Using geostatistics and clustering to design and optmize the environmental monitoring network for Hai Duong province (Vietnam)

Abstract

This paper presents some mathematical modeling approaches for designing an environmental monitoring network (EMN). As an example application region, Hai Duong province was chosen. The results of this paper include the general picture of air quality in Hai Duong province, a preliminary air quality monitoring network with 20 monitoring stations and a dendrogram for optimizing the network. The works proved that geostatistics and cluster analysis can be used for designing and optimizing the EMN from scratch. The procedure to generate a monitoring network includes: i) developing an air quality index (AQI) from historical data or surveying data; ii) determining practical range (optimum distance between monitoring station) from the variogram with AQI and design of the preliminary monitoring network; iii) operate the preliminary network; iv) using monitored data of the preliminary network to generate a dendrogram for optimizing network; v) re-arrange EMN.