Jm. Andrade et al., SELECTION OF ANALYTICAL VARIABLES TO OPTIMIZE LABORATORY EFFORTS IN FUTURE GROUNDWATER STUDIES, Analytica chimica acta, 292(3), 1994, pp. 253-261
The main objective of this work was to develop a way to select a subse
t of analytical variables from one initial broad study to monitor grou
ndwater quality. The study was applied to one Spanish area. The work w
as focused on selecting actual physico-chemical variables rather than
using typical multivariate techniques as a way to reduce costs, effort
, necessities facility and staff workloads. Several chemometric multiv
ariate techniques such as principal components analysis, factor analys
is, singular value decomposition of raw data matrix and procrustes rot
ation were used as the mathematical means to achieve the objectives. O
f the initial 29 variables analysed, six of them were selected to desc
ribe the system during five seasonal samplings and it is hoped that ea
ch season subset will allow the monitoring of groundwater quality duri
ng future seasons.