Skip to main content
Computing, Environment and Life Sciences

Environment and Climate

AI approaches will help improve our understanding and advance scientific discovery in Earth and environmental sciences.
(Image by Shutterstock / Dee Browning.)

Environmental and climate sciences address some of the world’s most pressing challenges — from use of our natural resources to protection of our infrastructure and environment. In recent decades, climate and environment observation capabilities have been revolutionized, employing a suite of novel sensor, analytics and telecommunication technologies. The United States now has access to several hundred petabytes of real-time observational data of the Earth system and predictive modeling capabilities have advanced significantly to simulate the complex Earth systems, facilitated by high-performance computing.

Together, these vast observation and simulation data offer unique opportunities to apply AI approaches for improved understanding and scientific discovery in Earth and environmental sciences. AI methods offer the promise to accelerate development of advanced tools, fast emulators, scalable simulations, ensembles, and the next generation of technology for assimilating observations and data-driven forecasting.

Related Project

Development of Emulators: Process Emulators

Argonne is employing deep neural networks to replace computationally expensive parameterizations of certain physical schemes in the Weather Research and Forecasting model.

Related Project

Development of Emulators: Model Emulators

Conditional, super-resolution application is bridging the gap between coarse resolution and convective-permitting scale in earth system models using deep learning.