Analysis of air quality monitoring networks by functional clustering

Abstract

SUMMARYAir quality monitoring networks are important tools in management and evaluation of air quality. Classifying monitoring stations via homogeneous clusters allows identification of similarities in pollution, of representative sites, and of spatial patterns. Instead of summaries by statistical indicators, we propose to consider the air pollutant concentrations as functional data. We then classify using functional cluster analysis, where Partitioning Around Medoids (PAM) algorithm is embedded. The proposed data analysis approach is applied to the air quality monitoring network in Piemonte (Northern Italy); we consider the three more critical pollutants: NO 2 , PM 10 , and O 3 .

Functional zoning for air quality

“…Functional clustering methods (see Abraham et al 2003 andIgnaccolo et al 2008) could be directly employed on each pollutant: indeed considering grid points' pollutant time series as functional data and clustering estimated coefficients, we obtain several zoning outcomes of the same land. However, decisors need to jointly consider different pollutants in order to get a multi-pollutant zoning: for this reason we propose two strategies for pollutant aggregation.…”

“…Specifically, we consider air pollutant time series provided by a deterministic air quality model on a regular grid, and preprocessed by assimilating observations, as functional data (Ramsay and Silverman 2005). We then classify them by using functional clustering, where the Partitioning Around Medoids (PAM) algorithm is embedded (as in Ignaccolo et al 2008) in place of the k -means one, as proposed by Abraham et al (2003). Thus the allocation to a specific zone preserves information about pollution temporal patterns and does not take into account any other information.…”

A comparison of clustering approaches for the study of the temporal coherence of multiple time series

“…As detailed in Ignaccolo et al (2008), the b i vector is estimated by means of the least squares method and the G i curve is approximated byĜ i t ð Þ ¼ s i t;b i .…”

“…Functional clustering based on the k-means algorithm has been introduced in Abraham et al (2003) and a similar approach which used partitioning around mediods rather than means has been applied in Ignaccolo et al (2008). K-means is applied to the spline coefficient vectors in the R KþdÀ1 space and the clustering result directly provides the clustering of the time series.…”

A thermal infrared instrument onboard a geostationary platform for CO and O<sub>3</sub> measurements in the lowermost troposphere: Observing System Simulation Experiments (OSSE)

“…Traditionally, AQ monitoring has been done using measurements from ground-based stations. Ground-based in situ observations have the disadvantage of an inhomogeneous spatial coverage, and present a strong variability in their spatial representativeness, their measurement methods and correction factors (Ignaccolo et al, 2008). The main advantage of satellite observations is the good spatial coverage.…”