Drought

Research, Initiatives, and Products

Adobe Stock

The Physical Sciences Laboratory (PSL) enhances drought preparedness and response through a forward-looking approach that targets the complex, cascading impacts of water shortages.

Drought occurs when an extended imbalance between water supply (precipitation) and water loss (evapotranspiration and/or human use) leads to an abnormal water deficit in critical hydrological components such as soils, streams, and reservoirs.

Drought poses a significant threat to the security and prosperity of the United States, both domestically and internationally, with far reaching consequences that include, but are not limited to:

  • decreased agricultural yields;
  • reduced water availability for consumption;
  • hindered transportation such as low waterway levels and road/rail destablization; and
  • wildfires.

These effects can make food and water insecurity worse, potentially leading to societal unrest (Kelley et al., 2026; Voice of America, 2011). The consequences are also financial. Between 1980 and 2024, drought-related economic losses in the U.S. amounted to nearly $370 billion (per National Centers for Environmental Information).

By conducting dedicated research, developing advanced predictive tools, and delivering key insights, PSL helps key community decision makers and local, state, and federal partners better anticipate and manage the effects of drought.


Initiatives and partnerships

PSL develops and improves on the evidence and delivery of early warning of drought and its compounding and cascading effects through impactful initiatives and partnerships.

Modernizing Drought Early Warning in the United States - Pilot Program

PSL leads an experimental drought resilience framework to produce sector-specific drought scenarios based on tailored monitoring and forecasting information, which will enable users to make proactive decisions ahead of drought. These sectors include water utilities, public health, and agriculture.

This pilot is being conducted in partnership with NOAA’s National Integrated Drought Information System, the California State Climatologist/California Department of Water Resources, NOAA’s National Weather Service and National Centers for Environmental Information, and the California-Nevada Adaptation Program (a NOAA CAP team)

Together the team delivers a user-oriented and evidence-based approach to drought early warning for sectors of our Nation’s economy susceptible to the hazards of too much and too little water.

Drought Early Warning pilot page (Drought.gov)

Piloting A Modern Approach to Drought Early Warning - PSL News


Famine Early Warning Systems Network (FEWS NET)

PSL leads the development of acute food insecurity scenarios based on current and evolving drought conditions in over 40 countries worldwide for the Department of State Famine Early Warning Systems Network. These scenarios inform decisions and responses in the world’s most food-insecure countries, including more accurate estimates of food aid.

FEWS NET (Dept. of State)

PSL FEWS NET page

FEWS NET monthly outlooks (Vimeo)


Experimental monitoring and forecasting products

The following experimental tools provide important data and visualizations to support the early warning of drought and its compounding and cascading effects.

Evaporative Demand Drought Index (EDDI)

EDDI quantifies the thirst of the atmosphere by measuring the demand of moisture from the land surface by the atmosphere. It is an especially useful tool for monitoring drought onset and providing early warning of conditions related to wildland fire.

Conditions Related to Large Wildland Fires in the United States

Documents historical drought conditions related to hundreds of large wildland fires in the United States that may provide insights into predicting the next large fires.

Vapor Pressure Deficit Forecasts

PSL provides experimental guidance on vapor pressure deficit (VPD) that is used to anticipate the potential for wildland fires. These forecasts are generated using a machine learning approach with a linear inverse model.

Subseasonal Precipitation Forecasts

Experimental subseasonal outlooks (weeks 1-4) showcase near real-time performance of conventional and deep-learning methods to calibrate raw precipitation accumulation forecasts from NOAA’s GEFSv12.

Monthly and Seasonal Forecasts

PSL provides experimental guidance on precipitation and temperature used to anticipate drought and its compounding and cascading effects.

ENSO-related Impacts

Documents historical conditions related to El Nino and La Nina across several global regions, which may be applied to future drought predictions.


Recent research and analysis

Weather whiplash in Texas: drought to flood
PSL researchers analyzed the extreme transition from drought to flood in central Texas in early July 2025 to improve understanding of "weather whiplash" extremes and advance early warning and planning.

Seasonal Predictability of Vapor Pressure Deficit in the western United States
As wildfire seasons grow longer, knowing what lies ahead is critical. This paper explains how vapor pressure deficit (VPD) can be forecasted months ahead of time, potentially giving fire managers a critical head start on resource planning. These forecasts are generated using a machine learning approach with a linear inverse model.

Breeden, M. L., Hoell, A., Worsnop, R. P., Albers, J. R., Hobbins, M., Robinson, R. M., and Vimont, D. J.: Seasonal Predictability of Vapor Pressure Deficit in the western United States, Weather Clim. Dynam., 6, 1443–1459, https://doi.org/10.5194/wcd-6-1443-2025, 2025.

Using Deep Learning to Identify Initial Error Sensitivity for Interpretable ENSO Forecasts
Better drought prediction starts with better ENSO forecasting. This study unveils a 'glass box' AI approach that doesn’t just predict El Niño and La Niña with high accuracy—it actually reveals the hidden ocean and wind patterns that drive them, offering a major leap forward for regional climate preparedness.

Toride, K., M. Newman, A. Hoell, A. Capotondi, J. Schlör, and D. J. Amaya, 2025: Using Deep Learning to Identify Initial Error Sensitivity for Interpretable ENSO Forecasts. Artif. Intell. Earth Syst., 4, e240045, https://doi.org/10.1175/AIES-D-24-0045.1.

Potential Predictability of Two-Year Droughts in the Missouri River Basin
Predicting a single dry summer is hard enough; predicting two in a row is even harder. This study identifies the specific 'March precursors'—including soil moisture and La Niña—that signal a high risk of 2-year droughts in the vital Missouri River Basin.

Hoell, A., X. Quan, R. Robinson, and M. Hoerling, 2024: Potential Predictability of Two-Year Droughts in the Missouri River Basin. J. Climate, 37, 3413–3432,https://doi.org/10.1175/JCLI-D-23-0588.1.

Flash drought: A state of the science review
A 'Flash drought' involves rapid onset, no warning, and severe impacts. This state-of-the-science review unifies the science behind these fast-moving climate events that have a significant effect on water security.

Christian, J. I., Hobbins, M., Hoell, A., Otkin, J. A., Ford, T. W., Cravens, A. E., Powlen, K. A., Wang, H., & Mishra, V. (2024). Flash drought: A state of the science review. WIREs Water, 11(3), e1714. https://doi.org/10.1002/wat2.1714

Rapid Development of Systematic ENSO-Related Seasonal Forecast Errors
El Niño models contain immediate, critical forecasting errors in the tropical Pacific, driven by the atmosphere rather than the ocean. This "westward shift" in simulated temperatures is a key factor limiting accurate worldwide seasonal weather prediction.

Beverley, J. D., Newman, M., & Hoell, A. (2023). Rapid development of systematic ENSO-related seasonal forecast errors. Geophysical Research Letters, 50, e2022GL102249. https://doi.org/10.1029/2022GL102249

More drought-related PSL publications (by title)

Page Last Updated: February 13, 2026