Quasi-2D block inversion of large-scale surface nuclear magnetic resonance profile data using a laterally constrained model

Abstract

The water content of underground structures can be imaged using surface nuclear magnetic resonance (NMR) in one dimension, two dimensions, or recently, three dimensions. The maximum efficiency can be obtained with a coincident loop configuration in a large-scale area, although the resolution is limited. We have developed a quasi-2D block inversion to characterize a horizontal continuous underground aquifer using a laterally constrained model to improve the resolution of thin aquifers and deep aquifers. Through the simulation of flat surface and irregular surface models, a horizontally layered aquifer with continuously variable water content and [Formula: see text] can be imaged through quasi-2D block inversion, which is more optimized than the traditional 1D block or 1D or 2D smooth inversion methods. Finally, using a series of profile data obtained from surface NMR field measurements in Mongolia, we determine that quasi-2D block inversion based on a laterally constrained model can significantly improve the discontinuity and heterogeneity caused by splicing images in a 1D block inversion and can provide a reliable result of the distribution of continuous aquifers for hydrogeology surveys over a large-scale area.