Novel changepoint methodology, and their application to environmental data.
Conference
Category: The International Environmetrics Society (TIES)
Abstract
With the evolving climate due to the continuing emission of greenhouse gasses and effects from various policy decisions, changes within observed environmental data are inevitable. Methods for detecting changepoints, which are abrupt shifts in the underlying distribution of a statistical process, continue to be key analysis tools. Changepoint detection is crucial across a wide variety of environmental inferences, from millenium-scale data in ice cores to sub-daily fluctuations in air pollution. These different scales and data structures motivate novel changepoint detection methods. This session highlights some recent advances in changepoint detection and their uses in a variety of environmental areas. Specifically, the session presents 1) a model that detects changes in the distribution of phytoplankton taxa, 2) a study on the changes of tropical storm counts in the North Atlantic Basin, and 3) methods to detect changes in air temperatures.