Interpretable Machine Learning Approaches for Multi-Year Climate Variability

Explore innovative machine learning approaches for identifying and understanding predictable multi-year climate variability in this 52-minute Allen School Colloquia Series talk by Emily Gordon from Stanford University. Discover how neural networks and explainable AI techniques can be applied to examine Pacific decadal variability, investigate predictable sea surface temperatures across oceans, and assess the impact of climate change on near-term climate variability. Learn about the creative experimental design that combines neural networks' ability to predict non-linear behavior with insights into sources of predictable internal variability. Gain valuable knowledge on how AI-driven SST predictions can provide constrained estimates of future climate variability in surface temperatures and precipitation, and explore the potential of data-driven tools in shaping future investigations of climate variability and predictability.

Interpretable ML approaches for multi-year climate variability: Emily Gordon (Stanford University)