Developing an advanced PM 25 exposure model in Lima, Peru

Vu, B.N.; Liu, Y.; Checkley, W.; Bi, J.; Hansel, N.N.; Gonzales Rengifo, Gustavo Francisco; Xiao, Q.; Steenland, K.; Sánchez, O.

dc.contributor.author	Vu, B.N.
dc.contributor.author	Sánchez, O.
dc.contributor.author	Bi, J.
dc.contributor.author	Xiao, Q.
dc.contributor.author	Hansel, N.N.
dc.contributor.author	Checkley, W.
dc.contributor.author	Gonzales Rengifo, Gustavo Francisco
dc.contributor.author	Steenland, K.
dc.contributor.author	Liu, Y.
dc.date.accessioned	2019-07-04T16:59:28Z
dc.date.available	2019-07-04T16:59:28Z
dc.date.issued	2019
dc.identifier.uri	https://hdl.handle.net/20.500.12866/6765
dc.description.abstract	It is well recognized that exposure to fine particulate matter (PM 2.5 ) affects health adversely, yet few studies from South America have documented such associations due to the sparsity of PM 2.5 measurements. Lima's topography and aging vehicular fleet results in severe air pollution with limited amounts of monitors to effectively quantify PM 2.5 levels for epidemiologic studies. We developed an advanced machine learning model to estimate daily PM 2.5 concentrations at a 1 km 2 spatial resolution in Lima, Peru from 2010 to 2016. We combined aerosol optical depth (AOD), meteorological fields from the European Centre for Medium-Range Weather Forecasts (ECMWF), parameters from theWeather Research and Forecasting model coupled with Chemistry (WRF-Chem), and land use variables to fit a random forest model against ground measurements from 16 monitoring stations. Overall cross-validation R 2 (and root mean square prediction error, RMSE) for the random forest model was 0.70 (5.97 μg/m 3 ). Mean PM 2.5 for ground measurements was 24.7 μg/m 3 while mean estimated PM 2.5 was 24.9 μg/m 3 in the cross-validation dataset. The mean difference between ground and predicted measurements was -0.09 μg/m 3 (Std.Dev. = 5.97 μg/m 3 ), with 94.5% of observations falling within 2 standard deviations of the difference indicating good agreement between ground measurements and predicted estimates. Surface downwards solar radiation, temperature, relative humidity, and AOD were the most important predictors, while percent urbanization, albedo, and cloud fraction were the least important predictors. Comparison of monthly mean measurements between ground and predicted PM 2.5 shows good precision and accuracy from our model. Furthermore, mean annual maps of PM 2.5 show consistent lower concentrations in the coast and higher concentrations in the mountains, resulting from prevailing coastal winds blown from the Pacific Ocean in the west. Our model allows for construction of long-term historical daily PM 2.5 measurements at 1 km 2 spatial resolution to support future epidemiological studies.	en_US
dc.language.iso	eng
dc.publisher	MDPI
dc.relation.ispartofseries	Remote Sensing
dc.rights	info:eu-repo/semantics/restrictedAccess
dc.rights.uri	https://creativecommons.org/licenses/by-nc-nd/4.0/deed.es
dc.subject	Lima	en_US
dc.subject	Peru	en_US
dc.subject	Air pollution	en_US
dc.subject	Remote sensing	en_US
dc.subject	Machine learning	en_US
dc.subject	Random forests	en_US
dc.subject	Decision trees	en_US
dc.subject	Image resolution	en_US
dc.subject	Land use	en_US
dc.subject	Learning systems	en_US
dc.subject	MAIAC AOD	en_US
dc.subject	PM 2.5	en_US
dc.subject	PM2.5	en_US
dc.subject	Random forest	en_US
dc.subject	Solar radiation	en_US
dc.subject	Topography	en_US
dc.subject	Weather forecasting	en_US
dc.subject	WRF-chem	en_US
dc.title	Developing an advanced PM 25 exposure model in Lima, Peru	en_US
dc.type	info:eu-repo/semantics/article
dc.identifier.doi	https://doi.org/10.3390/rs11060641
dc.subject.ocde	https://purl.org/pe-repo/ocde/ford#1.05.00
dc.relation.issn	2072-4292