Ph.D. candidate in the Department of Electrical and Computer Engineering
What to Expect
The Earth is a complex system with many physical processes interacting across space and time. Understanding these interactions in a causal perspective, i.e., as cause and effect, can help us get a deeper understanding of the mechanisms governing the Earth’s climate. This presentation focuses on the utility of observational studies, which do not rely on controlled experiments, to gain insights into these interactions. These approaches enable us to derive data-driven hypotheses of the key interactions between climate variables and the time scales at which they occur. This talk will demonstrate how methods based on probabilistic graphical models can be used to gain insights into different questions in climate science, as well as the limitations and challenges associated with them. The applications discussed will include the interactions between Planetary and Synoptic scale atmospheric disturbances, as well as the interactions between the Arctic temperature and midlatitude circulations.
Savini Samarasinghe is a Ph.D. candidate in the Department of Electrical and Computer Engineering at Colorado State University. Her research focuses on climate informatics, specifically the use of data-driven approaches to infer potential cause-effect relationships in climate. Her research interests lie broadly in using data science and machine learning to tackle questions in the Earth sciences. Savini’s work as a graduate student has been recognized by several awards, including an outstanding student presentation award at the 2019 AGU Fall meeting for her work on using causal discovery methods to explore MJO teleconnections in a changing climate. She received her undergraduate degree in Electronics and Telecommunication Engineering from the University of Moratuwa in Sri Lanka.