Magnetic resonance sounding: new method for ground water assessment.

Abstract

The advantage of magnetic resonance sounding (MRS) as compared to other classical geophysical methods is in its water selective approach and reduced ambiguity in determination of subsurface free water content and hydraulic properties of the media due to the nuclear magnetic resonance (NMR) principle applied. Two case examples are used to explain how hydrogeological parameters are obtained from an MRS survey. The first case example in Delft (the Netherlands) is a multiaquifer system characterized by large signal to noise ratio (S/N = 73), with a 24 m thick, shallow sand aquifer, confined by a 15 m thick clay layer. For the shallow aquifer, a very good match between MRS and borehole data was obtained with regard to effective porosity n(e) approximately 28% and specific drainage S(d) approximately 20%. The MRS interpretation at the level deeper than 39 m was disturbed by signal attenuation in the low resistivity (approximately 10 omega(m)) media. The second case of Serowe (Botswana) shows a fractured sandstone aquifer where hydrogeological parameters are well defined at depth > 74 m below ground surface despite quite a low S/N = 0.9 ratio, thanks to the negligible signal attenuation in the resistive environment. Finally, capabilities and limitations of the MRS technology are reviewed and discussed. MRS can contribute to subsurface hydrostratigraphy description, hydrogeological system parameterization, and improvement of well siting. The main limitations are survey dependence upon the value of the S/N ratio, signal attenuation in electrically conductive environments, nonuniformity of magnetic field, and some instrumental limitations. At locations sufficiently resistive to disregard the signal attenuation problems, the MRS S/N ratio determines how successfully MRS data can be acquired. Both signal and noise vary spatially; therefore, world scale maps providing guidelines on spatial variability of signal and noise are presented and their importance with respect to the MRS survey results is discussed. The noise varies also temporally; therefore, its diurnal and seasonal variability impact upon the MRS survey is covered as well.