There is a growing consensus among scientists that aspects of our planet's climate are changing due to human influences, though the scientific community acknowledges that substantial uncertainty exists regarding the forms, levels, and impacts of change. Quantifying these uncertainties requires new statistical research informed by climate science. Effective solutions to climate-change problems will rely on new methods for combining the information content of models and data in a fashion that quantitatively manages uncertainty. The research team of statisticians and climate modeling experts from the National Center for Atmospheric Research is developing new statistical strategies that combine observations with the information present in computer models for the climate system, while managing the uncertainties implicit in both. The research team will rely extensively on Bayesian hierarchical modeling and analysis strategies. Specific projects include (1) developing new probabilistic climate-change assessments based on an extensive suite of climate simulations; (2) statistical procedures for combining different climate models to produce climate projections; and (3) assessing regional and local impacts of global climate behavior.

• NSF Division of Mathematical Sciences