Viewing forced climate patterns through an AI Lens in Geophysical Research Letters


Preliminary Map of Neural Network Weight Intensity

Pattern Exploration recently contributed to an interdisciplinary publication, using machine learning techniques to separate climate ‘signals’ from internal noise across climate models. and isolate climate change patterns.
Viewing forced climate patterns through an AI Lens, published in Geophysical Research Letters

Interpretable analysis of neural networks is one of the foundational values of Pattern Exploration, and is an important aspect of deep analysis of complex systems and models.

A recent article that includes more on the interdisciplinary cooperation at Colorado State University that went into this publication can be found here https://engr.source.colostate.edu/atmospheric-science-meet-data-science/

Simple classification neural network across multiple global climate models

Code repository for this paper and project is publicly available through