Machine Learning Based Diagnostics of Climate Variability from Subseasonal to Seasonal Timescales

Explore machine learning applications for diagnosing climate variability patterns at subseasonal timescales in this conference talk delivered at the International Centre for Theoretical Sciences. Learn how advanced ML techniques can be applied to analyze and understand complex climate dynamics that occur over periods ranging from weeks to months. Discover methodologies for extracting meaningful climate signals from large-scale atmospheric and oceanic datasets using data-driven approaches. Examine the potential of machine learning algorithms to identify patterns in climate variability that traditional statistical methods might miss, particularly focusing on subseasonal phenomena that bridge the gap between weather prediction and seasonal forecasting. Understand how these diagnostic tools can enhance our comprehension of climate system behavior and improve predictive capabilities for intermediate-range climate forecasting. The presentation is part of the Advanced Machine Learning for Earth System Modeling program, which explores cutting-edge applications of artificial intelligence in climate science and Earth system research.