AI for Tipping Point Discovery

AI experts at APL and oceanographers from the Johns Hopkins University Department of Earth and Planetary Sciences leveraged AI and machine learning to better model the complex processes that affect Earth systems.

This work was supported by the Defense Advanced Research Projects Agency (DARPA).

To understand how tipping points can trigger cascading responses, APL researchers are developing an AI scientific discovery assistant named Causal-Cascading AI Tipping-Point Neuro-symbolic Intervention and Prediction (CATNIP).

CATNIP includes a generative adversarial network (TIP-GAN) that learns to predict what could tip the Atlantic Meridional Overturning Circulation (AMOC), an important system of ocean currents that circulates warm water from the South Atlantic to the colder North Atlantic. If the AMOC were to slow or collapse, there could be consequences for food security, sea levels, delicate ecosystems, and the Arctic.

The program was the first demonstration that AI can be used as an Earth systems modeling assistant, where researchers can run experiments by asking the AI modeling assistant natural language questions, similar to using a large language model. CATNIP extends this capability to examine the tipping points of coral reefs, sea-ice melt, and sea-level rise so it can study how tipping points are interconnected and interventions that could mitigate individual and cascading tips.