Research / Theme
Hydroclimate modeling, downscaling, and Earth system predictability
Illustration
Overview
Understanding and predicting hydroclimate extremes depends on tools that represent atmospheric and hydrologic processes in enough detail to capture individual events. We build and evaluate such tools, from kilometer-scale atmospheric simulations to machine-learning frameworks, and use them to push the limits of what can be predicted.
Research directions
Representative findings
Machine-learning clustering identifies distinct weather and moisture conditions associated with wildfires in the western U.S., improving seasonal predictions by 10%. Historically, some wildfire types have grown more common while those associated with wetter soils have declined, revealing changing patterns of wildfire risk.
Warmer coastal ocean temperatures bring wetter, warmer winters to the U.S. West Coast, reducing snowpack in the northern Cascades while increasing it in the southern Sierra Nevada. Changes in atmospheric rivers and precipitation interact with elevation to produce these contrasting responses.
Warmer coastal ocean temperatures strengthen landfalling atmospheric rivers along the U.S. West Coast, increasing their frequency, intensity, and spatial extent. Numerical experiments show that AR-related extreme precipitation increases by approximately 3% per degree of ocean warming, highlighting the influence of local sea surface temperatures on...
Atmospheric river occurrence, intensity, and duration help predict extreme precipitation in western U.S. watersheds at daily and monthly timescales. Machine-learning analysis highlights the importance of intense, long-lasting events, with stronger atmospheric rivers generally associated with heavier precipitation.
See DOE highlight
Projects
Evaluating high-resolution simulations of mesoscale convective systems
Our role Lead
Project detailsAdvancing credible and actionable regional climate information
Our role co-Investigator
Partners LBNL · PNNL
Project detailsPublications
AI agent for hydrologic modeling: definition, development, and application
Object-based Evaluation of Dynamical and Statistical Downscaled Precipitation Products over CONUS
High-resolution projections of extreme heat and thermal stress in southeastern U.S.
Potential weakening of the June 2012 North American derecho under future warming
