Large-scale Atmospheric Control on Non-Gaussian Tails of Midlatitude Temperature Distributions

Marianna Linz, Gang Chen and Zeyuan Hu

Abstract

Observed surface temperature distributions are non-Gaussian, which has important implications for the likelihood of extreme events in a changing climate. We use a two-dimensional advection-diffusion model of temperature stirred by stochastically generated Rossby waves with a sustained background temperature gradient to explore non-Gaussian temperature distributions. We examine how these distributions change with changes to thermal relaxation and eddy stirring. Weakening the background temperature gradient leads to decreased variance but no changes in other moments, while the eddy properties affect both the variance and skewness. A poleward movement of eddy stirring latitude leads to reduced skewness for most latitudes, implying a shift toward longer negative tails in temperature distributions, all else being equal. In contrast, the dependence of temperature skewness on eddy speed is a nuanced, nonlinear relationship.

Gang Chen

Professor, Dept. of Atmospheric and Oceanic Sciences, University of California, Los Angeles

Math Sci Building 7149, Los Angeles, CA 90095

Google Scholar ResearchGate ORCID GitHub Twitter

Large-scale Atmospheric Control on Non-Gaussian Tails of Midlatitude Temperature Distributions

Published in Geophysical Research Letters, 2018

Recommended citation: Marianna Linz, Gang Chen and Zeyuan Hu, 2018: Large-scale Atmospheric Control on Non-Gaussian Tails of Midlatitude Temperature Distributions, Geophysical Research Letters, 1–9, doi:10.1029/2018GL079324.

Abstract

Observed surface temperature distributions are non-Gaussian, which has important implications for the likelihood of extreme events in a changing climate. We use a two-dimensional advection-diffusion model of temperature stirred by stochastically generated Rossby waves with a sustained background temperature gradient to explore non-Gaussian temperature distributions. We examine how these distributions change with changes to thermal relaxation and eddy stirring. Weakening the background temperature gradient leads to decreased variance but no changes in other moments, while the eddy properties affect both the variance and skewness. A poleward movement of eddy stirring latitude leads to reduced skewness for most latitudes, implying a shift toward longer negative tails in temperature distributions, all else being equal. In contrast, the dependence of temperature skewness on eddy speed is a nuanced, nonlinear relationship.

Download paper