AL-Qadisiyah Journal For Administrative and Economic sciences

Spatio-Temporal Analysis of the Impact of Dust Storms on Hospital Admissions for Respiratory Diseases using Negative Binomial Models and Satellite Data within a Bayesian Probabilistic Framework

Document Type : Research Paper

Author

Ammar Kuti Nasser

AL- Mustansiriyah University

https://doi.org/10.33916/qjae.2026.01171183
Abstract
Iraq's environmental crisis, over the past 20 years, has become incomparable. The explosion of dust-affected days — 122 annually in 2000 to 283 by 2022, a rise of 132% — has been recorded by the European Centre for Medium-Range Weather Forecasts. During dust storm events, fine particulate matter (PM2.5) concentrations in the three provinces exceeded WHO air quality guidelines by 14 to 18 times. We obtained hospital admissions spanning three provinces (between January 2000 and December 2024) from 32 governmental hospitals, resulting in a total of 864 months for data collection. These clinical reports were cross-referenced with high-resolution satellite data: MODIS-MAIAC at a 1-km level, TROPOMI at a 3.5-km level, and ERA5 reanalysis datasets.The analytical strategy used a negative binomial count model in a Bayesian hierarchical framework with Markov Chain Monte Carlo simulation (50,000 iterations). The results indicate that for every 10 μg/m³ increase in PM2.5, there was a 3.74% increase in admissions of respiratory cases (95% BCI: 2.86-4.63%). The model had a good fit with the performance results: DIC = 8,342.6 and RMSE = 198.3. The peak health effects of exposure were found to be 2 day in the distributed lags models (coefficie nt=0.0142). The highest level of spatial variation was found between geographical areas; Basra, Baghdad, and Karbala were the most sensitive 0.187, 0.134, and 0.089 respectively.

Keywords

Applied statistics
Bayesian hierarchical modelling
credible intervals
distributed lag models
dust storms
overdispersion
PM2.5
respiratory diseases
spatiotemporal analysis
Iraq
Crossmark
AL-Qadisiyah Journal For Administrative and Economic sciences
Volume 28, Issue 1
Spring 2026
Pages 171-183

Home

Guide for Authors

Submit Manuscript

Reviewers

Contact Us