Hoover Dam spillway towers, 2024 Credit: Wikimedia/CC BY 4.0
With No Relief in Sight
From 2022 to 2025
We first posted about the Southwest megadrought in early 2022. At that time we thought that it was “likely the drought will continue through 2022.” Little did we know that we would still be living with it almost four years later.
Even worse, recent research suggests that the current drought-stricken environment of the Southwest could persist “through at least 2100,” driven by global warming.
Because of the megadrought the flagship water resource of the Southwest, the Colorado River, has been weakening for decades. Now, water managers in the Southwest are now facing the possibility that the river may never fully recover.
Water is a Perpetual Challenge in the Southwest
Since the early 20th century, the western US has steadily become drier, experiencing the most extreme drought intensification in the country. At the same time, the eastern US has become wetter. As we discussed in a recent post, global warming is amplifying evapotranspiration in many parts of the world — the warmer atmosphere able to extract more water from rivers, soil and vegetation, causing or magnifying drought. As we saw in the “flash flood summer” of 2025, the ability of the warmer atmosphere to hold more water can also lead to intense rainfall events in circumstances where water is more abundant.
In the last century, the principal causes of drought in the western US were periodic deficiencies in rainfall and snowfall. However, since 2000 the principal driver of drought has become the steadily warming atmosphere. A recent study found that during the 2020-2022 extreme drought, evaporation accounted for 61% of the drought’s severity, while reduced precipitation was responsible for only 39%.
Nowhere in the western US is facing a greater challenge from drought than the seven states that encompass the Colorado River Basin.
The Pivotal Role of the Colorado River
The Colorado River is approximately 1,400 miles long, running from the Continental Divide in Rocky Mountain National Park, Colorado, to the Gulf of California in Mexico. The Colorado River Basin, shown below, covers approximately 250,000 square miles.
Nearly all of the water in the Colorado River system comes from snow that falls in the Rocky Mountains in the Upper Basin. About 85% of the Colorado Basin’s flows come from just 15% of the basin. Large areas of the basin are arid or semi-arid.
The Colorado River basin is managed by the US Bureau of Reclamation (USBR) and other agencies to provide water and hydroelectric power to roughly 40 million people—most notably the cities of San Diego, Las Vegas, Phoenix, and Los Angeles—and irrigation for 4 to 5 million acres of farmland in the Southwest. The river water is allotted to states (including tribal lands) and Mexico through laws stemming from the 1922 Colorado River Compact.
The river is divided into two sub-basins. The principal “upper basin” states are Colorado, New Mexico, Utah, and Wyoming, while Arizona, California, and Nevada comprise the “lower basin” states. The Colorado River Compact allotted each basin 7.5 million acre-feet (maf) of water per year from the river, which the states share. (An acre-foot is the amount of water that would cover an acre to a depth of one foot.) The lower basin is allotted an additional 1 maf per year to account for diversions from the Gila River in Arizona. California was granted the largest state allotment — 4.4 maf. Native American tribes are provided additional allocations, along with a 1.5 maf allocation to Mexico. Agriculture consumes 70 – 80% of the river allotments, primarily for irrigation.
Unfortunately, the Colorado River Compact allotted basin states more water than the watershed can sustainably deliver today, under the assumption that the average annual outflow of the basin was 18 maf. This was true in the early 1900s, when the original Compact was drawn up. However, the long term average leading up to 2000 proved to be closer to 14.3 maf/year, leading to contentious negotiations between the basin states.
It gets worse. Since 2000, climate change, hotter and drier conditions and overuse have combined to reduce the recent average to roughly 12.5 maf/year — much lower than the 15 maf/year initially allocated to the basin states. The shortfall is leading to severe depletion of the principal Colorado River reservoirs, Lake Mead and Lake Powell, and critical water shortages.
Reservoirs and Dams
Numerous dams and reservoirs have been built in the Colorado River Basin to store spring snowmelt and the excess water of occasional wet years, to make that water available in other seasons and to manage the variable runoff. The larger dams also support hydroelectric power generation.
There are 8 major dams and reservoirs, and 7 diversion dams, on the main stem of the Colorado River, and hundreds of dams and reservoirs along its tributaries. In total, the reservoir system has the capacity to allow complete utilization of the watershed’s runoff. Total storage capacity across the basin (excluding the Gila River in Arizona) is about 62 maf. About 80% of the system’s storage capacity is in the two largest reservoirs, Lake Mead (26.1 maf) and Lake Powell (23.3 maf), with the remainder spread across dozens of major reservoirs (>0.1 maf) and hundreds of smaller ones, primarily in the Upper Basin.
Lake Powell and Lake Mead water levels are measured in feet above sea level, with maximum or “full pool” elevations roughly 3,700 ft for Lake Powell and 1,229 ft for Lake Mead. Water levels fluctuate significantly in both reservoirs. Both were last full in the 1980s, followed by decades of drought and decline. By January 1, 2026, USBR estimates Lake Powell will be at 3,538 ft., while Lake Mead will be in the range of 1,050 to 1,075 ft.
For the past 25 years, Colorado River water use has exceeded the river’s annual supply by an average of 1 maf/year, as shown in the graph below. This has created water shortages (especially in the lower basin and for Mexico), reduced hydroelectric power generation, and resulted in historically low water levels in Lake Powell and Lake Mead.
Hydroelectric Power Generation
Power generation by Colorado River dams is significant, both within and outside the basin.
Lake Powell’s Glen Canyon Dam can produce 1,320 megawatts (MW) when the lake is full — roughly the same as a large fossil fuel plant. The power is distributed to Wyoming, Utah, Colorado, New Mexico, Arizona, Nevada, and Nebraska.
When Lake Mead is full, the Hoover Dam has a generating capacity of 2,080 megawatts, serving cities and regions in Arizona, Nevada and Southern California. Notable users include Los Angeles, Phoenix, Las Vegas and the Metropolitan Water District of Southern California.
Of course, as a result of the ongoing drought, neither Lake Powell nor Lake Mead are full. As the water level drops, the water pressure that drives the turbine generators located near the base of the dams falls, and power production is reduced. Hoover Dam power generation has dropped to about 1,304 MW, while the Glen Canyon Dam is down to about 800 MW. The decline in power output forces distributors to buy more expensive power to meet demand, driving up the price of electricity for customers.
The figure below shows the current state of Lake Powell and the Glen Canyon Dam. The figure introduces two new terms: Minimum Power Pool, the lowest water level that will still run the turbine generators; and Dead Pool, the level below which the dam is no longer able to release water downstream and the Colorado River would stop flowing. Theoretically, dead pool would happen if water levels fell below 3,370 feet for Lake Powell or 895 feet for Lake Mead (as shown below).
Credit: Arizona Department of Water Resources/USBR
Credit: Arizona Department of Water Resources/USBR
The situation for the Hoover Dam is more complex. While the Minimum Power Pool is shown above at 950 ft, in reality the critical point for power generation is 1,035 ft — perilously close to the USBR prediction of the range of 1,050 to 1,075 ft by the end of 2025. At 1,035 ft, 12 older generators would have to be shut down, leaving only five newer generators capable of operating down to 950 ft. Consequently, at 1,035 ft, generating capacity will drop precipitously from 1,304 MW to 382 MW.
At the start of 2026, USBR estimates Lake Powell’s elevation will be 3,538 ft—just 48 ft above minimum power pool. At this writing, Lake Mead is at 1,060 ft—only 25 ft above the minimum power pool for its older turbines.
We are coming down to the wire…
Current Status of the Drought
The general consensus is that the Western US has been in the current megadrought since 2000. Now in its third decade, this megadrought is the worst in 1,200 years. This article focuses on the Colorado River Basin—a self-contained, heavily monitored part of the overall Western drought, and something of a “canary in the coal mine” for the future of the drought.
In late August, 2025, the National Integrated Drought Information System (NIDIS) reported that one hundred percent of the Colorado River Basin was in drought, including large areas of Extreme or Exceptional Drought in the upper basin, the primary source of the basin’s water supply.
In the span of a week in October 2025, US Southwest experienced an unusual period of heavy rainfall. Prolonged aridification of the Southwest had left large areas of the basin unable to absorb much of the intense rainfall, leading to widespread flooding.
The historic event set new local rainfall records (in the middle of a drought!) Phoenix, Arizona reported the wettest meteorological fall season (September through November) on record. For the first time in over 45 years, the mountains of Southwest Colorado received over 10 inches of rain in just 7 days, causing major river flooding.
This unusual “black swan” event is characteristic of a warmer atmosphere, as we discussed just two months earlier (2025’s Flash Flood Summer). As Prof. Andrew Dessler (Atmospheric Science, Texas A&M) puts it “…[climate warming] doesn’t cause rain events. Rather, the role of climate change is like steroids for the weather — it injects an extra dose of intensity into existing weather patterns.” 2023, 2024, and 2025 will be the three warmest years on record, so the October event is a not unexpected rogue rainfall event in a hotter, drier desert Southwest.
Nonetheless, the October rain event didn’t mark the end of the drought in the Colorado River Basin, as shown below.
Despite the extreme rainfall in October, near the end of 2025 79% of the basin is in drought. Lake Mead is reported to be at 33% of full pool, while Lake Powell is at 27%.
The following graph captures the cumulative impact of the megadrought on Lake Mead and Lake Powell.
Lake Powell Status
- Lake Powell peaked at only 5 feet above its winter level in 2025 — far below the typical 40–60 ft rise from snowmelt
- June inflows were only 44% of average
- Current elevation (December 11,2025): 3,542.06 ft (158 ft below full pool).
- Capacity: 27% — down about 32 ft from a year ago
Forecasts indicate continued decline into 2026, triggering shortage plans and impacting infrastructure.
Lake Mead Status
- Elevation: Around 1,063 feet in April 2025, dipping slightly lower by October/November to 1060.57 by December 12, still significantly below full capacity of 1,229 ft.
- Capacity: Approximately 33% full.
- Shortage: Remains in a Tier 1 Shortage Condition.
Under the USBR Shortage Condition, Arizona’s annual water delivery was cut by 512,000 acre-feet, while Mexico was required to forego 80,000 acre-feet (5%) of its Colorado River allocation. For the fifth year in a row, Nevada was required to forego 7% of its allocation.
As the following satellite images show, the ongoing megadrought has wreaked havoc on Lake Mead.
Lake Mead in 2000 – the start of the megadrought. The Colorado River enters the image at the right and exits Lake Mead at the Hoover Dam in the lower left. Credit: NASA
Lake Mead in 2022 – just the third year of the megadrought. Credit: NASA
What’s Going On?
Drought in the US Southwest is often linked to large scale meteorological conditions in the North Pacific—specifically a natural climate cycle called the Pacific Decadal Oscillation (PDO). Recent research has added anthropogenic climate change to the list of causal factors.
As its name implies, the PDO is a long-term climate pattern with cycles averaging 20 to 30 years. Much like the better known El Niño/Southern Oscillation (ENSO), the PDO cycles between warm and cold phases. However, PDO cycles are far longer than ENSO events. During a cold or “negative”PDO phase, the western Pacific is warm and the eastern Pacific, along the US west coast, is colder. During a warm or “positive” phase, the western Pacific is cool and the water along the eastern edge of the North Pacific is warmer.
It’s the negative phase that brings the drought – stifling the winter snowpack that is the source of most of the water in the Colorado River Basin. The cold air over cold water along the west coast of the US holds less moisture than warm air, reducing potential precipitation. The cool phase of the PDO also can also push storms further north that would otherwise have brought precipitation to the Southwestern US.
Human Impact
In the past, the PDO was driven by natural forces, such as the heat exchanges between the ocean and the atmosphere. Paleoclimate research found that 6,000 years ago, during the last period of northern hemisphere warming comparable to today, the Pacific Decadal Oscillation was forced out of rhythm, locking in a drought that lasted for thousands of years. Now, as the world warms under the effects of climate change, it appears to be happening again.
The graph below shows annual changes in the PDO since 1854. The numerical scale on the vertical axis indicates the strength of the PDO phase, whether positive (red) or negative (blue).
Research published in 2025 by Klaven et al in Nature uncovered a link between human activity and changes in the PDO. Their climate modelling showed that between 1870 and 1950, changes in the PDO were consistent with natural processes, with less than 1% of the variability attributed to external factors, such as air pollution. However, after the 1950s greenhouse gas and aerosol emissions accounted for 53% of the variations in the PDO.
From the 1950s through the 1980s, increasing aerosol emissions from rapid industrialization following World War II drove a positive trend in the PDO, making the Southwest wetter and less drought-prone. After the 1980s, Klaven and colleagues found that the combination of the sharp rise in greenhouse gas emissions from industries, power plants and vehicles and the reduction in aerosols as countries cleaned up their air pollution shifted the PDO toward the predominantly negative, drought-prone phase that continues today. Zooming in on the PDO data for the last 30 years makes clear the decades-long shift to the negative PDO phase that is powering the Southwest US megadrought.
Note that even the very strong El Niñ0 in 2015 barely shifted the PDO into the positive range, and the PDO rapidly switched back to the negative phase that has dominated the Southwest US climate for decades. In July 2025, the PDO hit the lowest negative value ever recorded. Klaven et al point out that the prolonged negative phase of the PDO generated colder air, which we know holds less moisture than warmer air, and led to declining precipitation across the US West.
The PDO set the stage for the megadrought, but a recent study by UCLA, NOAA and CIRES scientists shows that anthropogenic climate warming transformed an ordinary drought into the extraordinary drought that ravaged the Western US in 2020-2022.
The scientists determined that evaporative demand by a thirsty atmosphere played a bigger role than reduced precipitation in droughts since 2000. During the 2020-2022 drought, evaporation accounted for 61% of the drought’s severity, while reduced precipitation accounted for only 39%. The team speculated that we may have “…entered a new paradigm where rising temperatures are leading to intense droughts with precipitation as a secondary factor.”
Which takes us back to the map we introduced at the start of this post…
Many people still expect the Colorado River to bounce back, but our findings suggest it may not. Water managers need to start planning for the possibility that this drought isn’t just a rough patch — it could be the new reality.”
Prof. Timothy Shanahan, University of Texas at Austin