Detecting soil hydrological connectivity in a badland area by fast field cycling nuclear magnetic resonance relaxometry

Abstract

AbstractThe ‘hydrological connectivity inside the soil’ refers to both the spatial pattern inside the soil (structural component) and the physical–chemical process at a molecular level (functional component). Fast field cycling (FFC) nuclear magnetic resonance (NMR) relaxometry allows for measuring structural and functional connectivity by two suitable indexes named structural connectivity index (SCI) and functional connectivity index (FCI). In this study, FFC NMR relaxometry was applied to soils sampled in a very degraded environment (i.e., a badland area) to detect the capability of the measurement technique to distinguish the hydrological connectivity of these samples having different conditions (layer explored by roots, sparsely‐vegetated and bare soil). The relaxograms measured by the FFC NMR, using Proton Larmor frequencies in the range 0.01–10 MHz, were integrated and the resulting S‐shaped curves were analysed to obtain the connectivity indexes. Results showed that the ‘Sparsely vegetated’ sample is characterized by more small‐sized pores than the ‘Rooted’ one. The comparison between the ‘Sparsely vegetated’ and ‘Bare’ conditions pointed out that the presence of vegetation reduces the measured relaxation times and, as a consequence, the corresponding pore sizes and modifies the structural connectivity. The analysis also revealed that the three samples are characterized by similar values of SCI, which are independent of the proton Larmor frequency, while the FCI values of the ‘Bare’ soil are the lowest. Conversely, samples from soil with vegetation (‘Rooted’ and ‘Sparsely vegetated’) present comparable functional connectivity. Finally, the analysis of the frequency distribution of the ratio of each connectivity index and its mean value (SCI/m(SCI) and FCI/m(FCI)) allowed to establish its normal distribution. For the investigated samples, this result established that FCI and SCI can be represented by their mean value.