2005 Salt Lake City Annual Meeting (October 16–19, 2005)

Paper No. 8
Presentation Time: 3:45 PM

USING MULTIVARIABLE ANALYSIS AND GEOCHEMICAL MODELLING TO ASSESS INTERACTION OF GROUNDWATER AND SURFACE WATER IN SEMI ARID AREA


ACHOUR, Farid, ENVIRON International Corporation, 2010 Main Street, Suite 900, irvine, CA 92614 and MONTEIRO, J. Paulo, CVRM, University of Algarve, Campus de Gambelas, Faro, 8000, Portugal, fachour@environcorp.com

The Chelif Basin is located in western Algeria, North Africa, between the Dahra and Ouarsenis mountain ranges. Despite its proximity to the Mediterranean Sea, the Chelif Basin has a semi arid climate. The Chelif River considered to be the most important river in Algeria, more than 700 km long, flowing from its source in the Atlas Mountains through the Chelif Basin and discharging into Mediterranean Sea. The Chelif Plain is underlain by thick silty sand formations in the central and northern parts of the plain, and by calcareous formations in the southern part. A water resources vulnerability study was carried out in the plain. However, the question of the relationship between the river and the aquifer was not investigated as part of that study. The aim of this investigation was to define the connections between the river and groundwater in the alluvial sediment. Water samples were collected within the plain from 124 wells, and 37 water samples were taken from the River. Principal Component Analysis (PCA) was used to determine the different water types present within the dataset and thereby allowed the sources to be identified. Several successive factor analyses were necessary to select the appropriate samples and variables for extracting the hydrogeologic characteristics. Discriminant Factor Analysis (DFA) was applied to validate the results obtained by PCA by adding a qualitative variable to the analysis of a particular group, whereby the probability that given sample belongs to the specified group was calculated. Geochemical modelling was then used to compute the fictitious partial pressure of CO2 within the sampled waters; the resulting pCO2 should be low for surface water and high for groundwater. The values obtained from the study showed that most of the water samples obtained from the river have a local groundwater origin, confirming the results obtained using the multivariable analysis.