The Colorado River system’s combined reservoir storage has dropped to its lowest point since 1957, according to daily records maintained by the U.S. Bureau of Reclamation. Lake Mead and Lake Powell, the two largest reservoirs on the river, are both well below average capacity, putting pressure on hydropower generation, municipal water supplies, and irrigation deliveries across seven western states. The decline raises hard questions about whether existing shortage frameworks can keep pace with the rate of loss.
Storage declines and the shortage-tier clock
The speed of the drawdown is the central problem. Reclamation publishes Lake Mead storage data in a time-series that stretches back to 1937, making it one of the longest continuous hydrological records in the West. That dataset now shows storage figures not seen in nearly seven decades. Lake Powell storage, reported in thousand acre-feet and percent of capacity on the Bureau’s Lower Colorado operations page, tells a similar story for the Upper Basin.
Coordinated operations between the two reservoirs are governed by the 2007 Interim Guidelines, the 2019 Lower Basin Drought Contingency Plan, and the binational agreement known as Minute 323, all referenced in a Reclamation news release on the August 2023 24-Month Study. Those instruments set elevation thresholds that trigger progressively deeper cuts to water deliveries. When storage falls faster than the 24-Month Study’s modeling anticipated, the gap between projected and actual conditions compresses the timeline for reaching the next shortage tier. Daily tracking through Reclamation’s RISE platform, which pulls from regional Hydromet and HDB systems, allows outside analysts to compare real drawdown rates against the study’s baseline scenario in near-real time.
The practical effect for downstream users is straightforward: if storage crosses a threshold weeks or a full month ahead of schedule, states and water districts that planned around the study’s projected timeline face earlier and deeper mandatory reductions than their budgets assumed.
Reclamation’s daily datasets confirm the system-wide drop
Three primary data products anchor the record-low claim. First, the Lake Mead time-series dataset, cataloged through Reclamation’s RISE system, provides daily acre-feet readings updated from regional water operations databases. Second, the Lower Colorado teacup diagram offers a provisional, automated snapshot of reservoir elevations, contents, and flows across the system. Third, the Reservoir Storage Dashboard hosted on Drought.gov draws on the same Reclamation data to display current storage as a percent of average and to flag records for lowest observed storage at selected reservoirs.
The dataset underlying the dashboard is updated daily, according to its RISE catalog record, and is sourced from Reclamation’s own regional operations databases. Upper Basin reservoir data and projection products are accessible through Reclamation’s Upper Colorado water operations gateway, which includes modeling outputs that feed into the 24-Month Study. Reclamation’s general modeling information page explains the variables and hydrology uncertainty built into those projections, a detail that matters when comparing modeled trajectories to observed storage.
Taken together, these sources confirm that the system-wide decline is not an artifact of a single reservoir’s seasonal fluctuation. Both the Upper and Lower Basin anchors of the Colorado River system are drawing down simultaneously, a pattern that distinguishes the current period from earlier dry spells when one reservoir could partially compensate for the other.
Missing data and the next decision point for seven states
Several gaps limit how precisely anyone can pin down what comes next. The exact combined acre-feet total for all listed reservoirs on any given day requires pulling and summing individual readings from multiple catalog entries; no single Reclamation publication rolls them into one headline number with a date stamp. Real-time Hydromet readings beyond what the dashboard links provide are not included in the publicly summarized datasets, and basin-specific inflow forecasts from NOAA have not been paired with the storage data in a way that yields a firm crossover date for the next shortage tier.
Recreation and access impacts at Lake Mead and Lake Powell are another blind spot. The recreation.gov portal, linked from Reclamation’s operations pages, does not currently surface visitor data tied to low water levels, leaving the economic toll on marina operators and tourism-dependent communities difficult to quantify from federal sources alone.
The most consequential thing to watch is whether Reclamation’s next 24-Month Study update revises its projected shortage-tier timeline to reflect the pace of daily declines visible in the RISE data. If the study’s baseline scenario still assumes a slower drawdown than what the daily records show, water managers in Arizona, Nevada, California, and Mexico will need to plan for cuts that arrive sooner than the official calendar suggests. For residents and businesses that depend on Colorado River water, the gap between modeled and observed storage is not an abstraction. It determines when and how deeply their allocations get cut.
More from Morning Overview
*This article was researched with the help of AI, with human editors creating the final content.
