Atmospheric River Program

Atmospheric rivers (ARs), elongated transports of water vapor, are important for water supply in California, however, the most extreme ARs can lead to flooding. Better monitoring and prediction of ARs has the potential to enhance use of existing reservoirs as to increase water storage while maintaining flood control capabilities. Implementation of existing knowledge and research to extend that knowledge and develop tools could mitigate risks of too much or too little precipitation and could aid in adaptation to changes in climate. The AR Program aims to develop the science of atmospheric rivers to support planning, forecasts and warning elements of flood management and water management in California.

The California Atmospheric River (AR) Program was authorized by California Senate Bill SB-758 and signed by Governor Brown on October 9, 2015. The AR Program is housed in the CA Department of Water Resources and is managed by California State Climatologist Michael Anderson. The first phase of the AR Program was completed in March 2019. A summary report of the AR Program Phase I achievements can be found here.

The AR Program Phase II began in November 2019. This interdisciplinary project requires innovations in meteorology, hydrology, climate science, data science, oceanography, civil engineering, water resources management, fisheries management, and decision support systems. Major observational, computational, and educational capacity, including science and engineering, are core elements of the Program. Execution of AR Program Phase II builds on the partnerships, tools, staff, and project management built during Phase I, which included close and strong leveraging of federal facilities (e.g., weather reconnaissance aircraft, super computers, and Forecast Informed Reservoir Operations, FIRO, which enabled changes to the operations guidance and training of the U.S. Army Corps of Engineers, USACE). Research efforts in the AR Program focus on further improving situational awareness of ARs and the associated precipitation to enhance the resilience of California to extreme water related events. The research efforts of the AR Program Phase II are detailed below and organized into three main strategic themes: forecasting, observations, and decision support.

For more information on ARs visit the AR FAQs or watch this informational video about ARs.


Atmospheric river airborne reconnaissance: Lead the development of a new airborne weather reconnaissance method focused on better prediction of landfalling ARs on the US West coast in the cool season. Uses national research aircraft facilities normally deployed in summer and fall for hurricanes. Leveraging significant federal investment in AR-storm aircraft-based observations (funded by NOAA and Air Force). Supports evaluation and improvement of the recon methods, especially data assimilation method development, and focuses on emergency management and FIRO applications.

Land-based observations enhancements: Improve monitoring of atmospheric rivers and related conditions over California, including enhancement of the state-wide network of AR Observatories, snow-level radars, soil moisture and water vapor sensors, as well as developing new measurements, e.g., for snow albedo and density. Provides ability to fill observing system gaps identified over time.

Observation Support for AQPI: DWR is supporting the development of a decision support system for heavy precipitation in the San Francisco Bay area called the Advanced Quantitative Precipitation Information (AQPI). It integrates multiple observing systems including new radar units with modeling and decision support tools enabling multiagency response to mitigate economic losses from extreme precipitation events. This enables CW3E participation, training, and regional support of the AQPI venture.


West-WRF weather model: Continue development of a regional weather model tailored to predicting atmospheric rivers and their associated extreme precipitation in California, especially on the space and time scales of key watersheds. Includes scientific studies of underlying physical processes, optimized assimilation of observations, and modern post-processing techniques, and development of forecast skill performance metrics focused on California and Forecast-Informed Reservoir Operations needs.

S2S outlooks including atmospheric rivers and ridges: Develop and test new methods for predicting precipitation 2 weeks to a year ahead in areas impacting California water, including all of California and the Colorado River Basin. Conduct scientific studies and develop tools focused on wet periods (ARs) and on dry periods (ridges). For tools that have skill, provide outlooks in near real-time.

Evaluate/improve atmospheric river representation in climate models: Climate models vary substantially in their projections of annual precipitation across California, primarily because of differences in how they represent atmospheric rivers. Scientific studies, potentially leveraging federal and other investments, will identify the causes of these differences, determine which models are best.

Supercomputing capacity: Supports West-WRF, S2S, AR Recon and climate studies.

Decision Support:

Forecast-Informed Reservoir Operations (FIRO): ARs often lead to excess runoff causing flooding and is lost for supply or ecosystem benefit management. This explores the potential of better AR forecasts, used with forecast-informed reservoir operations (FIRO).

Debris flow risk assessment: Develop science to improve understanding of what combination of weather and post-fire hydrologic conditions trigger debris flows. This will entail developing specialized forecasts of intense precipitation, often associated with convection (associated with atmospheric rivers and other storm types), which has been identified as a trigger for debris flows.

DWR Informational Two-Pagers: