Penn State College Secures $1.23 Million NASA Grant to Revolutionize AI-Driven Satellite Weather Forecasting

Computer vision represents a subset of artificial intelligence that leverages machine learning and neural networks to enable computers to comprehend and analyze visual data and to enhance their capabilities through learning from that information.

The research group—comprising Steven Greybush, associate professor of meteorology at the Penn State College of Earth and Mineral Sciences, along with experts from Argonne National Laboratory, NASA Goddard Space Flight Center, the National Oceanic and Atmospheric Administration, and the University of Chicago—intends to integrate multiple data sources, such as satellite imagery, to enhance previous weather predictions that utilized transformer-based AI algorithms and machine learning frameworks.

“Generally, atmospheric and oceanic forecasts necessitate data assimilation—merging diverse information sources regarding weather to achieve a more precise outcome,” noted Romit Maulik, assistant professor in the College of IST. “Nonetheless, this data assimilation can considerably delay the forecasting process. We aim to employ computer vision to significantly expedite this procedure.”

“The initiative will involve retraining certain components of our model to accept these new data sets as inputs and enhance predictions,” stated Maulik. “After that, we will incorporate these refined algorithms into the NASA Goddard Earth Observing System to swiftly integrate satellite system observations into its operational data assimilation processes.”

Maulik and Greybush are also joint hires at the Penn State Institute for Computational and Data Sciences.