ground-water divides with respect to areas of potential flow
BARCASS presents a substantial change in previous thinking with regard to interbasin flow. Based on geology interpretations, BARCASS identifies areas that could allow flow through basin boundaries. Some are uncertain, and BARCASS identifies this. It should also be noted that the geology may consist of formations which are pervious enough to allow flow, but a groundwater divide could separate the basins. The BARCASS report should add a discussion about the location of groundwater divides with respect to areas of potential flow. This is important because it would aid an interpretation of how interbasin flow could change due to stresses. In other words, it is possible that pumping on one side of a range could draw flow from the other side by lowering the groundwater divide. This is particularly important in the south end of the Snake Range and in the south end of the Schell Creek Range where BARCASS indicates flow is possible but uncertain and may be constrained (elsewhere BARCASS indicates that water balance indicates flow through some of these boundaries).
RESPONSE
The summary report text has been revised to read “the source of ground water in the carbonate-rock aquifer within the study area is a relatively large recharge mound centered over Steptoe, Long, Jakes, and Butte Valleys. Within the large regionally extensive mound, small, high mounds are found over the Schell Creek and Egan Ranges. The recharge mounds form ground-water divides that separate the study area into multiple flow systems.” The general ground-water flow directions vary among the flow systems. A proposed new study should aid in the quantification of flow between southern Spring and southern Snake Valleys.
TM5. Need cross-sections along the crests of the major ridges
The geology presentation would be substantially improved if in addition to the east/west cross-sections the report provided profiles along the crests of the major ridges. This would aid in the interpretation of the potential for interbasin flow. A good example is the Fortification Range. The south half of the range has thick tuff through which interbasin flow projected to occur there would flow. The northern part of the mountain range is carbonate, however, except for potential thin intervals of Chainman Shale (BARCASS, page 40, description of unit 9). A cross-section would help the reader better interpret how the USGS feels this could impede the flow.
RESPONSE
This is beyond the time and budget constraints of this project. Cross sections along crests of major ridges were not identified by the BARCAS study team as priority locations for cross-section construction. Seven cross sections are shown in figure 9, covering much of the study area and extending across multiple HAs. These cross sections portray the three-dimensional shape of pre-Cenozoic basement, locations of major basin-bounding structures, and the presence of significant intrabasin faults.
TM6. BCM report needs to explain whether the soil always retains moisture at the minimum wilting point.
The water balance assumes that all precipitation, snowmelt and carry-over soil moisture is available water at the beginning of the time step. Potential evapotranspiration (PET) is removed from the available water, as is precipitation in the form of snow that does not melt. This water fills the soil water first: “[p]otential runoff was calculated as the available water minus the total storage capacity of the soil” (Flint et al 2004, page 165). Total storage capacity is the soil depth times the porosity. The potential runoff is subtracted from the available water to determine the amount of water available for recharge. Potential recharge is the remaining available water minus the field capacity of the soil. The maximum recharge rate is the “permeability of the bedrock” (Id.). If the available potential recharge exceeds this maximum recharge rate, the excess water remains in the soil until the next time step. One detail with the model not explained is whether the soil will always retain moisture at the minimum of the wilting point.
RESPONSE
The volume of available water per unit depth of soil in the BCM is the difference between field capacity and minimum wilting point. Moisture below the minimum wilting point is unavailable and is retained in the soil.
TM7. Figure 4 in BCM needs to be clearer so that amounts for a given area can be better read.
The BCM report has a map which shows the precipitation estimated using PRISM for the BARCASS study area (Figure 4). The scale is very hard to read; based on the scale and the amount of blue shown on the map, there are rather large areas in the mountains with more than three feet of precipitation (the top of the scale is 3.5 ft/y, or 42 in/y). Even if the ridges receive this much (they do not), the large area with this amount illustrates how PRISM may overestimate the precipitation.
RESPONSE
Figure 4 imparts the distribution of precipitation by scaling values from coolest colors (highest) to warmest colors (lowest). The size and intensity of cool and warm colors indicate the spatial distribution but is not intended to be used for quantifying precipitation; rather, the reader is referred to appendix A which lists precipitation by basins and subbasins. The summary report discusses the statistically derived differences between PRISM and measured precipitation.
See section titled “Water Balance”
RESPONSE
BCM generated runoff volumes are not directly comparable to gaged runoff because much of the runoff is not gaged in the study area and the BCM model results are long term (112-year) averages with a much longer period of record than any of the gaging stations. The authors state on page 4 of the BCM report, “No attempt is made in this study to route runoff, but only to estimate the amount and source of runoff that may be available for recharge downslope”. Text has been added to the summary report that describes how BCM input parameters have been compared to measured data in similar areas in the west.
See section titled “Water Balance”
RESPONSE
The BCM was not re-run for the following reasons. Each run using month-long time steps required several weeks of computation time. More importantly, the objective of the BCM simulations was to estimate long-term average recharge throughout the study area. The authors’ state that the recharge estimates do not necessarily reflect the current short-term average recharge for the study area. A regression approach to relate the shorter period (1970-2004) to the longer period (1895-2006) ignores the affects of antecedent conditions. This would be problematic if this study attempted to predict recharge in a particular year; however, the time step used should minimally affect an estimate of a 112-year average. Interflow and runoff routing that is hypothesized as not contributing to ground-water recharge is not included as part of the potential recharge quantified by the model. While the authors recognize the short comings of the PRISM map, the map is an improvement over the older Hardman precipitation map because geology and soil factors are considered thereby improving the spatial distribution of potential recharge.
See section titled “Water Balance”
RESPONSE
The text has been expanded to include a discussion of limitations and uncertainty. The authors do state that, “the uncertainties in the saturated hydraulic conductivity of bedrock was the greatest source of uncertainty for ground-water recharge estimates from the BCM because saturated hydraulic conductivity of bedrock partitions water between in-place recharge and runoff. Recharge from runoff ranges between 10 and 90 percent, which increases the uncertainty of ground-water recharge where runoff exceeds in-place recharge such as in Lake, Snake, and Spring Valleys”.
See section titled “Water Balance”
RESPONSE
Please see the explanation given by the authors on page 15 of the BCM report, “Although recharge estimates presented in this report are extension of Flint and others, 2004, results may differ due to different climatic data and an improved snow accumulation and snowmelt model used in the current evaluation”. Additionally, new information incorporated into the BCM for the BARCAS study and specific to the smaller study area as opposed to the earlier, larger regional model (Flint and others, 2004) will always influence model results. If future BCM simulations are done that incorporate new hydrologic and geologic data being collected by SNWA, model results will not be identical to the current results. This is true of all models due to the non-uniqueness of results.
See section titled “Uncertainty Estimates for ET Discharge”
RESPONSE
Ground-water discharge estimated from the ET unit referred to as “recently irrigated cropland—historically mixed phreatophytes” of 19,000 ac-ft/yr accounts for about 4 percent of discharge from the study area. Trying to group each of the 643 irrigated fields into one of the other 9 ET units would have been time consuming and would have minimally affected the ground-water discharge estimates.
TM13. Summary report should better explain how the ET unit rates were set.
A. ET rates were determined for 10 ET units ranging from playa soil to marshland. The ET Rate table in Appendix A shows different rates for marshland, meadowland, grassland, dense desert shrubland, moderately dense desert shrubland, and sparse desert shrubland. However, for moist bare soil, open water, dry playa and irrigated lands, the same rate is used for all basins and subareas. It is understandable that conditions such as aspect, elevation, and average temperature (local micrometeorological factors, BARCASS page 58) would cause the ET rate to vary for a specific ET unit for different subareas. However, these factors affect the evaporation from all ET units, not just the six mentioned (irrigated is a special case discussed below). The USGS should better explain how the different rates were determined and why site conditions would cause variation in some of the ET units but not in the others.
RESPONSE
ET estimated from the ET units referred to as “bare soil, open water, dry playa, and recently irrigated cropland—historically mixed phreatophytes” accounts for about 10 percent of ET from the study area. Data are very limited for these ET units. The ET rates developed for these ET units represent our best estimates of the long-term annual ET based on the limited data available. Additionally, the effects of depth to water and soil texture on ET rates in these areas are not well understood.
Ninety percent of the estimated ET is from the shrubland and riparian ET units. A range of ET rates could be assigned for these ET units because more data were available, particularly the ET rates measured directly by the USGS for this study. Given the confidence in these high-quality data (albeit only 1 year at six sites), a range of ET rates was developed in conjunction with remote sensing and fieldwork that varied with vegetation density.
B. BARCASS used recent literature values from four separate reports to determine the average ET rate to apply to different types of phreatophytic vegetation (BARCASS, page 54). The range is shown on Figure 27 of BARCASS. Figure 27 also shows a single line for “area-weighted average-annual evapotranspiration rate” which is confusing because it implies there is a single value per ET unit used for the entire BARCASS area. As discussed in the previous paragraph, it appears that a range was used for some ET units rather than a single value; the USGS should fix this discrepancy.
RESPONSE
ET rates used to estimate ground-water discharge are given in appendix A. Text revisions now reference appendix A.
C. A table with values from each literature source showing the value that could be used for each ET unit from that source would be more useful than the range shown on Figure 27. Our review of the sources suggests that ET units in those sources may have significantly varied from those described in BARCASS or the ET report. In other words, BARCASS may have used inappropriate ET units.
RESPONSE
Figure 27 is considered appropriate for the audience of the summary report. ET units described in the summary and ET reports are not inappropriate. Initially, the range in ET rate for each of the 10 ET units was from the literature. Subsequently, the ranges were modified using ET rate measurements made in the study area, and differences in vegetation density identified from satellite imagery using a MSAVI. Figure 27 shows where measured ET rates fall along the range and provides evidence that the ET rates developed for the selected ET units are in fact very reasonable. Readers are referred to Smith and others (2007) for additional and more detailed information on the selection and delineation process. This expanded information is not considered pertinent to a summary report and can easily be found elsewhere. The relation between MSAVI and measured ET rates is illustrated in Moreo and others (2007), another report often referenced throughout the summary report.
D. One of the referenced reports, Nichols (2000), estimated ET for various basins in 1985 and 1989. These years occurred during the second half of the wettest decade on record in the area. Based on statewide precipitation data downloaded from the National Climatic Data Center for Nevada and Utah and for Salt Lake City, the decade of the 1980s was extremely wet (www.ncdc.noaa.gov). For Nevada, the decade was the wettest; for Utah it was second wettest only to the 1990s (Table 2). In Salt Lake City, the nearest city included in the data base, the 1980s were also the second wettest. It is likely that the phreatophyte cover had expanded and its density had increased. The ET rates determined in that study would likely have reflected healthy vegetation. BARCASS should not rely on these estimates as accurate for long-term pre-development rates.
RESPONSE
Rates reported in Nichols (2000) were not directly used in determining ET rate ranges. Literature reported ET rates were assessed and modified with in-situ measurements and differences in vegetation density as explained in the previous response. Nichols results are included primarily for comparison. Note that the discharge values computed in the summary report generally are lower than those given in Nichols.
A. Accordingly Tom states that “the reality is that not all evaporation from open water sources is from ground water”. Tom further states that “surface water runoff during storm periods and snowmelt reaches the playa and that surface water runoff to open water area would add “area” delineated as open water causing an overestimation of GW ET discharge by including surface water evaporation.
BARCASS assumed that all precipitation is effective for satisfying the ET requirements for a ET unit. This is truly an incorrect assumption because most precipitation in the valleys occurs in short-duration, high-intensity storms at rates that exceed the infiltration capacity of the soil. The gradient is toward the playas. The presence of rivulets in the soil that may be observed by any visitor these areas make it clear that runoff does occur. Not all of the precipitation is effective. BARCASS should adjust its GW ET rates to account for runoff.
RESPONSE
Water entering a discharge area from outside the discharge area (run-on) that either recharges shallow aquifers or elevates soil moisture in areas where phreatophytic vegetation occur effectively decreases the ground-water discharge rate. Run-on discharging to playas or open water bodies has no net effect on the ground-water discharge rate. Precipitation falling directly on discharge areas potentially can infiltrate the basin fill, or elevate soil moisture downgradient from the area of origin but likely does not change the ground-water discharge rate. Conversely, precipitation that potentially can flow into playas or water bodies may increase the ground-water discharge rate. However, the net effect of these processes likely is small and would not significantly alter ground-water discharge estimates because some of the processes appear to cancel each other out and other processes have no effect. Nevertheless, this is a limitation of the method and is listed in the “Limitations and Considerations of Methodologies” section.
B. The USGS includes groundwater discharge from open water area. Presumably, this means playa lakes and open water in wetlands throughout the valleys. The ET rate ranges from 4.6 to 5.6 ft/y and is considered to be discharge from groundwater (BARCASS page 54). Open water accounts for just 0.1 percent of all ET units and only a few hundred acres. However, the reality is that not all evaporation from open water sources is from groundwater. Surface runoff, especially during storm periods and snowmelt, reaches the open water areas in these valleys. BARCASS assumes that surface water runoff that reaches “fine-grained playa sediments is assumed to evaporate and for the purpose of the water budget does not contribute to either ground-water recharge or discharge” (BARCASS, page 64). It seems that surface water runoff to any open water area would add to area for ET discharge estimates and that the methods used in BARCASS overestimated GW ET discharge by including surface water evaporation.
RESPONSE
The authors do not consider ground-water discharge by ET to be over-estimated in the BARCAS study primarily because ET units in the areas most likely to change annually or seasonally (open water bodies, marshland, and playa) were delineated using a land-cover classification that used remotely sensed imagery from multiple dates and years (SWReGAP; Smith and others, 2007). The authors feel strongly that that the area delineated as open water is supported primarily by ground water inflow or by springs which are fed by regional ground-water sources. The extent of these “wetter” ET units during anomalously wet years or during wetter periods of the year is not included as part of the delineated unit. In addition, Figure 32 shows the relatively tiny contribution of “open water” unit to ground-water discharge. Therefore, even if ground-water ET discharge is slightly overestimated for the “open water unit,” the overall affect to basin-wide estimates is minimal.
A. See section titled “Area of Irrigated Lands”
B. There is a significant probability that the actual discharge estimated for Snake Valley is closer to or even less than the recon discharge. See section titled “Snake Valley ET Discharge” and “Conclusion”
The difference is clearly due to the difference in the rates, especially the GW ET rate which is very sensitive to the low precipitation estimate. Hood and Rush’ estimates apparently accounted for precipitation, although it is not specifically stated in the document, as did BARCASS’ estimates. The low precipitation estimates in BARCASS for Snake Valley, less than 7 in/y, may have caused the GW ET rates to be higher than used by Hood and Rush.
The CV discussed above for Snake Valley subbasins is potentially higher than for the overall study area. Therefore, the high discharge estimate for Snake Valley may be quite variable; based on the high CV, there is a significant probability that the actual discharge could be closer to the recon discharge rate.
Due to the uncertainty, the BARCASS estimates for available water may be more uncertain than was perceived for previous estimates due to the lack of calibration. For example, the high discharge estimate for Snake Valley has been shown to have a broad range due to the high CV. Therefore, there is a significant probability that the actual discharge is closer to or even less than the recon discharge rate.
RESPONSE
The authors agree that the higher discharge estimated by the BARCAS study for Snake Valley is a result of using higher ground-water ET rates for shrubland vegetation than used by Hood and Rush (1966). Hood and Rush (1966) used 0.2 ft for all shrubland. The calculated area-weighted average ground-water discharge rate for shrubland in the BARCAS study is 0.39 ft. USGS measured ground-water discharge rates ranged between 0.30 ft and 0.36 ft at the Snake Valley ET site (SNV-1) which is located in an area where the shrub density is at the lower end of the moderately dense desert shrubland ET unit. The authors believe the area-weighted average annual ground-water discharge rate of 0.39 ft used for shrubland in the BARCAS study is closer to the true rate than the 0.2 ft assumed by Hood and Rush (1966). The higher rate of ground-water discharge in Snake Valley is supported by the low precipitation to the valley relative to other valleys in the area with very similar vegetation. Some of the variations in the ground-water discharge rate accounted for in the BARCAS study were not considered in the recon report estimates developed over 50 years ago.
The authors are confused by the argument that CV values calculated for this study lead to a “significant probability that the actual is closer to or even less than the recon discharge rate” (80,000 ac-ft). The computed subbasin CV values for the ground-water discharge estimate of 132,000 ac-ft in Snake Valley simply implies that there is a 67 percent likelihood that the ground-water discharge estimate for Snake Valley is between 100,000 ac-ft and 165,000 ac-ft per year. However, the authors believe there is significant room for the refinement and improvement of the estimated discharge. The estimate could be refined and significantly more confidence could be gained with additional spatial and temporal ET data.
TM16. Summary report should estimate the interbasin flow east through the Confusion Range shown on Plate 4.
Plate 4 shows interbasin flow from the Snake Valley through the Confusion Ranges. This appears to be separate from the interbasin flow from Snake Valley to the Great Salt Lake Desert. The map shows the entire boundary as likely to transmit groundwater. This flow may be the primary inflow to the Fish Springs Flat basin which features the substantial discharge at Fish Springs. BARCASS should estimate the flow east through the Confusion Range.
RESPONSE
Plate 4 does not contain flow arrows. Plate 3 contained flow arrows but these have been removed from the plate. The objective of plate 3 is to show the potentiometric surface of the carbonate-rock aquifer rather than the potential interbasin flow directions based on the DSC model. The potential interbasin flow through the Confusion Range was not predicted by the DSC model but rather flow was northerly within Snake Valley.
TM17. BARCAS should determine and show the uncertainty around the interbasin flow estimates
BARCASS discusses that the input to the water balance accounting has inherent uncertainties (BARCASS, page 74). However, it does not attempt to put a distribution around the estimated interbasin flow values. Utilizing the distributions determined for discharge and that should be determined for recharge, as recommended above, BARCASS should place uncertainty bands around the interbasin flow estimates.
RESPONSE
Additional text has been added to the report to describe the uncertainty of the interbasin flow. Uncertainty in net interbasin inflow and outflow estimates has been added to figure 46. The accuracy of the interbasin flow estimates depends on the uncertainty of the ground-water recharge and discharge estimates, hydraulic gradients, and deuterium concentrations. As reported by Lundmark and Pohll (2007), several HAs do not have wells and therefore the aquifer heads and isotope concentrations are lacking. As part of the DSC modeling, a Monte Carlo uncertainty analysis was performed to evaluate the effects of recharge rates and concentrations by random sampling from uniform distributions of potential recharge rates and deuterium values for each model cell for a given realization and running the model to achieve the best fit for that realization. One thousand realizations were run for each parameter (Lundmark and Pohll, 2007).
BARCASS discusses that the input to the water balance accounting has inherent uncertainties (BARCASS, page 74). Failing to do so, the interbasin flow numbers shown on plate 4 will be considered as exact estimates.
RESPONSE
Several figures including bar charts show the uncertainty for the various hydrologic components. Recharge and discharge estimates are shown on plate 4. Plate 4 will not include uncertainty so as not to further clutter an already busy graphic; however, significant figures used on the plate were reduced to indicate the lack of precision.
