2015 Island Park Dam Flow Releases: the Numbers

Wild trout need water of appropriate quantity and quality. Quantity and quality of water released from Island Park Reservoir during June and July of 2015 continue to be the subject of much discussion and concern, particularly following release of around 2,000 cfs July 21-23 to provide flow to test turbines at the Chester Dam hydroelectric project. This blog just covers water quantity, specifically flow releases from Island Park Reservoir. My objectives for this water-quantity blog are to present the data on flow releases from Island Park Dam from June 1 through August 6, 2015 and to put those releases in context of flow releases in some other years.

A future blog will cover water quality, as we are still awaiting test results from the lab that analyzes our water samples. HFF and Idaho Department of Environmental Quality, with assistance from Idaho Department of Fish and Game, have been monitoring water quality upstream of, in, and downstream of Island Park Reservoir all summer. We designed and planned an intensive water-quality monitoring campaign last winter in anticipation of high irrigation deliveries in 2015—and in response to numerous questions and well-founded concerns from outfitters, guides, and anglers over the past few years. So, stay tuned for water quality information coming in a few weeks.

Data sources and types

All of the data I report here came directly from the US Geological Survey water information database. Data for previous water years are reported as daily flow, which is the flow averaged over a given calendar day (midnight to midnight). Data from the current water year are known as “continuous” or “instantaneous” records; these data are actually recorded every 15 minutes. The 15-minute data are not available for water years very far back in the historical record, which is why daily flow is used for previous water years. However, I am presenting the 2015 data in its 15-minute form so that the exact dates and times of flow adjustments are readily visible. I downloaded the 2015 data on August 6; the most recent adjustment of the relationship between water depth and streamflow was made on July 28, so the 2015 flow data have all been recently adjusted for shifts in this relationship due to plant growth. Click here for details on this aspect of streamflow measurement. In fact, flow at the Island Park gage has been measured in the field five times so far this summer (Table 1), a very high frequency of measurement. Following each measurement, the gage was adjusted, resulting in a high degree of accuracy in the reported flow.

Table 1. Summary of USGS field measurements of flow at Island Park Dam made since June 1.



Measured flow (cfs)
















Flow releases at Island Park, June 1 through August 6, 2015 

Figure 1 shows 2015 continuous flow (15-minute data) as the thin black line. The five field measurements listed in Table 1 appear as red asterisks. These also indicate the times at which the gage was adjusted for plant growth in the channel. The highest degree of accuracy in the reported flow occurs near the times of these field measurements. The gold diamonds show mean flow for each day over water years 1933-2014. The gage station at Island Park was installed in 1933; Island Park Dam first stored water in 1938. Some comparison years are also included in this figure. However, before discussing the comparisons, here are some statistics from 2015.

Figure 1. Discharge in the Henry’s Fork at Island Park Dam, June 1 through August 6, 2015. Long-term mean daily flow, 2015 field measurements, 1999 daily flow, and 2001 daily flow are shown for comparison.

Flow adjustments

Fourteen flow adjustments were made at the dam in 2015 (Table 2). Eight of these were flow increases to accommodate increasing irrigation demand (click here for more detail on 2015 irrigation demand). One additional flow increase was made to deliver the water for the Chester hydroelectric test.  Five of the adjustments were flow decreases, made as irrigation demand decreased. On July 22, partway through the delivery of the Chester test flow, the Island Park hydroelectric plant automatically tripped offline due to an electrical supply disruption; this resulted in an unexpected decrease in streamflow that lasted about two hours before the plant was manually returned to operation.

Table 2. Flow adjustments made at Island Park Dam. Positive adjustments are flow increases; negative adjustments are flow decreases.


Adjustment (cfs)




Increased irrigation demand



Increased irrigation demand



Increased irrigation demand



Increased irrigation demand



Increased irrigation demand



Increased irrigation demand



Increased irrigation demand



Increased irrigation demand



Decreased irrigation demand



Decreased irrigation demand



Decreased irrigation demand



Chester hydroelectric test


-/+ 700

Unexpected shutdown/restart of IP hydroelectric plant



End of Chester test and decreased irrigation demand



Decreased irrigation demand

Early-season irrigation demand

Island Park Dam release was at its long-term mean between June 12 and June 17. Between June 17 and July 16, releases from Island Park Dam were above the long-term average due to irrigation demand that was higher than usual for that time of year. During this 30-day period, the maximum instantaneous flow reported by USGS was 1,650 cfs, recorded on June 30. Maximum daily flow was 1,630 cfs, also on June 30. Mean discharge between the respective flow adjustments on June 17 and July 16 was 1,428 cfs. Long-term mean daily flow over this time period is 1,014 cfs.

Chester hydroelectric test release

The release lasted 62 hours, from 7:15 a.m. on July 21 to 9:15 p.m. on July 23. Peak instantaneous flow was recorded as 2,060 cfs. Mean flow over the 62-hour period was 1,965 cfs. Maximum daily flow (24-hour average over a calendar day) was 1,940 cfs on July 22. The release used 4,408 acre-feet of storage above what was required for irrigation demand during the test period. Storing this amount of water over the October-March storage period will reduce outflow from the dam next winter by an average of 12.2 cfs. Based on our statistical relationship between winter flow and trout recruitment and on predicted outflow next winter, this reduction in winter flow is expected to reduce recruitment of 2-year old rainbow trout into the population in 2017 by 4.2%, from 2,030 to 1,945 fish.

As originally planned, the Chester hydroelectric release would have lasted 101 hours, from 4 p.m. on June 21 through 9 p.m. on June 25. Peak instantaneous flow would have been 2,400 cfs, and the release would have used 9,166 acre-feet of storage above what was required for irrigation. Storing this amount of water would have reduced winter outflow by 25.4 cfs and expected recruitment of 2-year old rainbow trout by 8.9%, from 2,030 to 1,849 fish. See Brandon’s blog for more information about the Chester hydroelectric test, both as originally scheduled and as it occurred.

Early summer 2015 in comparison to other years

There is no way to choose a set of comparison years that is relevant to everyone, because each person’s history with the river is unique. However, a 20-year comparison window is reasonable for several reasons: it is a commonly used planning horizon for natural-resources management agencies, it is long enough to include a variety of water and climatic conditions but short enough that climatic conditions have not shifted too much, and it is a time period over which water management policies and practices on the Henry’s Fork have been relatively stable.

Maximum daily flow

One comparison that can be made is magnitude of the maximum daily flow that has occurred in each of the last 20 water years. As mentioned above, the maximum daily flow recorded during 2015 was 1,940 cfs, on July 22, during the Chester flow test. This value tied 2006 for 6th place among peak flows recorded since water year 1996 (Table 3).

Table 3. Maximum daily flow releases at Island Park Dam, water years 1996-2015.



Maximum daily flow (cfs)





Jun 8-9




May 21




May 16




May 29




May 15

6 tie



May 23

6 tie



Jul 22




Jul 30




Jul 8




Jun 9




Jul 2




Jul 26-27




Jul 14




May 13




Aug 4-5

16 tie



Jul 1

16 tie



Aug 5




Jul 14




Jul 21




May 27

It is apparent from Table 3 that the highest peak flows over the past 20 years—those of 2,000 cfs or greater—have occurred in the spring. These events all occurred during wet years, when snowmelt runoff was high and Island Park Reservoir was full. When peak flows have occurred later in the summer, they are the result of irrigation delivery and typically are lower than the peak flows that occur because of runoff through a full reservoir. The 2015 peak flow release was neither the result of high spring runoff nor high irrigation delivery, although its timing fell in line with irrigation delivery peaks. In magnitude, it fell between the high spring-runoff peaks and the high irrigation-delivery peaks.

Figure 1 shows daily flow in 1999 (wet year) and 2001 (dry year), providing a more detailed view of the difference between high runoff and high irrigation delivery. In 1999, high runoff resulted in above-average flow for the first 27 days of June; flow exceeded equaled or exceeded 2,000 cfs that year from May 31 to June 15. In 2001, irrigation demand began early, although not quite as early as in 2015. Flow in 2001 was above average from June 20 until September 5 (not shown on graph), compared with June 17 to July 16 in 2015. Mean flow in 2001 over the June 17-July 16 period was 1,364 cfs, compared with 1,428 cfs in 2015. The maximum daily flow of 1,640 cfs on July 8, 2001 was essentially the same as the daily maximum of 1,630 cfs on June 30, 2015, which was the maximum irrigation flow recorded in 2015. Of course, higher flows were recorded on the three days of the Chester flow test, but these flows were above what was being delivered for irrigation.

Comparisons with the 1980s

For those of you whose history with the Henry’s Fork goes back into the 1980s, I have included 1983 and 1988 as comparative years in Figure 2. You will recall that 1983 was a moderately wet year and 1988 was a dry year—the year of the Yellowstone fires. In 1983, flow was above average during all of June and half of July, exceeding 2,000 cfs on several days during early June. Flow was also above average late in the summer, equaling or exceeding 2,000 cfs from August 3 through September 1. In 1988, irrigation demand began early, although again, not quite as early as in 2015. Nonetheless, irrigation-delivery flows in 1988 exceeded those in 2015 from June 16 through June 24. Flow releases in 1988 averaged 1,610 cfs over the month of July, peaking at 1,740 cfs July 8-10.

Figure 2. Discharge in the Henry’s Fork at Island Park Dam, June 1 through August 6, 2015. Long-term mean daily flow, 2015 field measurements, 1983 daily flow, and 1988 daily flow are shown for comparison.


As I predicted back in April and summarized two weeks ago, the spring and early summer of 2015 was the third driest on record. Irrigation demand was high in the middle of June, necessitating release of irrigation from Island Park Reservoir much earlier than usual. However, Island Park releases during mid-June of 2015 were lower than mid-June releases typical of wet years, including 1983, 1984, 1986, 1995, 1999, and 2011. Flows during late June of 2015 were similar to late-June flows in other dry years, including 1988, 1992, 1994, and 2001. Maximum irrigation-delivery flows in 2015 were typical of maximum irrigation-delivery flows in other dry years, but they occurred earlier in 2015 and were shorter in duration (about three weeks in 2015 versus six weeks in most years). Although high flow releases have occurred at Island Park Dam for reasons other than irrigation delivery (e.g., to empty the reservoir to make repairs to the dam, as occurred in 1984), short-duration flow events such the Chester test have been rare in recent years. Under previous management regimes, however, large changes in outflow for short periods of time were common, especially during the winter. The magnitude of the Chester release, around 2000 cfs, was very rare for July or August, but it was not unusual relative to peak flows that occur for extended periods of time during May and June of wet years.