Translator Disclaimer
February 2020 Some Statistical Issues in Climate Science
Michael L. Stein
Statist. Sci. 35(1): 31-41 (February 2020). DOI: 10.1214/19-STS730

Abstract

Climate science is a field that is arguably both data-rich and data-poor. Data rich in that huge and quickly increasing amounts of data about the state of the climate are collected every day. Data poor in that important aspects of the climate are still undersampled, such as the deep oceans and some characteristics of the upper atmosphere. Data rich in that modern climate models can produce climatological quantities over long time periods with global coverage, including quantities that are difficult to measure and under conditions for which there is no data presently. Data poor in that the correspondence between climate model output to the actual climate, especially for future climate change due to human activities, is difficult to assess. The scope for fruitful interactions between climate scientists and statisticians is great, but requires serious commitments from researchers in both disciplines to understand the scientific and statistical nuances arising from the complex relationships between the data and the real-world problems. This paper describes a small fraction of some of the intellectual challenges that occur at the interface between climate science and statistics, including inferences for extremes for processes with seasonality and long-term trends, the use of climate model ensembles for studying extremes, the scope for using new data sources for studying space-time characteristics of environmental processes and a discussion of non-Gaussian space-time process models for climate variables. The paper concludes with a call to the statistical community to become more engaged in one of the great scientific and policy issues of our time, anthropogenic climate change and its impacts.

Citation

Download Citation

Michael L. Stein. "Some Statistical Issues in Climate Science." Statist. Sci. 35 (1) 31 - 41, February 2020. https://doi.org/10.1214/19-STS730

Information

Published: February 2020
First available in Project Euclid: 3 March 2020

MathSciNet: MR4071356
Digital Object Identifier: 10.1214/19-STS730

Rights: Copyright © 2020 Institute of Mathematical Statistics

JOURNAL ARTICLE
11 PAGES

This article is only available to subscribers.
It is not available for individual sale.
+ SAVE TO MY LIBRARY

SHARE
Vol.35 • No. 1 • February 2020
Back to Top