氡
遥感
高分辨率
环境科学
分辨率(逻辑)
地理
气象学
计算机科学
物理
人工智能
核物理学
作者
Longxiang Li,Brent A. Coull,Carolina L. Zilli Vieira,Petros Koutrakis
标识
DOI:10.1073/pnas.2408084121
摘要
Radon, a common radioactive indoor air pollutant, is the second leading cause of lung cancer in the United States. Knowledge about its distribution is essential for risk assessment and designing efficient protective regulations. However, the three current radon maps for the United States are unable to provide the up-to-date, high-resolution, and time-varying radon concentrations. Tens of millions of radon measurements have been conducted as parts of property inspections in the past two decades, making it possible for us to improve the national radon map. We compiled a national database of over 6 million radon measurements conducted by independent laboratories during 2001 to 2021. A random forest model was built to predict monthly community-level radon concentrations based on nearly 200 geological, meteorological, architectural, and socioeconomical factors. Our radon map can accurately show the distribution of radon at higher spatial and temporal resolutions. We observed slight decreases in average radon concentrations in high-radon regions during the study period. But over 83 million people are living in residences with radon concentrations at screening floor over 148 Bq/m3 (the recommended action level). Most of these residences are in low-radon zones, highlighting the need for comprehensive radon surveys. The high-resolution radon maps can be used by federal and local governments to design, update, and improve the regulations. Furthermore, the model can be used to assess residential exposure to radon, thus facilitating studies to expand our understanding of radon's health effects.
科研通智能强力驱动
Strongly Powered by AbleSci AI