A Bayesian model averaging approach for estimating the relative risk of mortality associated with heat waves in 105 U.S. cities

Jennifer F. Bobb, Francesca Dominici, Roger Peng

Research output: Contribution to journalArticle

Abstract

Estimating the risks heat waves pose to human health is a critical part of assessing the future impact of climate change. In this article, we propose a flexible class of time series models to estimate the relative risk of mortality associated with heat waves and conduct Bayesian model averaging (BMA) to account for the multiplicity of potential models. Applying these methods to data from 105 U.S. cities for the period 1987-2005, we identify those cities having a high posterior probability of increased mortality risk during heat waves, examine the heterogeneity of the posterior distributions of mortality risk across cities, assess sensitivity of the results to the selection of prior distributions, and compare our BMA results to a model selection approach. Our results show that no single model best predicts risk across the majority of cities, and that for some cities heat-wave risk estimation is sensitive to model choice. Although model averaging leads to posterior distributions with increased variance as compared to statistical inference conditional on a model obtained through model selection, we find that the posterior mean of heat wave mortality risk is robust to accounting for model uncertainty over a broad class of models.

Original languageEnglish (US)
Pages (from-to)1605-1616
Number of pages12
JournalBiometrics
Volume67
Issue number4
DOIs
StatePublished - Dec 2011

Fingerprint

Infrared Rays
Bayesian Model Averaging
Relative Risk
relative risk
Mortality
Heat
heat
Posterior distribution
Model Selection
Model Averaging
Model Choice
Posterior Mean
Posterior Probability
Climate Change
Model Uncertainty
Time Series Models
Prior distribution
Statistical Inference
Model
Hot Temperature

Keywords

  • Climate change
  • Generalized additive models
  • Model uncertainty
  • Time series data

ASJC Scopus subject areas

  • Applied Mathematics
  • Statistics and Probability
  • Agricultural and Biological Sciences(all)
  • Biochemistry, Genetics and Molecular Biology(all)
  • Immunology and Microbiology(all)
  • Medicine(all)

Cite this

A Bayesian model averaging approach for estimating the relative risk of mortality associated with heat waves in 105 U.S. cities. / Bobb, Jennifer F.; Dominici, Francesca; Peng, Roger.

In: Biometrics, Vol. 67, No. 4, 12.2011, p. 1605-1616.

Research output: Contribution to journalArticle

@article{c0e24788d73145669400ed4beb8875cd,
title = "A Bayesian model averaging approach for estimating the relative risk of mortality associated with heat waves in 105 U.S. cities",
abstract = "Estimating the risks heat waves pose to human health is a critical part of assessing the future impact of climate change. In this article, we propose a flexible class of time series models to estimate the relative risk of mortality associated with heat waves and conduct Bayesian model averaging (BMA) to account for the multiplicity of potential models. Applying these methods to data from 105 U.S. cities for the period 1987-2005, we identify those cities having a high posterior probability of increased mortality risk during heat waves, examine the heterogeneity of the posterior distributions of mortality risk across cities, assess sensitivity of the results to the selection of prior distributions, and compare our BMA results to a model selection approach. Our results show that no single model best predicts risk across the majority of cities, and that for some cities heat-wave risk estimation is sensitive to model choice. Although model averaging leads to posterior distributions with increased variance as compared to statistical inference conditional on a model obtained through model selection, we find that the posterior mean of heat wave mortality risk is robust to accounting for model uncertainty over a broad class of models.",
keywords = "Climate change, Generalized additive models, Model uncertainty, Time series data",
author = "Bobb, {Jennifer F.} and Francesca Dominici and Roger Peng",
year = "2011",
month = "12",
doi = "10.1111/j.1541-0420.2011.01583.x",
language = "English (US)",
volume = "67",
pages = "1605--1616",
journal = "Biometrics",
issn = "0006-341X",
publisher = "Wiley-Blackwell",
number = "4",

}

TY - JOUR

T1 - A Bayesian model averaging approach for estimating the relative risk of mortality associated with heat waves in 105 U.S. cities

AU - Bobb, Jennifer F.

AU - Dominici, Francesca

AU - Peng, Roger

PY - 2011/12

Y1 - 2011/12

N2 - Estimating the risks heat waves pose to human health is a critical part of assessing the future impact of climate change. In this article, we propose a flexible class of time series models to estimate the relative risk of mortality associated with heat waves and conduct Bayesian model averaging (BMA) to account for the multiplicity of potential models. Applying these methods to data from 105 U.S. cities for the period 1987-2005, we identify those cities having a high posterior probability of increased mortality risk during heat waves, examine the heterogeneity of the posterior distributions of mortality risk across cities, assess sensitivity of the results to the selection of prior distributions, and compare our BMA results to a model selection approach. Our results show that no single model best predicts risk across the majority of cities, and that for some cities heat-wave risk estimation is sensitive to model choice. Although model averaging leads to posterior distributions with increased variance as compared to statistical inference conditional on a model obtained through model selection, we find that the posterior mean of heat wave mortality risk is robust to accounting for model uncertainty over a broad class of models.

AB - Estimating the risks heat waves pose to human health is a critical part of assessing the future impact of climate change. In this article, we propose a flexible class of time series models to estimate the relative risk of mortality associated with heat waves and conduct Bayesian model averaging (BMA) to account for the multiplicity of potential models. Applying these methods to data from 105 U.S. cities for the period 1987-2005, we identify those cities having a high posterior probability of increased mortality risk during heat waves, examine the heterogeneity of the posterior distributions of mortality risk across cities, assess sensitivity of the results to the selection of prior distributions, and compare our BMA results to a model selection approach. Our results show that no single model best predicts risk across the majority of cities, and that for some cities heat-wave risk estimation is sensitive to model choice. Although model averaging leads to posterior distributions with increased variance as compared to statistical inference conditional on a model obtained through model selection, we find that the posterior mean of heat wave mortality risk is robust to accounting for model uncertainty over a broad class of models.

KW - Climate change

KW - Generalized additive models

KW - Model uncertainty

KW - Time series data

UR - http://www.scopus.com/inward/record.url?scp=83655167427&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=83655167427&partnerID=8YFLogxK

U2 - 10.1111/j.1541-0420.2011.01583.x

DO - 10.1111/j.1541-0420.2011.01583.x

M3 - Article

C2 - 21447046

AN - SCOPUS:83655167427

VL - 67

SP - 1605

EP - 1616

JO - Biometrics

JF - Biometrics

SN - 0006-341X

IS - 4

ER -