A Bayesian Spatio-Temporal Model of Temperature- and Humidity-Related Mortality Using High-Resolution Climate Data

TL;DR

This paper presents a Bayesian spatio-temporal model that incorporates gender-specific effects of climate variables on mortality, utilizing high-resolution data to improve understanding of climate-related health risks at fine spatial and temporal scales.

Contribution

It introduces a novel Bayesian model that explicitly accounts for gender-specific meteorological effects and complex dependencies across space, time, age, and gender.

Findings

Identifies district-level mortality patterns related to climate variables.

Quantifies weekly climate-related mortality risks.

Provides detailed insights for public health surveillance.

Abstract

In this study, we introduce a novel and comprehensive extension of a Bayesian spatio-temporal disease mapping model that explicitly accounts for gender-specific effects of meteorological exposures. Leveraging fine-scale weekly mortality and high-resolution climate data from Austria (2002 to 2019), we assess how interactions between temperature, humidity, age, and gender influence mortality patterns. Our approach goes beyond conventional modelling by capturing complex dependencies through structured interactions across space-time, space-age, and age-time dimensions, allowing us to capture complex demographic and environmental dependencies. The analysis identifies district-level mortality patterns and quantifies climate-related risks on a weekly basis, offering new insights for public health surveillance.