Multivariate air pollution prediction modeling with partial missingness

Abstract

Missing observations from air pollution monitoring networks has posed a longstanding problem for health investigators of air pollution. Growing interest in mixtures of air pollutants has further complicated this problem, as many new challenges have arisen that require development of novel methods. The objective of this study is to develop a methodology for multivariate prediction of air pollution. We focus specifically on tackling different forms of missing data such as spatial (sparse sites), outcome (pollutants not measured at some sites), and temporal (varieties of interrupted time series). To address these challenges, we develop a novel multivariate fusion framework, which leverages the observed interpollutant correlation structure to reduce error in the simultaneous prediction of multiple air pollutants. Our joint fusion model employs predictions from the Environmental Protection Agency's Community Multiscale Air Quality model along with spatiotemporal error terms. We have implemented our models on both simulated data and a case study in South Carolina for eight pollutants over a 28‐day period in June 2006. We found that our model, which uses a multivariate correlated error in a Bayesian framework, showed promising predictive accuracy particularly for gaseous pollutants.

A spatio‐temporal model for the analysis and prediction of fine particulate matter concentration in Beijing

“…We note that, though the proportion of the missing data is small, the imputation for missing covariates (including the missing response of PM 2.5 concentration) before modeling may result in bias and sensitivity issues for inference. For potentially tackling the missing data in the future, we suggest possible avenues are through making use of the space‐time dependence structure of the missing response (Boaz, Lawson, & Pearce, 2019; Calculli et al, 2015; Padilla et al, 2020), or building a multiple regression model for the missing covariates (Yi, Liu, & Wu, 2011).…”

Data fusion with Gaussian processes for estimation of environmental hazard events

“…For example, Keller and Peng (2019) present a framework for evaluating the error in aggregating areal exposure concentrations for air pollution epidemiology. Boaz, Lawson, and Pearce (2019) develop a multivariate fusion framework to deal with air pollution prediction with partial missingness. Wilkie et al (2019) treat data for fusion as realisations of smooth temporal functions, Gilani, Berrocal, and Batterman (2019) accounts for nonstationarity by incorporating covariates, postulated to drive the nonstationary behavior, in the covariance function and Ma and Kang (2020) develop a stochastic expectation–maximization that facilitates the use of large spatiotemporal datasets.…”

Structural equation models for simultaneous modeling of air pollutants

“…In this study, the proposed methodology is applied to a dataset of pollutants in European Union metropolitan areas. The analyzed dataset includes Worldwide Air Quality data (https://aqicn.org/), which covers pollutants and atmospheric conditions around the world, providing unified and worldwide air quality information (Boaz et al, 2019). In addition, to take into account the features of the analyzed cities, two other sources of data at municipal level have been employed: the Organization for Economic Co-operation and Development (OECD) Metropolitan database, which provides socio-economic and environmental indicators in 36 countries (OECD, 2012), and the Eurostat metropolitan regions (NUTS3) data (Eurostat, 2019).…”