Ensemble averaging using remote sensing data to model spatiotemporal PM 10 concentrations in sparsely monitored South Africa.

Autor: Arowosegbe OO; Swiss Tropical and Public Health Institute, Allschwil, Switzerland; University of Basel, Basel, Switzerland., Röösli M; Swiss Tropical and Public Health Institute, Allschwil, Switzerland; University of Basel, Basel, Switzerland., Künzli N; Swiss Tropical and Public Health Institute, Allschwil, Switzerland; University of Basel, Basel, Switzerland., Saucy A; Swiss Tropical and Public Health Institute, Allschwil, Switzerland; University of Basel, Basel, Switzerland., Adebayo-Ojo TC; Swiss Tropical and Public Health Institute, Allschwil, Switzerland; University of Basel, Basel, Switzerland., Schwartz J; Department of Environmental Health, Harvard T.H. Chan School of Public Health, Boston, MA, USA., Kebalepile M; Department for Education Innovation, University of Pretoria, Pretoria, South Africa., Jeebhay MF; Centre for Environmental and Occupational Health Research, School of Public Health and Family Medicine, University of Cape Town, Cape Town, South Africa., Dalvie MA; Centre for Environmental and Occupational Health Research, School of Public Health and Family Medicine, University of Cape Town, Cape Town, South Africa., de Hoogh K; Swiss Tropical and Public Health Institute, Allschwil, Switzerland; University of Basel, Basel, Switzerland. Electronic address: c.dehoogh@swisstph.ch.
Jazyk: angličtina
Zdroj: Environmental pollution (Barking, Essex : 1987) [Environ Pollut] 2022 Oct 01; Vol. 310, pp. 119883. Date of Electronic Publication: 2022 Aug 03.
DOI: 10.1016/j.envpol.2022.119883
Abstrakt: There is a paucity of air quality data in sub-Saharan African countries to inform science driven air quality management and epidemiological studies. We investigated the use of available remote-sensing aerosol optical depth (AOD) data to develop spatially and temporally resolved models to predict daily particulate matter (PM 10 ) concentrations across four provinces of South Africa (Gauteng, Mpumalanga, KwaZulu-Natal and Western Cape) for the year 2016 in a two-staged approach. In stage 1, a Random Forest (RF) model was used to impute Multiangle Implementation of Atmospheric Correction AOD data for days where it was missing. In stage 2, the machine learner algorithms RF, Gradient Boosting and Support Vector Regression were used to model the relationship between ground-monitored PM 10 data, AOD and other spatial and temporal predictors. These were subsequently combined in an ensemble model to predict daily PM 10 concentrations at 1 km × 1 km spatial resolution across the four provinces. An out-of-bag R 2 of 0.96 was achieved for the first stage model. The stage 2 cross-validated (CV) ensemble model captured 0.84 variability in ground-monitored PM 10 with a spatial CV R 2 of 0.48 and temporal CV R 2 of 0.80. The stage 2 model indicated an optimal performance of the daily predictions when aggregated to monthly and annual means. Our results suggest that a combination of remote sensing data, chemical transport model estimates and other spatiotemporal predictors has the potential to improve air quality exposure data in South Africa's major industrial provinces. In particular, the use of a combined ensemble approach was found to be useful for this area with limited availability of air pollution ground monitoring data.
(Copyright © 2022 The Authors. Published by Elsevier Ltd.. All rights reserved.)
Databáze: MEDLINE
Externí odkaz: