Surface nuclear magnetic resonance (surface NMR)

Among the hydrogeophysical methods, the NMR applications (Fig. 1) have a unique benefit: water in the subsurface is detected directly by investigating the properties of the water molecules, in particular the protons in the water molecules, at the atomic scale. All other geophysical methods allow only indirect prospection of water by measuring physical quantities such as electrical conductivity or permittivity. These quantities depend on water content, but also on other properties of the rock, which increases the degree of ambiguity and thus the risk of misinterpretation. Using hydrogeophysical NMR applications, solely the proton spins of the water molecules are excited.

Figure 1: NMR-Measurement in the Döberitzer Heide Source: BGR

The first geophysical NMR methods were developed in the 1980s for the application in boreholes to explore hydrocarbon deposits (Dunn et al., 2002). Some years later the development of a non-invasive NMR-based method, the so called surface NMR, for prospection of groundwater began. The principle feasibility of the application was shown in Russia (Semenov et al., 1989). Since then, international research activities have started to continuously improve and extend the method. The simplest, one-dimensional application of surface NMR is also called magnetic resonance sounding (MRS). Similar to the TEM method, a cable loop is laid out on the surface in the field used as both transmitter and receiver (Fig. 2). A pulse-shaped alternating current is applied at the transmitter at the Larmorfrequency of the protons. The resulting electromagnetic (EM) field in the subsurface excites the proton spins and forces them to flip away from their position of equilibrium in the Earth’s magnetic field. After switching off the excitation, the spins move back to equilibrium and in doing so they induce a measurable voltage in the receiver loop. The measured signal is an exponentially decaying (relaxing) curve. Its initial amplitude is proportional to the number of excited protons and thus to the water content within the sensitive volume, while the relaxation behavior, represented by the relaxation time, is related to the mean pore size. By varying the current of the pulse, different penetration depths are can be reached.

Fig. 2: Measurement principle of surface NMR (Lange et al., 2007) Source: BGR

The maximum penetration depth corresponds approximately to the diameter of the surface loop, normally 10 to 150 m. However, the depth that can actually be reached also depends on the electrical conductivity of the subsurface. To invert surface NMR data correctly, it is thus necessary to conduct additional electrical or electromagnetic measurement such as geoelectrics or TEM. Normally, also magnetic measurements are combined with surface NMR, because the Larmor frequency, which must be adjusted to perform correct NMR measurements, depends on the strength of the Earth’s magnetic field. When applying surface NMR measurements in practice, additional surface loops are used to characterize the surrounding EM noise and to improve the data processing by optimized filtering.

Fig. 3: Schematic demonstrating the measurement layout of surface NMR in the field Source: BGR

Figure 4 demonstrates a surface NMR data example. The measured data after some processing (Fig. 4a) is approximated by inversion (Abb. 4b). The corresponding model consists of the water content and relaxation time distributions as functions of depth (Fig. 4c). The water content measured by surface NMR is an estimation of the effective porosity (Legchenko et al., 2004). Capillary bound water in small pores is associated with short relaxation times – too short to be measured by surface NMR due to the long dead times, i.e., the time necessary for switching from transmitting to receiving mode. Thus, the water content in our example is underestimated at the depth ranges corresponding to silty layers, while the effective porosity estimates for the sand aquifers are within a reliable range from 25 to 35 %. If calibration data is at hand, the relaxation times being a proxy for the mean pore size also allow the estimation of the hydraulic conductivity distribution as function of depth (Legchenko et al., 2004; Mohnke and Yaramanci, 2008).

Fig. 4: Data example from BGR testfield Barnewitz/Nauen Source: BGR

Many studies have shown the beneficial usage of surface NMR for various hydrogeological issues such as characterization of aquifers in sedimentary rocks (e.g. Mohnke und Yaramanci, 2008; Günther und Müller-Petke, 2012; Knight et al., 2012) as well as in fractured rock and karst aquifers (e.g. Vouillamoz et al., 2005; Vouillamoz et al., 2014; Girard et al., 2012). Moreover, the ongoing technical and methodological improvement of the method opens new and expands existing application fields. For instance, the decrease of instrumental dead-time meanwhile allows the investigation of the vadose zone (Walsh et al., 2014; Costabel and Günther, 2014). Novel pulse sequences improve the resolution and provide reliable results even in magnetically heterogeneous environments (Legchenko et al., 2010; Walbrecker et al., 2011; Grunewald et al., 2014; Grombacher et al., 2016). Strategies for increasing the measurement progress will improve the future applicability of surface NMR (Costabel et al., 2016; Grunewald et al., 2016). It is expected that the acceptance of surface NMR within hydrogeology will be increasing further.

Selected publications on the SNMR-method:

• References

Projects:

• Quantitative exploration of aquifer systems – Improvements of the surface nuclear magnetic resonance application
• Development of noninvasive methods for soil physical and hydraulic characterization of the vadose zone
• FLIN- Freshwater Lens Investigations