Assessment and visualization of threshold exceedance probabilities in complex space–time settings: A case study of air quality in Northern Italy

Abstract

Among the many atmospheric pollutants, fine particles are known to be particularly damaging to respiratory health, and therefore many efforts are being made worldwide to explore their spatio-temporal behavior. In this paper we focus on PM 10 , specifically addressing the probability (or risk) that such particles will exceed potentially harmful thresholds. We combine smoothing in the time domain with spatial interpolation to model threshold exceedance probabilities and their corresponding confidence regions in a flexible framework. We then present a comprehensive study of air quality in the North-Italian region Piemonte from October 2005 through March 2006. The proposed methodology, consisting of a two-stage modeling approach followed by a block bootstrap scheme, has a myriad applications to other research fields.

Spatial statistical modelling of urban particulate matter

“…A semi-parametric method for deriving exceedance probabilities and associated confidence intervals based on bootstrapping can also be applied in further work (Schelin and Sjöstedt-de Luna, 2010;Cameletti et al, 2013a). This will be an opportunity to evaluate differences in results from the two methods.…”