A geospatial approach to the prediction of indoor radon vulnerability in British Columbia, Canada

Michael C. Branion-Calles, Trisalyn Nelson, Sarah B. Henderson

Research output: Contribution to journalArticle

2 Citations (Scopus)

Abstract

Radon is a carcinogenic radioactive gas produced by the decay of uranium. Accumulation of radon in residential structures contributes to lung cancer mortality. The goal of this research is to predict residential radon vulnerability classes for the province of British Columbia (BC) at aggregated spatial units. Spatially referenced indoor radon concentration data were partitioned into low, medium, and high classes of radon vulnerability. Radon vulnerability classes were then linked to environmental and housing data derived from existing geospatial datasets. A Balanced Random Forests algorithm was used to model environmental predictors of indoor radon vulnerability and values at un-sampled locations across BC. A model was generated and evaluated using accuracy, precision, and kappa statistics. The influence of predictor variables was investigated through variable importance and partial dependence plots. The model performed 34% better than a random classifier. Increased probabilities of high vulnerability were associated with cold and dry winters, close proximity to major river systems, and fluvioglacial and colluvial soil parent materials. The Kootenays and Columbia-Shuswap regions were most at risk. Here, we present a novel method for predictive radon mapping that is broadly applicable to regions throughout the world.

Original languageEnglish (US)
Pages (from-to)554-565
Number of pages12
JournalJournal of Exposure Science and Environmental Epidemiology
Volume26
Issue number6
DOIs
StatePublished - Nov 1 2016
Externally publishedYes

Fingerprint

Radon
British Columbia
Canada
Uranium
Rivers
Lung Neoplasms
Classifiers
Soil
Gases
Statistics
Soils
Mortality

Keywords

  • environmental monitoring
  • exposure modeling
  • Keywordsempirical/statistical model

ASJC Scopus subject areas

  • Epidemiology
  • Toxicology
  • Pollution
  • Public Health, Environmental and Occupational Health

Cite this

A geospatial approach to the prediction of indoor radon vulnerability in British Columbia, Canada. / Branion-Calles, Michael C.; Nelson, Trisalyn; Henderson, Sarah B.

In: Journal of Exposure Science and Environmental Epidemiology, Vol. 26, No. 6, 01.11.2016, p. 554-565.

Research output: Contribution to journalArticle

@article{7ee95b4a100241f3995a6ae42a01a1c1,
title = "A geospatial approach to the prediction of indoor radon vulnerability in British Columbia, Canada",
abstract = "Radon is a carcinogenic radioactive gas produced by the decay of uranium. Accumulation of radon in residential structures contributes to lung cancer mortality. The goal of this research is to predict residential radon vulnerability classes for the province of British Columbia (BC) at aggregated spatial units. Spatially referenced indoor radon concentration data were partitioned into low, medium, and high classes of radon vulnerability. Radon vulnerability classes were then linked to environmental and housing data derived from existing geospatial datasets. A Balanced Random Forests algorithm was used to model environmental predictors of indoor radon vulnerability and values at un-sampled locations across BC. A model was generated and evaluated using accuracy, precision, and kappa statistics. The influence of predictor variables was investigated through variable importance and partial dependence plots. The model performed 34{\%} better than a random classifier. Increased probabilities of high vulnerability were associated with cold and dry winters, close proximity to major river systems, and fluvioglacial and colluvial soil parent materials. The Kootenays and Columbia-Shuswap regions were most at risk. Here, we present a novel method for predictive radon mapping that is broadly applicable to regions throughout the world.",
keywords = "environmental monitoring, exposure modeling, Keywordsempirical/statistical model",
author = "Branion-Calles, {Michael C.} and Trisalyn Nelson and Henderson, {Sarah B.}",
year = "2016",
month = "11",
day = "1",
doi = "10.1038/jes.2015.20",
language = "English (US)",
volume = "26",
pages = "554--565",
journal = "Journal of Exposure Science and Environmental Epidemiology",
issn = "1559-0631",
publisher = "Nature Publishing Group",
number = "6",

}

TY - JOUR

T1 - A geospatial approach to the prediction of indoor radon vulnerability in British Columbia, Canada

AU - Branion-Calles, Michael C.

AU - Nelson, Trisalyn

AU - Henderson, Sarah B.

PY - 2016/11/1

Y1 - 2016/11/1

N2 - Radon is a carcinogenic radioactive gas produced by the decay of uranium. Accumulation of radon in residential structures contributes to lung cancer mortality. The goal of this research is to predict residential radon vulnerability classes for the province of British Columbia (BC) at aggregated spatial units. Spatially referenced indoor radon concentration data were partitioned into low, medium, and high classes of radon vulnerability. Radon vulnerability classes were then linked to environmental and housing data derived from existing geospatial datasets. A Balanced Random Forests algorithm was used to model environmental predictors of indoor radon vulnerability and values at un-sampled locations across BC. A model was generated and evaluated using accuracy, precision, and kappa statistics. The influence of predictor variables was investigated through variable importance and partial dependence plots. The model performed 34% better than a random classifier. Increased probabilities of high vulnerability were associated with cold and dry winters, close proximity to major river systems, and fluvioglacial and colluvial soil parent materials. The Kootenays and Columbia-Shuswap regions were most at risk. Here, we present a novel method for predictive radon mapping that is broadly applicable to regions throughout the world.

AB - Radon is a carcinogenic radioactive gas produced by the decay of uranium. Accumulation of radon in residential structures contributes to lung cancer mortality. The goal of this research is to predict residential radon vulnerability classes for the province of British Columbia (BC) at aggregated spatial units. Spatially referenced indoor radon concentration data were partitioned into low, medium, and high classes of radon vulnerability. Radon vulnerability classes were then linked to environmental and housing data derived from existing geospatial datasets. A Balanced Random Forests algorithm was used to model environmental predictors of indoor radon vulnerability and values at un-sampled locations across BC. A model was generated and evaluated using accuracy, precision, and kappa statistics. The influence of predictor variables was investigated through variable importance and partial dependence plots. The model performed 34% better than a random classifier. Increased probabilities of high vulnerability were associated with cold and dry winters, close proximity to major river systems, and fluvioglacial and colluvial soil parent materials. The Kootenays and Columbia-Shuswap regions were most at risk. Here, we present a novel method for predictive radon mapping that is broadly applicable to regions throughout the world.

KW - environmental monitoring

KW - exposure modeling

KW - Keywordsempirical/statistical model

UR - http://www.scopus.com/inward/record.url?scp=84925645335&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=84925645335&partnerID=8YFLogxK

U2 - 10.1038/jes.2015.20

DO - 10.1038/jes.2015.20

M3 - Article

VL - 26

SP - 554

EP - 565

JO - Journal of Exposure Science and Environmental Epidemiology

JF - Journal of Exposure Science and Environmental Epidemiology

SN - 1559-0631

IS - 6

ER -