© 2013 European Association of Geoscientists & Engineers 143 * ferrara@fis.uniroma3.it Near Surface Geophysics, 2013, 11, 143-153 doi: 10.3997/1873-0604.2012051 Comparison of GPR and unilateral NMR for water content measurements in a laboratory scale experiment C. Ferrara 1 *, V. Di Tullio 2,3 , P.M. Barone 1 , E. Mattei 1 , S.E. Lauro 1 , N. Proietti 2 , D. Capitani 2 and E. Pettinelli 1 1 Department of Physics “E. Amaldi”, University of Roma Tre, Via della Vasca Navale, 84, 00146 Rome, Italy 2 Magnetic Resonance Laboratory “Annalaura Segre”, Institute of Chemical Methodologies, CNR, Research Area of Rome, Via Salaria km 29.300, 00015 Monterotondo, Rome, Italy 3 Department of Earth Sciences, Sapienza University of Rome, Piazzale Aldo Moro 5, 00185 Rome, Italy Received July 2011, revision accepted October 2012 ABSTRACT Several factors affect antenna-soil coupling in a Ground Penetrating Radar (GPR) survey, like sur- face roughness, lithology, lateral heterogeneities, vegetation, antenna height from the surface and water content. Among them, lithology and water content have a direct effect on the bulk electromag- netic properties of the material under investigation. It has been recently pointed out that the wavelet of the early-time portion of a radar signal is correlated to the shallow subsurface dielectric proper- ties of a material. This result indicates that some information on such properties can be directly extracted from the analysis of GPR early-time traces. In the present paper, we use the early-time GPR signal, in terms of average envelope amplitude computed on the first half-cycle, to map the near-surface (few centimetres) lateral distribution of dielectric parameters, induced by changing the shallow water content on a concrete slab. This con- trolled experiment was specifically designed to study the effect of water content variations on antenna-material coupling, minimizing the influence of both surface roughness and heterogeneity. The quantitative control of the water in the shallow portion of the slab is performed by using a port- able unilateral Nuclear Magnetic Resonance (NMR) sensor, which is able to determine the water content in the material on the basis of the measured proton density. The results show a matching pattern of the physical parameters measured with the two different techniques and a very high degree of linear correlation (r = 0.97) between the radar early-time signal average amplitude and the intensity of the NMR signal, which is proportional to the proton density, i.e., to the water content. This experiment suggests that the early-time approach could be used as a fast and high- spatial resolution tool for qualitatively mapping water content lateral variations in a porous material at shallow depth, using a ground-coupled single-offset antenna configuration and that a quantitative evaluation of the moisture content would require a calibration procedure. In a survey design several factors should be taken into consid- eration, like the investigation depth, the spatial resolution, the characteristics of the medium (e.g., granular or solid), the physi- cal properties of the material, the site conditions and the reliabil- ity of the retrieved physical parameter in terms of water content estimator. If the required investigation depth is limited to a few metres, GPR represents one of the best options in terms of spatial resolution, fast acquisition time, extension of the investigated area and repeatability of the measurements (Annan 2004; Jol 2009; Barone et al. 2010). Such a technique is based on radio waves propagating through the medium and the water content value can be indirectly retrieved from the measurement of a sig- nal velocity. In particular, several radar methodologies like the INTRODUCTION Assessing water content in porous materials by means of non- destructive methods is of great importance for many applied sciences like hydrology, environmental physics, civil engineer- ing, agriculture and cultural heritage protection. Several geo- physical techniques, particularly those based on the interaction between electromagnetic (EM) fields and matter, can be success- fully applied to indirectly estimate water content in a material (Rubin and Hubbard 2005). The choice of a specific method depends on the goal of the research and the characteristics of the site (or of the material under investigation).