California is lucky to host some of the most innovative universities in the world. More specifically, our surrounding counties boast various institutions prided on their commitment to diversity, progress, and achievement. The University of California: Santa Cruz (UCSC) is one of these institutions, serving as a model innovator of climate solutions and employing cutting-edge technology that is changing the landscape of sustainability science as we know it. In September 2024, two UCSC professors were awarded nearly $2 million in funding for their projects centered around climate change. This funding, generously allotted by the U.S. National Science Foundation’s Collaborations in Artificial Intelligence and Geosciences (CAIG), is meant to aid geoscientists in advancing AI technology to increase our understanding of Earth.

Chattopadhyay: Avoiding AI “Hallucinations”

Ashesh Kumar Chattopadhyay is an associate professor at UCSC affiliated with the Baskin School of Engineering in Applied Mathematics. Having received his PhD in Mechanical Engineering from Rice University, Chattopadhyay is well-versed in scientific machine learning, deep learning theory, and dynamical systems. The project he is leading, which is receiving funding from CAIG, involves AI modeling for weather and climate. However, his project focuses on a significant limitation to previous AI models of the same variety.

AI weather and climate predictors help tremendously in forecasting climate-change-related natural disasters. Geoscientists can glean early warning signs of potential floods and heatwaves from the models generated by AI technology, allowing at-risk localities to evacuate affected areas before disaster strikes. These insights can also illuminate climate patterns that may emerge in the future, allowing scientists, sustainability experts, and policymakers to prepare their communities and instate emergency plans.

However, one glaring problem with this technology is its inability to root its predictions in reality. When looking 15-20 days in the future, the AI begins to “hallucinate” scientifically implausible models that are misleading and, in worst cases, outright wrong. Chattopadhyay and fellow associate professor Nicole Feldl seek to amend this issue by building consistent AI emulators that can reliably forecast weather and climate at a reduced price. The $898,000 award will enable Chattopadhyay and Feldl to expand their research and create realistic, cost-effective solutions to global climate challenges.

Prochaska: Addressing Warming Oceans

Jason Xavier Prochaska is a professor at UCSC in the Astronomy and Astrophysics Department. He was named a Distinguished Professor in 2020, and before that, he received an Outstanding Faculty award in 2017, which speaks volumes of his commitment to furthering education and research in the physical and biological sciences division. Originally hailing from New England, Prochaska received his PhD in Physics at the University of California, San Diego, and completed a postdoctoral fellowship at Carnegie Observatories before joining the UCSC faculty. In line with his passion at the intersection of AI and oceanographic research, his project will utilize AI to detect and measure ocean heat storage.

Prochaska and his research team will develop an AI to measure ocean fronts using satellite data. Ocean fronts are boundaries between water masses of different temperatures, densities, and salinities, and they play an essential role in regulating the climate and shaping ocean currents. Scientists can more accurately predict climate change or potential natural disasters that threaten coastal communities by investigating ocean fronts and how they impact weather systems.

Like Chattopadhyay’s project, Prochaska also seeks to mitigate AI’s uncertainty when predicting future models. The team plans to pair the AI model with over 15 years of satellite data to enhance its predictive capabilities and improve the accuracy of its measurements. That way, Prochaska’s AI can get the most conclusive results to explore how the heat content in the upper half of the ocean fuels the climate crisis and vice versa. This project’s $1 million award will contribute to the expanding body of oceanographic climate research and the refinement of AI technologies to collect accurate and trustworthy data.

Learn More!

Professors like Chattopadhyay and Prochaska are paving the way for innovative climate solutions by employing AI. If you are interested in projects similar to those at UCSC, take a look at the U.S. National Science Foundation’s official announcement to invest $20 million to advance AI in the field of geoscience. Universities from all over the nation are trying to push the boundaries of emerging technologies to solve the most pressing environmental concerns plaguing Earth. The efforts of these awardees reflect a promising future where AI safeguards the health of our planet, bolstered by the brilliant minds of researchers and educators who seek to eliminate the climate crisis once and for all.