Surface-Nuclear Magnetic Resonance (NMR) in an area with low geomagnetic field and low water content – a case history from Namibia

Map showing global anomalies of the Earth magnetic field Source: BGR

Surface Nuclear Magnetic Resonance (SNMR) is a new method for groundwater exploration. Not only can the distribution of groundwater be determined, but also water content, porosity and hydraulic conductivity of the rocks. NMR is already routinely used in borehole geophysics and laboratory investigations. In contrast to borehole NMR, in which a strong magnetic field is produced, SNMR uses the weak magnetic field of the Earth. The nanovolt range signal is approximately proportional to the square of the total intensity of the geomagnetic field, which varies considerably at different parts of the globe. Therefore, the geomagnetic field at the sounding site must be taken into consideration. The field intensity in Namibia is about 60 % of that in Europe. When the hydrogeological conditions are the same, the signal strength is about 35 % of that measured in Europe only. The low porosity of the aquifers causes additional problems. Thus, previous experience in Europe with SNMR is of limited use in southern Africa. In order to properly apply the method, the hydrogeological conditions at the survey site were modelled. At the Omaruru 5 site good results were obtained by using extremely high stacking numbers (n = 250). The site is characterised by fluvial sands, 20 m thick above fractured bedrock. The SNMR result is in very good agreement with vertical electrical sounding (VES) data, but gives the water content in the vadose zone and aquifer directly, even though it is low. Decay times indicate medium-grained sand, i.e., the water is abstractable.

At the Omdel 2 site the VES data suggest an aquifer with very low resistivity down to about 27 m. However, this is in contradiction to the SNMR results, which show mobile water only below this depth. The low resistivity depth range is caused by the high salinity of the water. This water is adsorbed and, therefore, does not show up in SNMR, and the high salinity is due to salts dissolved from the matrix by seeping water. The higher resistivity below this depth indicates the presence of groundwater. This is unambiguously recognised by SNMR, even though the water content is unusually low. The SNMR result is verified by information from a nearby borehole. The decay times at this site show some interesting features. The high measured values of the very shallow range are not reliable because there is no mobile water. The aquifer has decay times around 80 ms, corresponding to fine-grained sand, indicating a slightly higher proportion of adsorbed water and a low proportion of mobile water, explaining the low water content derived from the amplitudes.

At this location water is artificially recharged. The regional extent of the aquifer can be found only by using SNMR. VES data alone would give an incorrect water distribution, owing to the ambiguity in geoelectrical methods. This site is a good example that only by using SNMR is a reliable estimate of the stored amount of water possible.

SNMR and VES at OMARURU test site. Source: BGR

SNMR and VES at OMDEL test site Source: BGR