Abstract

Health effect estimates depend on the methods of evaluating exposures. Due to non-linearities in the exposure-response relationships, both the predicted mean exposures as well as its spatial variability are significant. The aim of this work is to systematically quantify the impact of the spatial resolution on population-weighted mean concentration (PWC), its variance, and mortality attributable to fine particulate matter (PM2.5) exposure in Finland in 2015. The atmospheric chemical transport model SILAM was used to estimate the ambient air PM2.5 concentrations at 0.02° longitudinal × 0.01° latitudinal resolution (ca. 1 km), including both the national PM2.5 emissions and the long-range transport. The decision-support model FRES source-receptor matrices applied at 250-m resolution was used to model the ambient air concentrations of primary fine particulate matter (PPM2.5) from local and regional sources up to 10 km and 20 km distances. Numerical averaging of population and concentrations was used to produce the results for coarser resolutions. Population-weighted PM2.5 concentration was 11% lower at a resolution of 50 km, compared with the corresponding computations at a resolution of 1 km. However, considering only the national emissions, the influences of spatial averaging were substantially larger. The average population-weighted local PPM2.5 concentration originated from Finnish sources was 70% lower at a resolution of 50 km, compared with the corresponding result obtained using a resolution of 250 m. The sensitivity to spatial averaging, between the finest 250-m and the coarsest 50-km resolution, was highest for the emissions of PPM2.5 originated from national vehicular traffic (about 80% decrease) and lowest for the national residential combustion (60% decrease). Exposure estimates in urban areas were more sensitive to the changes of model resolution (14% and 74% decrease for PM2.5 and local PPM2.5, respectively), compared with estimates in rural areas (2% decrease for PM2.5 and 36% decrease for PPM2.5). We conclude that for the evaluation of the health impacts of air pollution, the resolution of the model computations is an important factor, which can potentially influence the predicted health impacts by tens of percent or more, especially when assessing the impacts of national emissions.

Highlights

  • Air pollution is one of the biggest environmental health risks globally (Gakidou et al 2017)

  • We study the sensitivities of predicted population-weighted PM2.5 concentrations, originated (i) from the national Finnish sources and dispersion and the long-range transport modelled with system for integrated modelling of atmospheric composition (SILAM), and (ii) from local Finnish sources modelled with FRES

  • Difference to the corresponding results computed on 1-km resolution was modest, but decrease of estimated concentration was substantially larger for the results computed on coarser resolutions

Read more

Summary

Introduction

Air pollution is one of the biggest environmental health risks globally (Gakidou et al 2017). In Finland, fine particles (PM2.5) were recently evaluated to be the most harmful ambient air pollutant (Lehtomäki et al 2018). Health impact estimates of air pollution are based on ambient concentrations which are used as exposure estimates. Concentrationresponse functions are needed to link the exposure to health outcomes. Concentrationresponse relationships for PM2.5 exposure have been assumed to be linear or log-linear (Pope 3rd et al 2015). These mainly work for low-exposure levels but lead to unrealistically large impacts in areas where the concentrations are very high (Burnett et al 2014)

Objectives
Methods
Results
Discussion
Conclusion
Full Text
Published version (Free)

Talk to us

Join us for a 30 min session where you can share your feedback and ask us any queries you have

Schedule a call