Premium
Combined Use of Hydraulic and Electrical Properties of an Aquifer in a Geostatistical Estimation of Transmissivity
Author(s) -
Ahmed Shakeel,
Marsily Ghislain,
Talbot Alain
Publication year - 1988
Publication title -
groundwater
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 0.84
H-Index - 94
eISSN - 1745-6584
pISSN - 0017-467X
DOI - 10.1111/j.1745-6584.1988.tb00370.x
Subject(s) - kriging , aquifer , soil science , aquifer properties , geostatistics , multivariate statistics , permeability (electromagnetism) , estimation theory , statistics , mathematics , geology , spatial variability , groundwater , geotechnical engineering , groundwater recharge , membrane , biology , genetics
Many previous attempts have been made to establish an empirical relation between the electrical and hydraulic properties of aquifers. However, only regression models between transmissivity or permeability and a few electrical parameters have been used on the basis of the available pairs of data. Kriging, a geostatistical technique, estimates a regionalized variable at any point in space, and multivariate geostatistical techniques allow one to use several variables together to estimate any ‘spatial parameter. One such method, cokriging, is used to estimate the transmissivity based not only on measurements of transmissivity, but also on measurements of specific capacity and electrical transverse resistance. The studied aquifer is situated in the Medjerda Valley in Tunisia where very few data on transmissivity and specific capacity are available, but resistivity data are relatively abundant. It is shown that with the geostatistical technique, one can: (1) use several electrical or elastic properties, which are easily measured, in the estimation of the desired parameter without establishing any empirical relation; and (2) make the estimation at any point where none of these properties has been sampled and, at the same time obtain a variance of the estimation error. The method is also compared with the usual regression method.