Developing an Advanced PM2.5 Exposure Model in Lima, Peru

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 km² 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 the Weather 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 R² (and root mean square prediction error, RMSE) for the random forest model was 0.70 (5.97 μg/m³). Mean PM_2.5 for ground measurements was 24.7 μg/m³ while mean estimated PM_2.5 was 24.9 μg/m³ in the cross-validation dataset. The mean difference between ground and predicted measurements was -0.09 μg/m³ (Std.Dev. = 5.97 μg/m³), 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 km² spatial resolution to support future epidemiological studies.