TY  - JOUR
T1  - The Use of Environmetric Techniques Combined with Sensitivity Analysis for the
Discrimination of Groundwater Quality Parameters
AU - Nasiru Usman, Usman AU - Juahir, Hafizan 
JO  - Environmental Research Journal
VL  - 13
IS  - 3
SP  - 73
EP  - 78
PY  - 2019
DA  - 2001/08/19
SN  - 1994-5396
DO  - erj.2019.73.78
UR  - https://makhillpublications.co/view-article.php?doi=erj.2019.73.78
KW  - Sensitivity analysis
KW  -cluster analysis
KW  -discriminant analysis
KW  -groundwater quality
KW  -groundwater pollution
AB  - Understanding the most effective pollutants
affecting groundwater quality is of utmost importance in
promoting sustainable development of groundwater
resource. The study was performed to reduce the less
significant parameter and give a preliminary judgment on
the most significant water quality parameters
discriminating the groundwater regions based on ANN
model. This, study shows the use of sensitivity analysis
combined with environmetric techniques such as Cluster
Analysis (CA), Discriminant Analysis (DA). The water
quality data was obtained from 10 different wells, over
the period of 6 years (2006-2011) using 24 water quality
parameters. Sensitivity analysis was carried out for nine
models (ANN-R-AP, ANN-R-Na<sup>+</sup>, ANN-R-Ca<sup>+</sup>,
ANN-R-HCO<sub>3</sub>, ANN-R-Cl-, ANN-R-SiO<sub>2</sub>, ANN-R-TDS,
ANN-R-pH, ANN-R-EC). Percentage of contribution and
R<sup>2</sup> was used for model performance evaluation criterion.
The CA allowed the formation of two clusters between
the sampling wells. The Low Contaminant Level as LCL
and moderate contaminant level as MCL reflecting
differences on water quality at different locations. DA as
a data reduction techniques was used to evaluate the
spatial variability in water quality as it uses 6 parameters
(SO<sub>4</sub><sup>-</sup>,Cl<sup>-</sup>, As, Mn, NO<sub>2</sub> and total dissolved solid)
affording 90.00% correct assignation to discriminate
between the clusters using forward stepwise mode from
the original 24 parameters. The sensitivity analysis
reveals that Na<sup>+</sup>, HCO<sub>3</sub>, SiO<sub>2</sub> and EC are the four most
effective parameters for discriminating groundwater
quality regions with a percentage of contribution of 17.49,
17.50, 17.57 and 17.46%, respectively. This study reveals
the significance of sensitivity analysis and multivariate
techniques for the use of less parameter for understanding
the most effective pollutant in water resource
management, since, its time and cost consuming.
ER  -