Abstract
Air pollution poses a major problem in modern cities, as it has a significant effect in poor quality of life of the general population. Many recent studies link excess levels of major air pollutants with health-related incidents, in particular respiratory-related diseases. This introduces the need for city pollution on-line monitoring to enable quick identification of deviations from “normal” pollution levels, and providing useful information to public authorities for public protection. This article considers dynamic monitoring of pollution data (output of multivariate processes) using Kalman filters and multivariate statistical process control techniques. A state space model is used to define the in-control process dynamics, involving trend and seasonality. Distribution-free monitoring of the residuals of that model is proposed, based on binomial-type and generalised binomial-type statistics as well as on rank statistics. We discuss the general problem of detecting a change in pollutant levels that affects either the entire city (globally) or specific sub-areas (locally). The proposed methodology is illustrated using data, consisting of ozone, nitrogen oxides and sulfur dioxide collected over the air-quality monitoring network of Athens.
Highlights
In recent years, Statistical Process Control (SPC) has been proposed in environmental related monitoring problems (Pan and Chen 2008; Paroissin et al 2016)
(ii) Control Statistics Based on Ranks Regrouping the components of et in such a way that the same air quality variable from all the recording stations are placed together and establishing a similar in nature to the rule of Section 3.2.2(ii), we develop immediately a control procedure which aims in identifying if a specific air quality characteristic is out of control
We note that using the TB,3 statistic we may sum runs of excess length that belong to consecutive blocks of variables
Summary
Statistical Process Control (SPC) has been proposed in environmental related monitoring problems (Pan and Chen 2008; Paroissin et al 2016). Distribution-free statistical procedures for process-monitoring are favoured to parametric-based methods, as they relax the distributional assumption of the observed data. Such non-parametric procedures usually focus on univariate i.i.d. processes. As it is well known in non-industrial processes several variables are often observed and exhibit significant autocorrelation, i.e. the observed process is a multivariate time series. Examples of such process include, but are not limited to, environmental and financial processes
Sign-in/Register to view highlights & summary Sign-in/Register to access full text options 
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
