Near Surface Water Content Measurement And Contamination Detection Using Ground-Penetrating Radar-A Simulation Study

ecent years, great progresses are made in the research of near surface water content. More and more electromagnetic methods, such as ground-penetrating radar (GPR), have been widely used in hydrological problems of the near surface zone. GPR technique is an indirect method to measure water content, changes in velocity and dielectric constant can be converted into volumetric water content. GPR makes use of high frequency EM wave as exploration source, two physical parameters effect the behavior of the EM wave are conductivityσand dielectric constantε, conductivity dominants the attenuation of EM wave, and dielectric constant determines the velocity, and the dielectric constant is mostly determined by water content of materials. Therefore, GPR is an effective and fast-taking geophysical method to measure near surface water content. In this paper, we engage in bridging the velocity parameters that determined by GPR common midpoint measurement and the models for mixing materials which represent the relationship between dielectric constant and volumetric water content, further, the near surface water content information can be gained, and the forward modeling analysis achieve promising results.Conventional water content measurement methods at small scales include: gravimetric, capacitance or resistivity probe, tensiometer, frequency- and time-domain reflectometry (FDR and TDR) and nuclear technology. The greatest limitation of these techniques is that they are not capable of capturing information about field-scale moisture content variations; all of these are invasive and destroy the structure of soil layer. Remote sensing is the most promising technique for measuring water content variations over large regions, with the spatial resolution of meter. Compared with the conventional methods, GPR is an intermediate-scale technique that could accurately measure water content over large areas. The main advantages are larger sampling volume, faster measurement and higher in resolution.There exists four methodologies of GPR to measure near surface water content, that are: water content form reflected wave velocity, water content form ground wave velocity, water content form transmitted wave between boreholes, and water content form surface reflection coefficient. Due to the widely applications in different geological conditions, and larger exploration depth, the reflected wave method is most commonly applied. In this paper, we fully discuss the principium and characteristics of the GPR reflected wave method.Under low-loss medium assumption, the interval velocity is largely related to the real part of the dielectric constant. In order to obtain the dielectric constant distribution of the subsurface materials, GPR reflect method needs to accurately determine the velocity that the EM wave transmitted through the subsurface. The determination of EM wave velocity is based on velocity analysis theory. To obtain the precise EM wave velocity, firstly, raw common-midpoint profile, after the process of wavelet denoise, and auto gain control et al, the reflect waves of the interfaces are more clear and easy to identify, then, the velocity information is easy to pick up in the stacked velocity spectrum. Under the condition of approximate horizontal layered materials, the stack velocity is equal to root mean square velocity, which can be converted to interval velocity using the Dix formula. Therefore the 2D velocity profile of the subsurface materials can be created, and the dielectric constant distribution can be further gained.Four models for mixed materials are introduced to indicate the petrophysical relationship between the dielectric constant and volumetric water content: Topp formula, complex refractive index model (CRIM), Hanai-Bruggman model and self-similar model, and the first model is most commonly used. Using the models for mixed materials, along with the velocity parameters obtained from the result of velocity analysis, the distribution and variations of near surface water content can be achieved. Using GPR technique to mapping and locating the organic contamination detection is a topic of active research these years, petroleum and its chemical engineering products often pollute the soil, aquifer and ground water in the form of light nonaqueous phase liquids (LNAPLs). Since the distribution and transfer mechanism of the subsurface organic contaminations are extremely complex, we give a brief analysis and discussion of the states and transfer mechanism of LNAPLs.GPR numerical modeling is an effective way of analyzing the detecting problems and researching the propagation rule of EM wave in the subsurface materials. We use the forward modeling software GprMax2.0 developed by Edinburgh University, England, to model GPR response under different measurement acquisitions. The models being analyzed include: local aquiferous bodies model, two-layer horizontal model, multiple-layer horizontal model, groundwater table model, local aquifer model, and contamination model.Multiple-layers horizontal model consists of six horizontal layers, made up of sandstone of different dielectric constant. From the result of the velocity analysis, the designed velocity spectrum arithmetic is easy and reliable, and the velocity spectrum does not disturb by the multiples. After the process of wavelet denoise and linear gain, the stacked velocity and average amplitude energy can be pick-up from the velocity spectrum profile. We can conclude from the results of four models of mixed material that the Hanai-Bruggman model with m=2.0 reaches the highest value, and closes to the result of Topp formula which accords with the relative theory; and the value of Hanai-Bruggman model with m=1.5 closes to self-similar model.Groundwater table is the external represent of the phreatic water, we must concern the effect of the groundwater table for most of the environmental problems. From the analysis of the modeling result we can conclude: from the common-offset measurement we can acquire the variations of groundwater depth in a short time period, and from the multiple-offset data, after velocity analysis, accurate velocity of the interfaces can pick-up. The hydrological information can be effectively reflected with the combination of common-offset and multiple-offset measurement.The perched aquifer often develops in the upper layer of loam lens in the vadose zone. Because of its lower depth of burial, the perched aquifer is closely related to the ground surface. We can conclude from the modeling result that common-offset profile can reflect each part of the lens clearly, and image is higher in resolution. Through the calculation of ten midpoints, the interfaces of the local aquifer fitting profile are rather clear and easy to identify. Therefore, the velocity distribution of the lens area can accurately be calculated, combined with the models of mixed material, the 2D water content profile can be created, the water content information of the lens area is then obtained.The distribution condition and migration mechanism of the LNAPLs in the subsurface is rather complex, so in forward modeling, we make certain simplification. Common-offset profile can clearly reflect each phase of the contamination, and the radar wave appears considerate attenuation through the top of the contamination, forward modeling results are promising.Synthetically analyzing the numerical modeling results: GPR technique is an effective geophysical method in the use of measuring near surface water content, and deserves to popularize. The designed velocity spectrum is simple and reliable, which can be used to accurately pick up the stacked velocity of the interfaces. The inverse modeling results of dielectric constant and volumetric water content are promising.