A geostatistical approach for assessing population exposure to NO2 in a complex urban area (Beirut, Lebanon)

Few studies in the Middle East region estimated the spatial distribution of air pollutants for exposure studies. This paper presents a geostatistical approach to assess background NO2 spatial distribution and the associated exposed population in a Mediterranean city with a complex topography, Beirut. Such modeling gave an accurate mapping of the 2010 yearly background average value of NO2: it varies between 35 and 67 mu g m(-3) with a mean of 53 mu g m(-3). The mean SD of the estimated error was about 3 mu g m(-3). The results showed that the spatial distribution of NO2 follows a nested structuring, with a major structure related to topoclimatic characteristics (interaction topography/atmospheric flow at large scale) and a minor one linked to micro-environment and micro-climatic characteristics (interactions urban morphology/atmospheric flows at fine scale). The probability for the city's inhabitants to be exposed to NO2 levels exceeding 40 mu g m(-3) threshold limit set by the World Health Organization (WHO) showed that Beirut city has a real sanitary risk to the NO2 pollution. 93 % of the population (around 358,459 people) is 100 % sure to be exposed to a yearly average exceeding 40 mu g m(-3). This knowledge will be certainly useful for developing a tool for decision support in order to implement policies of reducing air pollution in Beirut, which is, given the results, very urgent.