Drought and the Power Grid: How Hydrology Becomes a Forecasting Variable

Drought rarely shows up on a single operating day the way a heat wave or a winter storm does, but it shapes the operating environment for months at a time. A dry winter sets the available hydro fleet for the following summer. A multi-year drought tightens cooling-water limits at thermal plants. A dry, windy spring sets the wildfire risk that drives summer transmission de-rates and Public Safety Power Shutoffs. This article walks through how hydrology becomes a forecasting variable for power markets, and how the indicators on Weather Workbench translate into grid consequences.

The U.S. Drought Monitor, published weekly by the National Drought Mitigation Center in partnership with NOAA and USDA, is the standard national snapshot of drought conditions. The categories run from D0 (abnormally dry) through D4 (exceptional drought). Each map is built by a rotating team of authors from federal and academic institutions, using a blended methodology that combines short-term indicators (the Standardized Precipitation Index over recent months, soil moisture anomalies, vegetation health), long-term indicators (multi-year SPI, groundwater levels, reservoir storage), and ground-truth reports from state climatologists, agricultural extension agents, and water-resource agencies.

Two features of the Drought Monitor matter for power-market use. First, it is intentionally slow to change. A region in D3 does not drop to D1 because of a single rain event, and a region in D0 does not jump to D3 because of a single dry month. The categories are designed to track the cumulative water balance, which is what matters for reservoir storage, snowpack, and groundwater. Second, the same category can mean operationally different things in different regions. D2 in the Pacific Northwest, where the grid leans hard on hydro, is a much bigger deal than D2 in PJM, where hydro is a small share of the resource mix.

Hydro is the most direct connection between drought and grid output. The hydro fleet across the western U.S. — Bonneville Power Administration's Federal Columbia River Power System in the Pacific Northwest, the Bureau of Reclamation's Hoover and Glen Canyon dams on the Colorado, California's State Water Project pumping and generation, and the smaller dam fleets across CAISO and SPP — ultimately depends on precipitation that fell in the preceding water year, often months or seasons earlier.

The seasonal cycle matters. In the Columbia Basin, most of the year's runoff arrives as snowmelt between April and July. By April 1, the snow water equivalent measurements at the network of SNOTEL stations across the basin are highly predictive of the summer hydro fleet's energy production. A snowpack at 70 percent of normal on April 1 will, with high confidence, deliver less hydro generation between June and September than a 110 percent snowpack would. BPA's seasonal fish-passage and flood-control obligations further constrain how the available water can be used. The hydro forecast for the Pacific Northwest summer is largely written by April 1.

On the Colorado, the situation is different. Lake Powell and Lake Mead store multiple years of runoff, so single-year deficits do not immediately show up as generation losses. But a multi-year drought eventually drives reservoir elevations toward the minimum power pool, the elevation below which intake structures cannot reliably move water through the turbines. Glen Canyon and Hoover have both spent extended periods over the past two decades with reduced generation capability because reservoir elevations sat near minimum power pool. When reservoir elevations are projected by the Bureau of Reclamation's 24-month study to approach those thresholds, the western interconnection's summer resource adequacy outlook tightens accordingly.

California operates somewhere between these two cases. The State Water Project and Central Valley Project have multi-year storage capacity, but the smaller hydro plants scattered across the Sierra Nevada are essentially run-of-river with limited storage and respond directly to current-year precipitation. A dry winter cuts late-summer hydro contribution materially. CAISO's summer assessments reflect this directly: a low snowpack year is also a tight summer reliability year.

Thermal generators — coal, gas-fired combined cycle, nuclear — depend on cooling water for plant operation. Plants on rivers, lakes, and estuaries use either once-through cooling (water drawn from the source, passed through condensers, and returned warmer) or closed-loop cooling with cooling towers (water cycled through a tower and partly evaporated, with periodic blowdown back to the source). Both depend on the source body having enough water at low enough temperature.

Drought tightens both constraints. River and lake levels drop, which reduces intake capacity. Source-water temperatures rise as flows shrink and warm air heats the remaining water faster. Discharge permits typically cap how much temperature increase a plant can add to the receiving water and what the maximum return temperature can be. As source temperatures rise toward the permit limit, the plant must reduce output to comply.

These derates are most familiar at midwestern and southeastern coal and combined-cycle plants drawing from the Mississippi, Ohio, Tennessee, and Missouri river systems. They affect ERCOT lake-cooled plants in a multi-year drought. They occasionally affect ISO-NE coastal plants when offshore tropical-water intrusions raise estuary temperatures. Nuclear plants face additional layers of regulation, and plant-specific waivers from the NRC during extended heat events are not unprecedented.

The forecasting takeaway is that the same hot afternoon that drives cooling load also stresses thermal plant output through cooling-water limits. The grid loses capacity at exactly the hours it needs it most. This is one of the structural reasons why summer scarcity events tend to be worst on multi-day heat waves rather than on single hot days — the cooling-water source has time to warm up across consecutive days of high air temperatures and reduced flow.

Drought is the upstream driver of the wildfire risk that has reshaped western transmission operations over the past decade. Dry fuels, low overnight humidity recoveries, and elevated daytime winds combine to push the National Weather Service Fire Weather products — Red Flag Warnings, Fire Weather Watches, and the Storm Prediction Center's fire-weather outlooks — into elevated and critical categories. Each of those categories has direct operating consequences in the West.

Public Safety Power Shutoffs, pioneered by California utilities and now used by several western utilities, de-energize transmission and distribution circuits during the most dangerous fire-weather conditions to prevent ignition from line failures. PSPS events are forecast against a combination of red-flag conditions, fuel moisture observations, and engineering judgment. They directly remove load and sometimes generation from the system, and they shift load profiles in unpredictable ways.

Fires that do start can de-rate or take out transmission lines, either through direct damage or through smoke-induced flashovers. Smoke from large fires dramatically reduces the dielectric strength of the air around energized conductors; on bad days, transmission operators reduce line ratings or take lines out of service preemptively. The Pacific Northwest, California, and the desert Southwest have all experienced summers where wildfire smoke alone — without any direct line damage — drove material reductions in transfer capability across major intertie paths.

All of this is downstream of drought. A wet winter does not eliminate fire-weather days, but it materially reduces the number of critical-category days, the duration of the fire season, and the size of the fuel base available to burn. The Drought Monitor map at the start of fire season is the single best leading indicator of the operational fire risk a western ISO will face in the months that follow.

Drought-driven grid risk shows up on a different cadence from weather-driven grid risk. The relevant indicators do not change much from one day to the next. They are best monitored on a weekly-to-monthly cadence, and the questions you ask of them are seasonal rather than hourly.

Snow water equivalent at SNOTEL sites tells you, by April 1, what the western summer hydro fleet will look like. Reservoir elevations from the Bureau of Reclamation's 24-month studies tell you whether multi-year storage is approaching minimum power pool. The Drought Monitor's category map tells you the spatial footprint of stress. The Climate Prediction Center's 30-day and 90-day outlooks tell you whether the trend is improving or worsening. The NWS Fire Weather Outlook from the Storm Prediction Center tells you, day by day, where critical fire-weather conditions are developing.

The drought monitor panel surfaces the weekly Drought Monitor map for the U.S. and for each ISO footprint. The CPC 6–10 and 8–14 day outlooks, together with the 30-day and 90-day outlooks, indicate whether building cold or warm patterns will compound or relieve current drought conditions. Active NWS alerts surface Red Flag Warnings, Fire Weather Watches, and Excessive Heat Warnings — the latter are particularly relevant when source-water temperatures are already elevated. Watching these together gives you a credible read on whether drought is currently a marginal background condition or an operational variable that should be near the top of the watch list.

Concepts and data sources discussed in this article are drawn from the public-use federal and ISO materials listed below.