Eastern and southern Idaho
Eastern Snake Plain and Upper Snake
The Eastern Snake Plain Aquifer lost an estimated 600,000 acre-feet of stored water from spring 2025 to spring 2026. State modeling credits about a decade of recharge and pumping cuts with keeping 2.8 million more acre-feet underground.
Research snapshot ·
Ready to paste into a message or email.
Copy this address: https://idahowaterstatus.com/explore/eastern-snake-plain
TL;DR
The short answer
- The state water agency estimated that the aquifer lost 600,000 acre-feet from spring 2025 to spring 2026, using water levels in 329 wells.
- By the state’s model, recharge and pumping cuts have kept 2.8 million more acre-feet in the aquifer. That is a running total over about a decade, not a 2026 gain.
- Seven Upper Snake reservoirs held about 612,000 acre-feet on November 5, 2026, by provisional federal readings. That is about 15% of capacity and about 40% of the median for that date.
What this means for you
Before you compare aquifer numbers, check whether each one is a one-year change or a model comparison built up over years. Read reservoir storage with its date, and compare it with the median for that date as well as percent full. If you farm or depend on a canal or well, ask your irrigation district or canal company what this season’s numbers mean for your water.
Location context · north is up
The Snake River arcs across southern Idaho
Henrys Fork joins the Snake in the east. The Big Lost River runs out onto the plain and ends there, short of the Snake.
- Snake River
- Henrys Fork
- Big Lost River
The Idaho outline comes from the U.S. Census Bureau. The river lines use Idaho Department of Water Resources (IDWR) data that is current as of 1996. Both are simplified. Orientation only: the lines do not mark today’s river channels, property lines, aquifers or water-provider service areas. Your water bill names the provider that serves you.
Terms used in this guide
- Acre-foot
- Enough water to cover one acre of land one foot deep, about 326,000 gallons.
- Storage change
- How much the water stored in an aquifer or reservoir rose or fell between two dates.
- Managed recharge
- Water sent into the ground on purpose, under a planned program, to add to an aquifer.
- Without-management case
- A model estimate of what aquifer storage would be with no recharge program or pumping cuts.
The Eastern Snake Plain Aquifer is the groundwater beneath the eastern Snake River Plain. Farms and towns pump from it, and it pours out as springs that keep the Snake River flowing below Milner Dam. Upstream, the Bureau of Reclamation tracks stored water in seven Upper Snake reservoirs. Amounts here are in acre-feet. One acre-foot would cover an acre of land one foot deep, about 326,000 gallons.
How is the aquifer tied to the Snake River?
Water soaks into the Eastern Snake Plain Aquifer from leaky canals, irrigated fields and managed recharge. It leaves through wells and through springs along the Snake River, including Thousand Springs. The Idaho Department of Water Resources (IDWR) publishes the Eastern Snake Plain Aquifer Model. A committee of state, federal, private and university hydrologists works to improve it. Their goals include showing how groundwater and rivers interact and estimating how uncertain the results are.
Irrigation water seeping into the ground once made the springs grow. The Twin Falls North Side tract was first irrigated in 1908. By 1913, the flow of Blue Lakes Spring had more than doubled. Blue Lakes and three nearby springs went from a combined 680 cubic feet per second (a flow rate) in 1902 to about 1,400 by 1917. That record comes from a 1964 U.S. Geological Survey (USGS) report. A 1992 USGS fact sheet says spring flows began to fall in the early 1950s, when well pumping for irrigation grew quickly. Later, the move from flood to more efficient sprinkler irrigation meant less water soaked in. The 1987–91 drought made the decline worse.
In July and August, irrigators above Milner Dam often divert the Snake’s entire flow. IDWR says flow past Milner drops to zero under normal irrigation-season operations. This is legal. The 1984 Swan Falls settlement kept the minimum daily flow at Milner at zero, which let existing uses above the dam continue.
Below the dam, springs rebuild the river. In July, it grows from essentially zero at Milner to about 4,500 cubic feet per second at King Hill, mostly from springs. In high-runoff years, water from above Milner makes up a large share of the river below it. In 2009, the state and Idaho Power reaffirmed the settlement. Their agreement says the river below Milner can at times be almost entirely groundwater. For that reason, the aquifer “must be managed as an integral part of the Snake River.”
Farther downstream, the settlement protects minimum flows at the Murphy gage, just below Swan Falls Dam. The floors are 5,600 cubic feet per second from November through March and 3,900 from April through October. In late March 2026, the river fell below the 5,600 floor on six days, measured as a three-day average. A May 8 staff memo to an Idaho Water Resource Board committee said low inflows and high irrigation diversions caused the shortfall. Spring flow from the aquifer was below average. On May 21, the board agreed to make its Palisades Reservoir storage available to Idaho Power to offset the shortfall.
Rivers and groundwater are connected
Water can move between the surface and the aquifer below. A change in one place can show up somewhere else or in a later season.
- Canal and land surfaceSeepage and infiltration can recharge groundwater.
- Pumping wellWater is withdrawn from connected pores or fractures underground.
- River exchangeGroundwater can feed a river; other reaches can lose water to the aquifer.
The drawing is general and does not match the ground under any one place. Arrows show where water can go, not how much or how fast. Where the aquifer’s water comes from differs by region. In the Treasure Valley, the largest sources are seepage from canals and from irrigated fields. Under the Rathdrum Prairie and Spokane Valley, seepage from the Spokane River and nearby lakes supplies about two-thirds of the water entering the aquifer.
Evidence: Treasure Valley model Rathdrum–Spokane framework Eastern Snake Plain model
How much water did the aquifer lose in 2025–26?
On August 26, 2026, IDWR reported that aquifer storage fell by 600,000 acre-feet from spring 2025 to spring 2026. The estimate, from IDWR hydrogeologist Mike McVay, rests on water levels in 329 monitoring wells measured before the 2026 irrigation season. From 2024 to 2025, storage showed no net change. The figure is a change in stored water, not the total amount left underground.
IDWR also expected spring flows from the aquifer to decline during 2026, both at Thousand Springs and in the reach from Blackfoot to Minidoka. That was a forecast in late August. Later flow records will show whether it came true.
Sources: [2]
If storage fell, are recharge and pumping cuts helping?
IDWR’s modeling says yes, over the long run. It estimates that three efforts together have kept 2.8 million acre-feet more water in the aquifer. They are the Idaho Water Resource Board’s managed recharge program, recharge by cities and groundwater districts, and pumping cuts by groundwater districts. “We are definitely moderating the decline,” said IDWR hydrogeologist Alex Moody.
That 2.8 million is a running total, not one year’s effect. The model compares observed storage with an estimate of storage without that work. In the release’s chart, the two lines split after 2014, and the gap grows over the following decade. The gap was about the same in 2025 as in 2026. Read it as water kept over many years, not as savings against this year’s 600,000 acre-foot loss.
A savings account shows the idea. Say a balance fell from 100 to 90 over ten years, but without deposits it would have fallen to 70. The balance still dropped, yet the deposits kept 20 in the account over those ten years. These numbers are made up to show the comparison.
Gains can fade. McVay said the aquifer constantly “leaks” at many locations along the Snake River, so gains can be “fleeting.” IDWR Deputy Director Brian Patton said “we need to stay the course” with recharge and pumping cuts. Otherwise, he said, “we will be seeing our aquifer volume going down much lower than it is now.”
How much water is in the Upper Snake reservoirs?
On November 5, 2026, the Bureau of Reclamation’s provisional readings showed 611,537 acre-feet in seven Upper Snake reservoirs. They are Jackson Lake, Palisades, Grassy Lake, Island Park, Ririe, American Falls and Lake Walcott. Together they hold up to 4,045,695 acre-feet, so they were 15.1% full. Henrys Lake and Little Wood appear on the same diagram but are not in this total. Jackson Lake is in Wyoming, so this water system crosses state lines.
American Falls held the most, 293,878 acre-feet, and Palisades held 146,160. Those two held about 72% of the group’s water. The 15.1% figure divides total contents by total capacity. Averaging the seven reservoirs’ own percentages would give tiny Grassy Lake the same weight as American Falls.
Storage is also low for the date. For November 5, Reclamation’s median for this group over water years 1991–2020 (October 1, 1990, through September 30, 2020) is about 1.54 million acre-feet. This year’s storage was about 40% of that. Reclamation adds the seven reservoirs together for each day before finding the median. The chart below tracks 2026 readings against that baseline.
A November reading shows where storage stands going into winter. Next season’s supply still depends on winter snow and spring runoff. Keep reservoir storage separate from the aquifer estimate, too. One is the water held on a single day and the other is a change over a year, so adding them gives a meaningless number.
How 2026 reservoir storage compares with past years
On November 5, the Boise reservoirs were right at their historical median for the date. The Upper Snake reservoirs held about 40% of their median.
Acre-feet stored · selected daily readings and the historical median for the same calendar day
An acre-foot covers one acre one foot deep. The median is the middle historical value for that calendar day; it is not the reservoir’s capacity.
Boise system
November 5: 367,906 acre-feet · historical median 368,137
Anderson Ranch, Arrowrock and Lucky Peak. The vertical scale differs between the two charts.
Read all Boise system values and download the dated originals
| Date in 2026 | Observed | Historical median |
|---|---|---|
| Jan 1 original CSV | 528,853 | 436,279 |
| Feb 1 original CSV | 601,246 | 492,527 |
| Mar 1 original CSV | 662,860 | 518,354 |
| Apr 1 original CSV | 808,101 | 581,517 |
| May 1 original CSV | 888,032 | 639,161 |
| Jun 1 original CSV | 917,723 | 808,635 |
| Jul 1 original CSV | 794,089 | 872,167 |
| Aug 1 original CSV | 640,842 | 706,061 |
| Sep 1 original CSV | 467,765 | 529,970 |
| Oct 1 original CSV | 348,138 | 376,142 |
| Nov 1 original CSV | 364,131 | 359,742 |
| Nov 5 original CSV | 367,906 | 368,137 |
Upper Snake system
November 5: 611,537 acre-feet · historical median 1,539,741
American Falls, Jackson Lake, Palisades, Ririe, Grassy Lake, Island Park and Minidoka (Lake Walcott). The vertical scale differs between the two charts.
Read all Upper Snake system values and download the dated originals
| Date in 2026 | Observed | Historical median |
|---|---|---|
| Jan 1 original CSV | 1,849,545 | 2,266,566 |
| Feb 1 original CSV | 2,214,638 | 2,563,354 |
| Mar 1 original CSV | 2,519,299 | 2,793,720 |
| Apr 1 original CSV | 2,990,030 | 2,890,573 |
| May 1 original CSV | 2,925,230 | 3,016,603 |
| Jun 1 original CSV | 2,471,577 | 3,349,054 |
| Jul 1 original CSV | 2,029,214 | 3,461,258 |
| Aug 1 original CSV | 1,300,834 | 2,680,106 |
| Sep 1 original CSV | 600,773 | 1,957,941 |
| Oct 1 original CSV | 349,192 | 1,454,511 |
| Nov 1 original CSV | 557,356 | 1,498,711 |
| Nov 5 original CSV | 611,537 | 1,539,741 |
The points are single-day readings, not monthly averages. The median comes from water years 1991–2020 (October 1, 1990–September 30, 2020). The U.S. Bureau of Reclamation adds up each system’s reservoirs before it calculates the median. The Upper Snake group is a separate system farther east, including Jackson Lake in Wyoming. Its low storage describes that system, not the Boise River reservoirs. The readings are provisional and can change. They are not a water allocation or a forecast.
Evidence: Reclamation seasonal records
What should a recharge or water-saving plan tell you?
A clear plan says where the water comes from, when it is available and where the benefit should appear. Model results should name the model version, the comparison case and the time span. A regional model helps compare strategies. Questions about one well, spring or headgate need more local evidence.
Ask how a plan changes the water used up, which can differ from the water diverted. USGS keeps withdrawals, the water taken from a river or aquifer, separate from consumptive use. Consumptive use is water that evaporates, is taken up by plants or crops, or otherwise leaves the nearby water system. Water that is withdrawn but not used up returns to streams or groundwater as return flow. So a field’s diversion is not the same as what it removes from the system. Cutting irrigation can also cut the recharge and return flows that other water users depend on.
To judge progress, follow the yearly aquifer estimate, measured spring and river flows, and the modeled benefit together. Over the years, they show whether recharge and pumping cuts are keeping pace with the aquifer’s losses.
Follow the sources
Read our source summaries, or open the original documents to check the details.
-
IDWR: Eastern Snake Plain Aquifer Model
Model overview, committee goals, model reports and recharge tools
Original document: IDWR: Eastern Snake Plain Aquifer Model (opens in a new tab) -
IDWR: August 2026 aquifer update
News release, August 26, 2026, pages 1–3: storage change from 329 wells, recharge benefit and charts
Original document: IDWR: August 2026 aquifer update (opens in a new tab) -
Reclamation: Upper Snake reservoirs
Daily page with provisional readings for November 5, 2026, read November 6. The diagram changes each day. Seven reservoirs added up; Henrys Lake and Little Wood left out.
Original document: Reclamation: Upper Snake reservoirs (opens in a new tab) -
USGS: water withdrawn, used up and returned
Introduction: what withdrawal, consumptive use and return flow mean
Original document: USGS: water withdrawn, used up and returned (opens in a new tab) -
Reclamation: reservoir storage by season
November 5, 2026 data file and first-of-month readings; medians for water years 1991–2020; provisional
Original document: Reclamation: reservoir storage by season (opens in a new tab) -
USGS: groundwater for irrigation in the Snake River basin (1964)
PDF page 180: how irrigation raised spring flows, 1902 to 1917
Original document: USGS: groundwater for irrigation in the Snake River basin (1964) (opens in a new tab) -
USGS: springs from Milner Dam to King Hill (1992)
Two-page fact sheet, linked from the publication page: diversions above Milner, and why spring flows rose and then fell
Original document: USGS: springs from Milner Dam to King Hill (1992) (opens in a new tab) -
Swan Falls settlement: 2009 framework
Exhibit 2, PDF page 19: zero minimum flow at Milner and the Murphy flow floors
Original document: Swan Falls settlement: 2009 framework (opens in a new tab) -
IDWR: 2024 Snake River Basin moratorium order
Findings 25 and 71, PDF pages 15 and 23: flow past Milner in irrigation season
Original document: IDWR: 2024 Snake River Basin moratorium order (opens in a new tab) -
Water Resource Board committee packet, May 2026
Staff memo, PDF pages 8–11: river flow from Milner to Swan Falls and the March 2026 shortfall
Original document: Water Resource Board committee packet, May 2026 (opens in a new tab) -
Water Resource Board: Resolution 30-2026
PDF pages 49–50: offset for the March 2026 Swan Falls shortfall
Original document: Water Resource Board: Resolution 30-2026 (opens in a new tab)
Cite or share this guide
Idaho Water. “Eastern Snake Plain and Upper Snake.” Research snapshot: 2026-11-06. Cite the original source for a specific figure.