Page 173 - IJES Special Issues for AIEC2016
P. 173
International Journal of Environment and Sustainability, 2016, 5(2): 72-88 73
(1982) based on the concept of a spatial correla- compromise programming technique for as-
tion analysis. The same concept combined with sessing the relocation of strategy of an urban air
potential of violation was used by Arbeloa et al., quality monitoring network with respect to the
(1993) to design an optimal air quality monitor- multi-objective and multi-pollutant design crite-
ing network. Noll and Mitsutome (1983) devel- ria. Silva and Quiroz (2003) used a Shannon in-
oped a method to design AQMN based on ex- formation index to optimize atmospheric moni-
pected ambient pollutant dosage. This method toring networks by excluding the least informa-
ranks potential locations by calculating the ratio tive stations with respect to different air pollu-
of a station’s expected dosage over the study tants. An optimal design of AQMN was conducted
area’s total dosage. Another methodology devel- around a refinery using a mathematical model
oped by Nakamori and Sawaragi (1984) deter- based on the multiple cell approach with simul-
mines the representative areas of monitor sta- taneous multiple pollutants by Elkamel et al.,
tions in urban areas. A different perspective, (2008). Lu et al., (2011) used principal compo-
based on the use of Shannon information, was in- nent analysis (PCA) and cluster analysis to opti-
itiated with the results of Caselton and Zidek mize the air quality monitoring network in Hong
(1984), applied in a univariate setup by Sampson Kong. PCA and fuzzy c-means clustering was ap-
and Guttorp (1992) and Guttorp et al., (1993) plied for assessment of air quality monitoring in
and later extended to a multivariate context by Turkey by Dogruparmak et al., (2014). The au-
Perez-Abreu and Rodriguez (1996). The concept thors showed that a number of monitoring sta-
of sphere of influence and figure of merit was ap- tions can be decreased using the methodology.
plied by McElroy et al., (1986) to calculate the Ferradas et al. (2011) developed a methodology
minimum number of air quality monitoring sites. based on a self-organizing map (SOM) artificial
A simple methodology for siting ambient air neural networks for integrating data about mul-
monitors at the neighborhood scale was devel- tiple measured pollutants to group monitoring
oped by Richard et al., (2002) and applied as a stations according to their similar air quality.
case study in Hudson County. The proposed method considered the subse-
quent geographical mapping of the clusters of
A linear programming approach was used by stations observed with the SOM, which made it
many researchers to site optimum AQMN. A possible to detect geographically different areas
multi-attribute utility function method was used that share similar air pollution problems.
for siting the air quality network by Kainuma et
al., (1990). Trujillo-Ventura and Ellis (1991) ap- The strides that the field Geographical Infor-
plied multiple objectives, including spatial cov- mation System (GIS) and its components (such
erage, violation detection, data validity and a as interpolation methods) are making as an ap-
weighting method, to determine the most suita- plication in almost every field are incredible. GIS
ble network for Tarragona, Spain. Chen et al., and spatial interpolation techniques were also
(2006) developed a multiple objective optimiza- used in AQMN. Bayraktar et al., (2005) used a
tion model for air quality monitoring networks Kriging-based approach to locate sampling sites
in Taoyuan County, Taiwan. A multiple objective for assessing air quality. Long years of the smoke
optimization model and a procedure for sustain- data measurements were used in the determina-
able air quality monitoring networks planning tion of the region for the representation by draw-
were developed. A holistic approach was ing contours through the Kriging method. Then
adapted for optimal design of air quality moni- the sampling site in this region was selected
toring network expansion in an urban area by based on the EPA criteria. External drift Kriging
Mofarrah and Husain (2010). In this approach a of NOx concentrations with dispersion model
multiple-criterion method with the spatial corre- output was used by Kassteele et al., (2009) to re-
lation technique was implemented to design an duce the uncertainties in parameter estimation
expanded air quality monitoring network of due to the reduction of air quality networks. GIS
Riyadh city in Saudi Arabia. Tseng and Ni-Bin ancillary variables were used to predict volatile
(2001) proposed a Genetic Algorithm (GA) based organic compounds and nitrogen dioxide levels
at unmonitored locations (Smith et al., 2006).
Science Target Inc. www.sciencetarget.com
