The technique of surface nuclear magnetic resonance (surface NMR) is the only geophysical exploration method providing direct and nondestructive information on subsurface aquifer properties due to the method’s unique sensitivity to hydrogen protons. The method combines the information content accessible via nuclear magnetic resonance (NMR) measurements and the nondestructive approach to derive subsurface information from surface-based measurements. Because of this, surface NMR became a useful tool for hydrogeophysics during the last decade. Two different inversion schemes exist. The initial value inversion (IVI) extracts the water content distribution from the surface NMR information content by estimating a sounding curve from surface NMR data. The time step inversion (TSI) extracts the distribution of both water content and decay time by separating the surface NMR data into several time steps. Both solve the inverse problem using independent steps and by separating subdata sets from the complete data. In this paper, a new inversion scheme — the QT inversion (QTI) — is found to solve to inverse problems by taking the complete surface NMR data set into account at once. QTI extracts water content and decay time and satisfies the complete data set jointly. We examine and compare QTI to IVI and TSI by a synthetic data set and a field data set. Our results find that the QT inversion approach increases both spatial resolution of the subsurface decay time distribution and stability of the inverse problem.

The technique of magnetic resonance sounding (MRS) has shown several improvements in data processing, inversion and interpretation during the last years. Along with these improvements, detailed innovations on instrumentation have been demanded to support their use. Latest developments in surface nuclear magnetic resonance (NMR) instrumentation promise to fulfil these hardware requirements such as decreased dead time, improved digital signal detection, multi‐channel capabilities and improved reference techniques with the second generation surface NMR instruments. In this paper, we compare data from two generations of instruments and assess the impact of the improvements on practical issues, i.e., the increased accuracy of data due to shorter dead times and new noise reduction approaches and the feasibility for efficient 2D measuring schemes. Well‐known and documented test sites and synthetic considerations are used to evaluate these developments. First, the relaxation signals of different devices using the same loop match each other. The inversion results coincide within the range of data errors. Decay time estimation appears to be more stable for the new generation instrument. Second, the potential of shorter effective dead times (considering a relaxation of the protons during the pulse) is investigated using statistical analysis of synthetic data sets with different decay times and noise levels. The additionally measured data at early times significantly improve the scope and accuracy of the determined parameters initial amplitude and T2* time and thus extend the range of formations to be characterized. A field example comparing an effective dead time of 18 ms and 45 ms is presented. Two different reference techniques were successfully applied for noise cancellation at the very noisy test site Nauen. We observed an equivalent signal improvement using the software‐based and hardware‐based technique. However, software noise cancellation approaches are easily adaptable and extendable. Finally, considerations are given how to efficiently carry out 2D surveys using multi‐channel instruments. A 2D field data set using the GMR demonstrates that 2D surveys can already be realized in moderate measuring times. The new generation of instruments provides comparable results and improved capabilities that will enable surface NMR measurements to be applied in a wider range of applications.

No abstract

No abstract