Back to search

Cornell University

Climate Extremes In A General Climate Model With Stochastic Parameterizations

Abstract

dc:description.abstract

This work employs techniques from extreme value theory to evaluate the representation of temperature and precipitation extremes in two climate models and an observational dataset. The climate models correspond to the general climate model, the NCAR Community Atmosphere Model version 4 (CAM4), with two stochastic parameterizations of sub-grid scale processes: the stochastic kinetic energy backscatter (SKEBS) scheme and the stochastically perturbed parameterization tendency (SPPT) scheme. The observational dataset is version 7 of the satellite-based Tropical Rainfall Measuring Mission (TRMM) Multisatellite Precipitation Analysis (TMPA) 3B42 research product, developed at the National Aeronautics and Space Administration Goddard Space Flight Center. Temperature extremes are described in terms of the 95th percentile (20-yr return level) of the distribution of annual extremes of near-surface temperature, while precipitation extremes are characterized in terms of the analogous percentile with respect to daily precipitation amounts, in addition to less extreme precipitation statistics. The distribution of annual extremes is assumed to be well-approximated by the Generalized Extreme Value (GEV) distribution, which is fit at each gridpoint using a "block maxima" approach. A Bayesian hierarchical approach is used to estimate the parameters of the distribution of annual extremes in the 3B42 dataset. CAM4 overestimates warm and cold extremes over land regions, particu- larly over the Northern Hemisphere when compared against observations and reanalysis. The addition of a stochastic parameterization generally produces a warming of both warm and cold extremes relative to the unperturbed configuration, however, neither of the proposed parameterizations meaningfully reduce the biases in the simulated temperature extremes of CAM4. Similarly, the precipitation response to the use of stochastic parameterizations is remarkably muted, particularly that to SPPT. SKEBS is shown to enhance the dry bias of annual precipitation in CAM4 over the central contiguous United States, and also exacerbates the shortfall of moderate precipitation extremes over the same region. The 3B42 dataset shows severe overestimation of 20-yr return levels over eastern Asia. The analysis of the 3 parameters that define the GEV distribution enhances the understanding of the behavior of extremes, revealing valuable information that may potentially help modeling centers improve the simulation of extremes in the climate models.

Degree

thesis:*
Name thesis:degree_name
Ph. D., Operations Research
Level thesis:degree_level
Doctor of Philosophy
Discipline thesis:degree_discipline
Operations Research
Grantor
Cornell University
Year dc:date.issued
2016

Author and committee

dc:creator, dc:contributor.*
Author dc:creator
  • Tagle, Felipe
Chair dc:contributor.chair
  • Samorodnitsky,Gennady
Committee members dc:contributor.committeemember
  • Gross,Leonard
  • Mahowald,Natalie M

Subjects

dc:subject × 2

Rights

Language dc:language.iso
en_US

Identifiers

dc:identifier.*
Dc Identifier Bibid
9597040

Chain of custody

source
Harvested from
Cornell University
Base URL
ecommons.cornell.edu/server/oai/request
Last updated
2026-11-19
Source record
OAI-PMH GetRecord
citation

Tagle, Felipe. Climate Extremes In A General Climate Model With Stochastic Parameterizations. Doctor of Philosophy thesis, Cornell University, 2016. https://hdl.handle.net/1813/43610