Draft

Analysis of Recent and Long-Term Natural Hydrology

Eastern Snake Plain Aquifer and Upper Snake Basin

January 5, 2007

Introduction

The Eastern Snake Plain Aquifer Ground Water Model (ESPAM) uses a period of record from 1980 to 2002 to calculate aquifer recharge. This same period of record is used for the Base Case Scenario.[1] The ESHMC[2] is considering the options available to update the model to 2006 and to revise the Base Case Scenario. At the last ESHMC meeting, the Committee discussed whether the “natural hydrology” for the 1980 to 2006 period is representative of the longer hydrologic record for the ESPA and Upper Snake Basin. This question is relevant for deciding whether or not the 1980 to 2006 period is long enough to represent the natural hydrology components influencing aquifer recharge in the model.

This memo provides information to assist in evaluating whether the post-1980 period is representative of the natural hydrologic record for the basin and aquifer. Specifically, we analyzed whether the 1980 to 2006 and 1985 to 2006 periods are representative of the longer record from 1929 to 2006. We used the 1930 to 2006 period as a standard of comparison for the more-recent periods because it includes many dry and wet periods and because some of the gage records are discontinuous prior to 1929.

Our analysis is based on examination of the following three variables:

1.  Unregulated Snake River inflow above American Falls (defined below)

2.  Northern tributary inflow to the ESPA

3.  Direct precipitation onto the ESPA

It is important to note this memo only evaluates the natural hydrology components of aquifer recharge in the post-1980 record. We did not examine water management components affecting aquifer recharge (such as reservoir storage, surface water diversions, irrigation efficiency, etc.). These water management components have changed significantly even during the last 25 years and should be adjusted as appropriate for each model scenario being considered- regardless of the period of record that is chosen for the scenario.

The results of the analysis show that the 1980 to 2006 period is generally representative of the 1929 to 2006 record, although the more-recent period exhibits greater variability than the longer-term record. The results are discussed in greater detail below.

Unregulated Snake River Inflow

The “unregulated Snake River inflow above American Falls” includes the following components:

·  Snake River near Heise (Gage No 13037500) minus reservoir storage and evaporation loss

·  Henry’s Fork River near Rexburg (Gage No. 13056500) minus diversions and reservoir storage and evaporative loss

·  Portneuf River at Pocatello (Gage No. 13075500)

·  Blackfoot River near Blackfoot (Gage No. 13068500)

The “unregulated Snake River inflow above American Falls” was used as an indicator of Snake River inflow because it provides a good approximation of the unregulated inflow to the Upper Snake River (absent river reach gains from ground water).

Analysis and Results

The annual series for the 78-years of record (1929-2006) is shown in Figure 1. The average flow is 7.15 million acre-feet for the 78-year record, 7.41 million acre-feet for the 1980-2006 period and 7.10 million acre-feet for the 1985-2006 period. The annual flow volumes are ranked in Table A-1 in Appendix A. This table also displays the ranking for averaged two-year and three-year periods. Table 1 compares annual statistics of annual unregulated inflow for the 1929-2006 period, with statistics for the 1980-2006 and 1985-2006 periods. The annual flow values for each period of record are plotted against the percent of years with smaller flows in Figure A-1.

This analysis considers the average, median, maximum, minimum, and standard deviation of each dataset, as well as certain measures of dryness, wetness, and percent exceedance. The following definitions were used to designate wet or dry years in the record:

·  Very Wet: The wettest 8 years in the record

·  Wet: The wettest 20 years in the record

·  Dry: The driest 20 years in the record

·  Very Dry: The driest 8 years in the record

The following procedure was used to compare the frequency of wet or dry years in the shorter periods against the longer period. The number of years designated as very wet, wet, dry or very dry are counted for each period. The number of years designated as very wet, wet, dry or very dry are then computed as a percentage of the total years in each period.

The same examination described above was conducted for two-year and three-year averages to examine multiple-year trends in the unregulated inflow data. Table 2 compares statistics on the two-year flows for the 1929-2006 period, with statistics for the 1980-2006 and 1985-2006 periods. Table 3 summarizes statistics on the three-year average unregulated inflow data. Figures A-2 and A-3 compare the three periods of record (1929 to 2006; 1980 to 2006; 1985 to 2006) based on average unregulated inflow, for two-year and three-year averages, based on frequency statistics.

The results in Table 1 show the single year average unregulated inflows during the 1980 to 2006 and 1985 to 2006 periods are within 4% of the longer-term 1929 to 2006 average. The 1980-2006 and 1985-2006 periods have similar standard deviations to the 1929-2006 record. All three periods of records experience very wet, wet, dry, and very dry years within 5% of the same frequency, except in the following two cases. The 1985-2006 period has 10% more dry years than the full 78-year period. The 1980-2006 period has 9% more very wet years than the full period.

Based on the results for the two-year and three-year average unregulated flow shown in Tables 2 and 3, both the more recent periods have averages within 4% of the longer period. The standard deviation for the shorter periods is similar but slightly more than the longer period. The three periods of record experience very dry years within 3% of the same frequency. The 1980-2006 and 1985-2006 periods both have approximately 10% to 15% more dry, wet, and very wet periods for the two-year and three-year averages compared with the full 78-year period.

Northern Tributary Inflow

The same comparison for the three periods of record and three durations was completed for the total Northern Tributary Inflow. Total Northern Tributary Inflow is the sum of the annual volume of flow for the gages on the Big Wood, Little Wood, and Big Lost rivers (USGS gages 13142500, 13148500, and 13127000).

Analysis and Results

The annual volume of flow is shown on Figure 2. The average flow for the 1929-2006 period is 671,800 acre-feet. The average flow is very similar for the 1980-2006 period (~3% difference), but is 9 percent lower for the 1985-2005 period. The average annual flow volumes for the one, two, and three year average flows are presented in Table B-1. Statistics for the annual flows for the three periods of record are summarized in Table 4. The annual flow values for each period of record are plotted against the percent of years with smaller flows in Figure B-1.

The same examination was conducted for two-year and three-year averages of the Northern Tributary inflow. Table 5 presents a statistical summary of the two-year average flows for the three periods of record. Table 6 presents the same information for the three-year average flows. Figures B-2 and B-3 compare the three periods of record based on inflow frequency statistics.

The results in Table 4 show the single year average northern tributary inflows during the 1980 to 2006 and 1985 to 2006 periods are within 3% and 10%, respectively of the longer-term 1929 to 2006 average. The 1980-2006 and 1985-2006 periods have standard deviations similar to the 1929-2006 record. The 1980-2006 period of record experiences 5% more very dry years and 7% more wet and dry years than the longer record. The 1985-2006 period of record experiences 8% more very dry years, 15% more dry years, and 5% less very wet years than the 78-year period.

Based on the results for the two-year and three-year average tributary flow shown in Tables 5 and 6, 1980-2006 is 2% more and 1985-2006 is 6% to 9% less than the longer period. The standard deviation for the shorter periods is similar but slightly more than the longer period. The three periods of record experience very dry and wet years within 12% of the same frequency. The 1980-2006 and 1985-2006 periods both have between 11% and 24% more dry, and 4% to 15% more very wet periods for the two-year and three-year averages compared with the full 78-year period.

ESPA Precipitation

Shoshone Precipitation Record

Analysis and Results

The same comparison for the three periods of record and three durations was completed for the annual precipitation at the Shoshone gage. Total Shoshone precipitation is shown on Figure 3. The average for the 1929-2006 period is 10.12 inches. The average precipitation is very similar for both the 1980-2006 and 1985-2006 periods, 10.33 inches and 9.99 inches, respectively. The average annual precipitation values for the one, two, and three year average are presented in Table C-1. Statistics for the annual precipitation for the three periods of record are summarized in Table 7. The annual values for each period of record are plotted against the percent of years with smaller values in Figure C-1.

The same examination was conducted for two-year and three-year averages of the Shoshone precipitation. Table 8 presents a basic statistical summary of the two-year average values for the three periods of record. Table 9 presents the same information for the three-year average values. Figures C-2 and C-3 compare the three periods of record based on a frequency analysis for two and three year average precipitation at Shoshone gage, respectively.

The results in Table 7 show the single year average precipitation gage during the 1980 to 2006 and 1985 to 2006 periods are within 2% of the longer-term 1929 to 2006 average. The 1980-2006 and 1985-2006 periods have standard deviations, 9% and 3% greater than the 1929-2006 record. All three periods of records experience very wet and very dry years within 5% of the same frequency. All three periods of records experience wet and dry years within 8% of the same frequency.

Based on the results for the two-year and three-year average precipitation shown in Tables 8 and 9, both the more recent periods have averages within 1% of the longer period. The standard deviation is 11% to 23% greater for the shorter periods than the longer period. The three periods of record experience very dry years within 5% of the same frequency. The 1980-2006 and 1985-2006 periods have very wet years 4%-12% more than the full 78-year period. The two-year average precipitation for all three periods has wet and dry years within 4% of the same frequency. The three-year average precipitation for all three periods has wet and dry years within 10% of the same frequency.

Aberdeen Experiment Station Precipitation Record

Analysis and Results

The same comparison for the three periods of record was completed for the annual precipitation at the Aberdeen gage. Total Aberdeen precipitation is shown on Figure 4. The average for the 1929-2006 period is 8.80 inches. The average precipitation is very similar for both the 1980-2006 and 1985-2006 periods, at 9.14 inches and 8.69 inches respectively. The average annual precipitation values for the one, two, and three year average are presented in Table C-2. Basic statistics for the annual precipitation for the three periods of record are summarized in Table 10. The annual values for each period of record are plotted against the percent of years with smaller flows in Figure C-4.

The same examination was conducted for two-year and three-year averages of the Aberdeen precipitation. Table 11 presents a basic statistical summary of the two-year average values for the three periods of record. Table 12 presents the same information for the three-year average values. Figures C-5 and C-6 compare the three periods of record based on a frequency analysis.

The results in Table 10 show the single year average precipitation gage during the 1980 to 2006 and 1985 to 2006 periods are within 4% of the longer-term 1929 to 2006 average. The 1980-2006 and 1985-2006 periods have standard deviations, 26% and 30% greater than the 1929-2006 record. All three periods of records experience very dry years within 3% of the same frequency. All three periods of records experience dry, wet, and very wet years within 16% of the same frequency.

Based on the results for the two-year and three-year average precipitation shown in Tables 11 and 12, both the more recent periods have averages within 2% of the longer period. The standard deviation is 14% to 17% greater for the shorter periods than the longer period. The three periods of record experience very dry years within 8% of the same frequency. The 1980-2006 and 1985-2006 periods have very wet years 3%-8% more than the full 78-year period. The two-year average precipitation for all three periods has wet and dry years within 11% of the same frequency. The three-year average precipitation for all three periods has wet and dry years within 6% of the same frequency.

Dubois Precipitation Record

Analysis and Results

The same comparison for the three periods of record and three durations was completed for the annual precipitation at the Dubois gage. Total Dubois precipitation is shown on Figure 5. The average for the 1929-2006 period is 12.06 inches. The average precipitation is 8 percent higher (13.25 inches) for the 1980-2006 period, and 6 percent higher (12.53 inches) for the 1985-2006 period. The average annual precipitation values for the one, two, and three year average are presented in Table C-3. Basic statistics for the annual precipitation for the three periods of record are summarized in Table 13. The annual values for each period of record are plotted against the percent of years with smaller flows in Figure C-7.