Superstatistical approach to air pollution statistics

Abstract

Air pollution by nitrogen oxides (NO_{x}) is a major concern in large cities as it has severe adverse health effects. However, the statistical properties of air pollutants are not fully understood. Here, we use methods borrowed from nonequilibrium statistical mechanics to construct suitable superstatistical models for air pollution statistics. In particular, we analyze time series of nitritic oxide (NO) and nitrogen dioxide (NO_{2}) concentrations recorded at several locations throughout Greater London. We find that the probability distributions of concentrations have heavy tails and that the dynamics is well described by χ^{2} superstatistics for NO and inverse-χ^{2} superstatistics for NO_{2}. Our results can be used to give precise risk estimates of high-pollution situations and pave the way to mitigation strategies.