Observation-Based Simulations of Future Temperatures with Changes in Variability Predicted by a Climate Model

• Activity Number: 91
• Type: Invited
• Date/Time: Sunday, August 9, 2015 : 9:30 PM to 10:15 PM
• Sponsor: Section on Statistics and the Environment
• Author(s): Andrew Poppick* and Michael L. Stein and Elisabeth Moyer
• Companies: The University of Chicago and The University of Chicago and The University of Chicago
• Keywords: climate change ; climate variability ; data-driven simulation ; evolutionary spectrum ; nonstationary processes

Abstract

Future climate change impacts depend on changes in the temporal covariance structure of future temperatures in addition to changes in mean temperature. While general circulation models (GCMs) predict changes in both mean and variability, GCMs cannot fully reproduce the distribution of observed temperatures. Our work addresses an ensuing need for simulations of future temperatures that combine properties of both the observational record and a GCM's projections. We advocate that such simulations be based on transformations of observational data that reflect GCM projected changes, in contrast to methods that transform GCM realizations in an attempt to resolve GCM/observation discrepancies. Leeds et al. (2014) described a method for transforming observed temperatures based on GCM predicted changes in means and covariances between two equilibrium (stationary) climates. Since our climate is instead in a transient (nonstationary) state, we build on that methodology for transient climates, describing a model for a GCM's projected changes in covariance structure across an ensemble of future emissions scenarios.