Hostname: page-component-cd9895bd7-gbm5v Total loading time: 0 Render date: 2024-12-26T10:38:14.897Z Has data issue: false hasContentIssue false

Characterizing Nuclear Fallout Patterns in Atlanta, Georgia

Published online by Cambridge University Press:  13 July 2023

Morgan Taylor
Affiliation:
University of Georgia, Athens, USA
William Bell
Affiliation:
University of Georgia, Athens, USA
Curt Harris
Affiliation:
University of Georgia, Athens, USA
Rights & Permissions [Opens in a new window]

Abstract

Core share and HTML view are not available for this content. However, as you have access to this content, a full PDF is available via the ‘Save PDF’ action button.
Introduction:

Weather significantly affects the distribution of fallout radiation resulting from a nuclear detonation. Prior nuclear detonation models have either utilized a “typical” day for the city of interest or have chosen conditions that optimize fallout radiation. However, models that aid emergency planners should utilize representative weather conditions to capture the most likely distribution of fallout radiation for the region of interest.

Method:

Fallout radiation resulting from an improvised nuclear device detonation in Atlanta, Georgia, USA was simulated for each day in 2019 using the Hazard Prediction and Assessment Capability (HPAC) software and weather from Modern-Era Retrospective analysis for Research and Applications, Version 2 (MERRA-2). A partition around medoids cluster analysis was conducted, based on the characteristics of the plumes, population at risk, and estimated proportion of fatalities. A multinomial logistic regression, a decision tree, and a random forest model were then used to predict the cluster from surface-level weather data.

Results:

On average, the fallout plume was 160.25km long, had an area of 3,174.44 km2, and was angled 83.5° from due north. The plume on average contained 3,668,173 individuals at risk for exposure and caused 416,8908 casualties. Four clusters were identified to represent the distribution of fallout radiation. The random forest model was best able to predict the cluster using surface-level weather data, with an average accuracy of 57.24% (kappa = 0.385). The variable importance plot suggests north-westerly winds, cloud coverage at detonation, whether it is summer, and average temperature are among the most important variables for classification.

Conclusion:

Meaningful representation of the variation in the distribution of fallout radiation is imperative while creating nuclear detonation models. While an analysis of the fallout distribution throughout a calendar year provides important insight, future research may examine longer study periods to better understand the climatological impacts on fallout radiation.

Type
Poster Presentations
Copyright
© The Author(s), 2023. Published by Cambridge University Press on behalf of World Association for Disaster and Emergency Medicine