Predicting the impact of geology and housing characteristics on indoor radon concentration in Metropolitan Atlanta Area, US: using Multilevel Binary Regression Model

Keywords: Radon, Geology, Housing, Granite, Stories, Basement, Multilevel Regression Model
Session Type: Virtual Paper Abstract
Day: Sunday
Session Start / End Time: 2/27/2022 08:00 AM (Eastern Time (US & Canada)) - 2/27/2022 09:20 AM (Eastern Time (US & Canada))
Room: Virtual 3

Abstract

Radon is a harmful radioactive gas that leads to lung cancer. According to U.S. Environment Protection Agency, it causes more than 2,000 Americans to die every year. Radon originated from geological factors such as the soils and the fractures of bedrock and housing characteristics such as the housing materials and structure. This study aims to predict the impact of geology and housing characteristics on indoor radon concentration.
13 counties in the Atlanta Metropolitan Area were considered a study area. The radon data (n = 3,559) was obtained from a private vendor, Air Check, Incorporated (1990–2015). The geological data, including rock types and faults, was obtained from the Atlanta geologic map of the U.S. Geological Survey. The housing characteristics data, including housing age, exterior wall, stories, basement, and finished basement size, was obtained from Beacon and qPublic.net, Tax Assessor’s Office in Gwinnett County, and three private real estate websites.
The analysis methods are as follows: 1) Pearson chi-square test was used to examine the correlation between radon and the factors of geology and housing. 2) Only factors associated with radon were used for the multilevel binary regression analysis. It evaluated the impact of both each factor and the interaction effect between factors on radon.
It was proved that indoor radon levels were likely to be clustered in the areas with high fault density. The result showed that three factors of granite, stories, and the basement greatly impacted indoor radon levels, and their confluent impact was also significant statistically.