Understanding factors influencing groundwater levels in hard-rock aquifer systems by using multivariate statistical techniques

Abstract

Spatial and temporal dynamics of groundwater levels provides vital information required for management of fast depleting groundwater resources in hard-rock aquifer systems. This study demonstrates application of multivariate statistical techniques to analyze spatial and temporal variations of a 39-month period (May 2006–July 2009) monthly groundwater levels at 50 monitoring sites and to understand principal factors most influencing the groundwater system in Ahar River catchment of Udaipur district, Rajasthan, India. Box-whisker plots drawn for mean monthly groundwater levels revealed that spatial variation of the groundwater levels was less during rainy season in comparison to that during dry season. The groundwater levels in the aquifer system were found to be largely influenced by rainfall occurrences in the area. Firstly, hierarchical cluster analysis technique was applied to classify 50 monitoring sites into different clusters according to behaviour of the groundwater levels. This resulted into four clusters of the groundwater levels at less than 22 linkage distance. The most (25.29 m) and the least (6.48 m) spatial variability of the groundwater levels were observed for clusters III and I, respectively. Furthermore, principal component analysis (PCA) technique was utilized to understand and identify the most significant variables influencing the groundwater levels in each of the four clusters of the monitoring sites. The first two principal components (PCs) explained 43–55 % of the total variance. Based on the PCA, the significant PCs for clusters I and II were characterized as ‘topography factor’. On the other side, the significant PCs for clusters III and IV were termed as ‘geomorphologic’ and ‘land use’ factors, respectively.