Integrating concepts of population exposure into atmospheric dispersion models at different spatial scales, taking into account individual mobility

The traditional approach of using static maps of residential population and annual average
concentrations to determine population exposure levels is not capable of taking into account the
spatial heterogeneity and the temporal variability of both ambient air pollutant concentrations, and the
fact that populations are highly mobile. People spend substantial amounts of time at work places,
schools, universities, often far away from their residence. In the United Kingdom, the 2011 census
revealed that for some local authorities in the city of London, the population during a working day was
tens of times larger than outside of working hours. This is, to a varying degree, the case in all urban
areas. As pollution levels vary due to the temporal profile of emissions (driven by human activities),
meteorology, physical transport and chemical transformation as well, applying state-of-the-art
atmospheric chemistry transport models (ACTMs), integrated with the latest information on population
distribution, offer the capability of quantifying human exposure in a dynamic fashion and with high
spatial resolution. However, spatial and temporal resolution are related to at times substantial costs,
in computing time, in the amount and degree of detail of input data required, and output data
generated. For this reason, applying a nested approach with urban scale dispersion models (e.g.
ADMS-Urban) within regional ACTMs (e.g. EMEP4UK) provides a suitable balance by providing the
necessary resolution where it matters, while being efficient with regard to computing time and data
needs overall.
In this paper, we focus on two aspects, first, we introduce the state of work on integrating data from
the 2011 census to generate a consistent, detailed population data product for ingestion in our air
pollution models. Secondly, we demonstrate the approach taken for a one-way nesting of the ADMSUrban model within EMEP4UK. Finally, we illustrate the direct relevance and application of this
approach for the development of national air pollution control policies on the example of identifying
options for reducing population exposure to fine particulate matter (PM2.5) in the United Kingdom.
The research described here is work in progress, as the census 2011 data have only recently been
made available. Data processing is currently being completed with the results being computed in time
for both the submission of the final version of this paper, as well as for presentation at iEMSs in San
Diego. This paper will be revised accordingly for final submission to include these results.