System Maintenance Notice

Please note that maintenance work is scheduled between 4 AM and 5 AM - CET. During this time, you may experience reduced performance. Thanks for your patience.

Publications

Journal Article Research Support, N.I.H., Extramural Research Support, Non-U.S. Gov't Research Support, U.S. Gov't, Non-P.H.S. Ppublish 181 citations

PubMed: 32808786 DOI: 10.1021/acs.est.0c01791

An Ensemble Learning Approach for Estimating High Spatiotemporal Resolution of Ground-Level Ozone in the Contiguous United States.

Environ Sci Technol 54(18):11037-11047

Authors: Weeberb J Requia, Qian Di, Rachel Silvern, James T Kelly, Petros Koutrakis, Loretta J Mickley, Melissa P Sulprizio, Heresh Amini, Liuhua Shi, Joel Schwartz

Affiliations

  • Department of Environmental Health, Harvard University, TH Chan School of Public Health, Boston, Massachusetts 02115, United States.
  • School of Public Policy and Government, Fundação Getúlio Vargas, Brasília, Distrito Federal 72125590, Brazil.
  • Department of Environmental Health, Harvard University, TH Chan School of Public Health, Boston, Massachusetts 02115, United States. Research Center for Public Health, Tsinghua University, Beijing 100084, China.
  • Harvard University, John A. Paulson School of Engineering and Applied Sciences, Cambridge, Massachusetts 02138, United States.
  • U.S. Environmental Protection Agency, Office of Air Quality Planning and Standards, Research Triangle Park, Durham, North Carolina 27709, United States.
  • Department of Public Health, Faculty of Health and Medical Sciences, University of Copenhagen, Copenhagen 1165, Denmark.
  • Gangarosa Department of Environmental Health, Rollins School of Public Health, Emory University, Atlanta, Georgia 30322, United States.

Abstract

In this paper, we integrated multiple types of predictor variables and three types of machine learners (neural network, random forest, and gradient boosting) into a geographically weighted ensemble model to estimate the daily maximum 8 h O 3 with high resolution over both space (at 1 km × 1 km grid cells covering the contiguous United States) and time (daily estimates between 2000 and 2016). We further quantify monthly model uncertainty for our 1 km × 1 km gridded domain. The results demonstrate high overall model performance with an average cross-validated R 2 (coefficient of determination) against observations of 0.90 and 0.86 for annual averages. Overall, the model performance of the three machine learning algorithms was quite similar. The overall model performance from the ensemble model outperformed those from any single algorithm. The East North Central region of the United States had the highest R 2, 0.93, and performance was weakest for the western mountainous regions (R 2 of 0.86) and New England (R 2 of 0.87). For the cross validation by season, our model had the best performance during summer with an R 2 of 0.88. This study can be useful for the environmental health community to more accurately estimate the health impacts of O 3 over space and time, especially in health studies at an intra-urban scale.

MeSH

Algorithms, Learning, Environmental Health, Sanitation

Cited by