Search

Geophysical methods are routinely applied for investigation of near surface in areas of infrastructure, water supply, artificial infiltration, farming, waste deposits, construction, etc. A new towed transient electromagnetic tool, called ‘tTEM’, provides rapid, efficient, high‐resolution imaging of subsurface hydrogeology, and can deliver densely spaced profiles of resistivity. These profiles can be used to map a three‐dimensional subsurface in high resolution. In this paper, we present three case studies where the towed transient electromagnetic system was used to map the subsurface at the hectare scale. In the first case, we used towed transient electromagnetic to map raw materials in the northern Jutland, Denmark. The survey was carried out to map possible sand and gravel deposits. In the towed transient electromagnetic models, the potential sand/gravel targets are identified as high resistive bodies in the top 30 m. These bodies can have significant lateral variation at scales much less than one hundred metres. In the second case study, towed transient electromagnetic was used to map the thickness of a protecting clay layer above an aquifer in Vildbjerg, a town in the central part of Denmark. Results show that the overlying clay layer has sufficient thickness (>15 m) to protect the underlying aquifer from pesticide pollution in the area. Finally, in the third case study, we used towed transient electromagnetic for mapping geology in the vicinity of a landfill. The inversion results reveal a hitherto unknown buried valley‐like feature within the top 30 m of the subsurface that was not identified by older, regional TEM surveys – a feature that can have significant impact on the water flow.

Paleochannels typically act as pathways for groundwater movement and provide a potential source of groundwater. Their presence can be helpful in identifying areas suitable for recharge and at times in mitigating contamination problems in afflicted regions. Thus, mapping of paleochannels is significant in the planning and management of groundwater resources. An airborne electromagnetic (AEM) system employing dual pulse moments has been used extensively for this purpose in India. This paper presents the results over paleochannels defined in three different terranes. In northwest India, a 100 m wide by 80 m deep paleochannel within alluvium overlaying a Proterozoic basement illustrates the impact of neotectonic disturbances in changing the river course. In northeast India's Ganga Plains, a paleochannel is mapped that provides insight into managing groundwater resources of areas polluted with arsenic. In south India, a paleochannel buried under similar to 100 m thick sequence of coastal sediments is imaged with implications on submarine groundwater discharge.

SUMMARY A surface nuclear magnetic resonance (NMR) forward model based on the full-Bloch equation improves the accuracy of the forward response given an arbitrary excitation pulse and a wider range of relaxation conditions. However, the full-Bloch solution imposes a significant slowdown in inversion times compared to the traditional forward model. We present a fast-mapping approach capable of dramatic increases in inversion speeds with minimal sacrifices in forward response accuracy. We show that the look-up tables used to calculate the transverse magnetization and the full surface NMR forward response are smoothly varying functions of the underlying T2* and T2 values. We exploit this smoothness to form a polynomial representation of the look-up tables and surface NMR forward responses, where a fast-mapping approximation of each are reduced to a simple matrix multiplication. Accurate approximations with less than 1 per cent error can be produced using 21 coefficient representations of the look-up tables for each B1 value and for the signal expected from a particular depth layer for a particular pulse moment. In essence, the proposed fast-mapping approach front-loads all expensive calculations and stores the results in a compressed form as a coefficient matrix containing less than a half a million elements. This allows all subsequent inversions to be performed at greatly improved speeds.

The hydrogeology below large surface water features such as rivers and estuaries is universally under-informed at the long reach to basin scales (tens of km+). This challenge inhibits the accurate modeling of fresh/saline groundwater interfaces and groundwater/surface water exchange patterns at management-relevant spatial extents. Here we introduce a towed, floating transient electromagnetic (TEM) system (i.e. FloaTEM) for rapid (up to 15 km/h) high resolution electrical mapping of the subsurface below large water bodies to depths often a factor of 10 greater than other towed instruments. The novel FloaTEM system is demonstrated at a range of diverse 4th through 6th-order riverine settings across the United States including 1) the Farmington River, near Hartford, Connecticut; 2) the Upper Delaware River near Barryville, New York; 3) the Tallahatchie River near Shellmound, Mississippi; and, 4) the Eel River estuary, on Cape Cod, near Falmouth, Massachusetts. Airborne frequency-domain electromagnetic and land-based towed TEM data are also compared at the Tallahatchie River site, and streambed geologic scenarios are explored with forward modeling. A range of geologic structures and pore water salinity interfaces were identified. Process-based interpretation of the case study data indicated FloaTEM can resolve varied sediment-water interface materials, such as the accumulation of fines at the bottom of a reservoir and permeable sand/gravel riverbed sediments that focus groundwater discharge. Bedrock layers were mapped at several sites, and aquifer confining units were defined at comparable resolution to airborne methods. Terrestrial fresh groundwater discharge with flowpaths extending hundreds of meters from shore was also imaged below the Eel River estuary, improving on previous hydrogeological characterizations of that nutrient-rich coastal exchange zone. In summary, the novel FloaTEM system fills a critical gap in our ability to characterize the hydrogeology below surface water features and will support more accurate prediction of groundwater/surface water exchange dynamics and fresh-saline groundwater interfaces.

Summary 3D transient electromagnetic (TEM) modelling and inversion has been continuously developed in the last decade, together with the undergoing rapid developments in computational resources. In this work we present an efficient and practical finite element 3D implementation, applicable to both airborne and ground-based TEM, developed with an octree-based hexahedral meshing that implements topography through a mesh deformation. This new implementation is compared to both 1D modelling and 3D finite-element tetrahedral modelling, in terms of accuracy, speed and memory requirement and inversion results on synthetic modelling. Furthermore, a verification of the topography implementation obtained comparing results with and without deformation is presented. Overall, the octree-based implementation is ~5x faster than tetrahedral simulation, while the memory cost in the forward modelling process is ~6x smaller.

Permafrost is ubiquitous at high latitudes, and its thickness is controlled by important local factors like geothermal flux, ground surface temperature and thermal properties of the subsurface. We use airborne transient electromagnetic resistivity measurements to determine permafrost thickness on the coast of Ross Island, Antarctica, which contains the active volcano Mt Erebus. Here, resistivity data clearly distinguish resistive permafrost from the electrically conductive fluid-saturated materials underlying it. For our study, we define permafrost as frozen material with a resistivity > 100 Ω·m; more conductive material contains a significant fraction of water or (more likely) brine. We observe that permafrost is very thin near the coast and thickens within several hundred metres inland to reach depths that are typically within the range of 300–400 m. We attribute the sharp near-shore increase in permafrost thickness to lateral heat conduction from the relatively warm ocean, possibly combined with seawater infiltration into the near-shore permafrost. We validate this result with a two-dimensional heat flow model and conclude that away from the thermal influence of the ocean, the local geothermal gradient and heat flux are about 45 ± 5 °C/km and 90 ± 13 mW/m2, respectively. These values are in line with published estimates in the vicinity of Mt Erebus and within the actively extending Terror Rift, but do not reflect a strong heat flow anomaly from volcanic activity of Mt Erebus. Measurements made previously in the McMurdo Dry Valleys, on the other side of McMurdo Sound, tend to be a few dozens of mW/m2 lower, likely reflecting its different tectonic setting on the uplifted rift shoulder of Transantarctic Mountains. Our study demonstrates a new approach towards constraining geothermal flux in polar regions using airborne electromagnetic (AEM) data that can be relatively efficiently collected on regional scales where ice coverage does not exceed the penetration limits of the AEM device, which for the device used is ∼ 500 m under the favourable conditions in the study area.

We describe a new approach for modelling airborne transient electromagnetic (TEM) data which combines the use of on- and off-time data for inversion. Specifically, the response is modelled using system response convolution both during and after transmitter ramp-down. High near-surface sensitivity can be achieved through a combination of fast transmitter ramp-down, broad receiver system bandwidth, efficient suppression or explanation of the primary field, and by combining the use of on-time gates with accurate knowledge of the system response. The system response can either be calculated based on the transfer function of the individual system components (i.e. receiver coil, amplifiers, low-pass filters and current waveform) or it can be measured at high altitude. The latter approach has the advantage of avoiding the specific modelling of individual system components. By comparing model parameter uncertainty when the on-time gates are included in the inversion versus when they are not, we show that a significant improvement in near-surface sensitivity is obtained. The method is used to invert both synthetic and field data. In the inversion of synthetic data, we see clear improvements in the determination of thin shallow layers, especially when they are resistive. This is confirmed by inversion of field data where we observe more pronounced structures with better definition of layer boundaries and layer resistivities.

Soil contamination from industrial activities is a large problem in urban areas worldwide. Understanding the spreading of contamination to underlying aquifers is crucial to make adequate risk assessments and for designing remediation actions. A large part of the northern hemisphere has quaternary deposits consisting of glacial clayey till. The till often has a complex hydrogeological structure consisting of networks of fractures, sand stringers and sand lenses that each contribute to a transport network for water, free phase and dissolved contaminants. Thus, to determine the possible flow-paths of contaminants, the geology must be described in great detail. Normally, multiple boreholes would be drilled in order to describe the geology, but boreholes alone do not provide the needed resolution to map such sand lenses and their connectivity. Cross-borehole full-decay time-domain induced polarization (TDIP) is a new tool that allows for quantitatively mapping not only contrasts in bulk resistivity, but also contrasts in spectral IP parameters. We present a feasibility study with synthetic tests and a field application on a clayey moraine environment with embedded sand lenses, with hitherto unseen ground-truth verification. Indeed, the investigated area was above the water table, which allowed for digging out the entire area after the investigation for an unprecedented description of the lens interconnectivity. The TDIP data were acquired with a full-waveform acquisition at high sampling rate, signal-processed by harmonic denoising, background removal, and de-spiking, and subsequently the full-waveform data were stacked in log-increasing tapered gates (with 7 gates per decade). The resulting TDIP decays, with usable time-gates as early as two milliseconds, were inverted in terms of a re-parameterization of the Cole-Cole model. The inverted models of the field data show a remarkable delineation of the sand lenses/layers at the site, with structure in both the resistivity and the IP parameters matching the results from the ground-truthing. The synthetic examples show that in models both below and above the groundwater table, sand-lenses with thicknesses comparable to the vertical electrode spacing can be well resolved. This suggests that full-decay cross-borehole TDIP is an ideal tool for high-resolution sand-lens imaging.

This study presents a novel expansion of the clay-fraction (CF)/resistivity clustering method aiming at developing realizations of subsurface structures based on multiple point statistics (MPS). The CF-resistivity clustering method is used to define a data driven training image (TI) for MPS simulations. By combining this TI with uncertainty estimates obtained from correlation between the resistivity models and the unique categories in the TI, subsurface realizations are generated honoring all geophysical and lithological data including any equivalent resistivity layer sequences. The generated subsurface realizations were calibrated in a steady state groundwater model. Forecasts of well catchment zones were derived based on two wells located in areas with different levels of structural uncertainty. The catchment probability maps derived from the structural realizations were compared with the well catchment forecasted by a deterministic subsurface structure, and we are able to capture this catchment within the estimated uncertainties. We believe that this study is the first to combine MPS methods with a complete data driven workflow going directly from lithological and geophysical data to realizations of the subsurface structures. The main benefits of this is that it is data driven, fast, reproducible, and transparent.

Recently, the interest in the induced polarization (IP) phenomenon in airborne time-domain electromagnetic (ATEM) data has increased considerably. IP may affect the ATEM data significantly and mask underlying geologic structures. To simulate 2D airborne IP data, a 2D finite-element forward-modeling algorithm has been developed with the dispersive conductivity described by the well-known Cole-Cole model. We verify our algorithm by comparison with the 1D solution of the AarhusInv code. Two-dimensional forward responses on six synthetic models, mimicking archetypal 2D conductive and chargeable anomalies, have been generated, and the results indicate that 2D IP affects the data significantly. Differences between the 2D IP responses and the 1D IP responses are evident above the 2D anomalies and at their edges. These differences are similar to what is found when comparing 2D and 1D forward responses over conductive 2D anomalies without considering IP. We evaluate an effective robust inversion scheme to recover the 2D IP parameters using the 1D laterally constrained inversion (LCI) scheme. The inversion of the synthetic data using the robust scheme indicates that not only can the IP parameters be recovered, but also the IP inversions can provide more accurate resistivity sections than a resistivity-only inversion, in terms of resistivity values and anomaly thickness/depth. The field example from Hope Bay area in Canada is even more valuable, considering that part of the profile consists of only negative data, which cannot be inverted with a resistivity-only scheme. Furthermore, the edge effects at the anomaly boundaries are less pronounced in the IP parameters than in the resistivity parameter on the synthetic models with more conductive backgrounds.

Surface nuclear magnetic resonance (NMR) is a powerful technique providing non-invasive imaging of groundwater. One challenge with the method is that it commonly suffers from low signal to noise ratios (SNR). Two methods to increase SNR are to either develop noise cancellation approaches to reduce the noise level, or to perform the experiment in a manner capable of increasing the signal amplitude. A recent study adopted the latter approach by employing a novel transmit strategy. An adiabatic pulse was employed and observed to produce significant signal improvements compared to the standard transmit method in surface NMR. The advantage of an adiabatic pulse is that it is capable of producing a uniform excitation in the presence of a heterogeneous magnetic field, which describes exactly the transmit conditions in surface NMR. Given the great potential of adiabatic pulses for surface NMR, we explore several factors related to the design of adiabatic pulses intended for application in surface NMR conditions. We investigate how various adiabatic pulses perform in a heterogeneous magnetic field given the limitation that current instrumentation couples the modulation of the current amplitude during the pulse to the instantaneous transmit frequency. That is, only the duration of the adiabatic sweep, the bandwidth through which the pulse sweeps, and frequency modulation throughout the pulse can be directly controlled. A numerical sensitivity analysis of each of these parameters is performed to gain insight into how to design optimal adiabatic pulses for surface NMR. Additionally, a numerically optimized modulation (NOM) approach is implemented to optimize the frequency sweep. The spatial resolution and depth penetration provided by an example adiabatic pulse is also investigated. A trade off between signal amplitude and spatial resolution is observed to be present when employing adiabatic pulses.

We describe the setup for inversion of ground based TEM data using a 3D modelling code and a full description of the measurement system using the system response. The response is calculated using a finite volume method where we solve for the electric field on the edges of a staggered grid and time step solutions using backward Euler steps. For the forward calculation we use an iterative solver and a preconditioner which solves the Helmholtz decomposed E-fields. In the iterative process we take advantage of the similarity between different time steps and always use the previous result as the starting point of the next iteration. To compare the calculated fields with measured data we interpolate the fields to the receiver positions and convolute the calculated fields with the system response. In this way we include all system related effects in the calculation and this is in particular relevant for the early time signal. We show that the forward code is in good agreement with the analytic response from a half space and sketch the layout of survey setup with a large, centrally positioned transmitter and many receivers located around the transmitter. We demonstrate that the code can be used to invert data from a single loop system with multiple receivers as commonly used in ground follow-up surveys in mineral exploration.

The extraction of spectral information in the inversion process of time-domain (TD) induced polarization (IP) data is changing the use of the TDIP method. Data interpretation is evolving from a qualitative description of the subsurface, able only to discriminate the presence of contrasts in chargeability parameters, towards a quantitative analysis of the investigated media, which allows for detailed soil- and rock-type characterization. Two major limitations restrict the extraction of the spectral information of TDIP data in the field: i) the difficulty of acquiring reliable early-time measurements, in the millisecond range and ii) the self-potential drift in the measured potentials distorting the shape of the late time IP responses, in the second range. Recent developments in TDIP acquisition equipment have given access to full waveform recordings of measured potentials and transmitted current, opening a breakthrough for data processing. For measuring at early times, we developed a new method for removing the significant noise from powerlines contained in the data through a model-based approach, localizing the fundamental frequency of the powerline signal in the full-waveform IP recordings. By this, we cancel both the fundamental signal and its harmonics. Furthermore, a novel and efficient processing scheme for identifying and removing spikes TDIP data is developed. The noise cancellation and the de-spiking allow the use of earlier and narrower gates, down to a few milliseconds after the current turn-off. Furthermore, tapered windows are used in the final gating of IP data, allowing the use of wider and overlapping gates for higher noise suppression without signal distortion. For measuring at late times, we have developed an algorithm for removal of the self-potential drift. Usually constant or linear drift-removal algorithms are used, but these algorithms fail in removing the background potentials due to the polarization of the electrodes previously used for current injection. We developed a drift-removal scheme that model the polarization effect and efficiently allows for preserving the shape of the IP responses at late times. Uncertainty estimates are essential in the inversion of IP data. Therefore, in the final step of the data processing, we estimate the data standard deviation based on the data variability within the IP gates and the misfit of the background drift removal Overall, the removal of harmonic noise, spikes, self-potential drift, tapered windowing and the uncertainty estimation allows for doubling the usable range of TDIP data to almost four decades in time (corresponding to four responses in frequency), and will significantly advance the science and the applicability of the IP method. (Less)

The nuclear magnetic resonance sounding (MRS) method is used increasingly as a tool for hydrological investigations. Compared to other geophysical methods, the advantage of MRS is that it is directly sensitive to the presence of water in the subsurface. Data interpretations can also be used to get information about the subsurface pore structures, which under special conditions can be related to hydraulic properties such as aquifer transmissivity. However, to broaden the usage of this information in hydrological modeling, the uncertainties related to these transmissivity estimates must be determined. Otherwise, properly balanced weights cannot be given to the prior information obtained from MRS transmissivity estimates as compared to the hydrological data sets when used for groundwater model calibration. We have developed a methodology to estimate the uncertainties of MRS-based transmissivity estimates. Compared to previous studies, the methodology is well defined, and it takes into account important factors such as the uncertainties of the hydraulically estimated transmissivities, the uncertainty of the correlation factor in the petrophysical relation, and the uncertainties and correlations of the geophysically estimated parameters. We have determined the correlations and uncertainties of the geophysical parameters using a linear and a nonlinear method, and we find that the results are comparable.

In hardrocks that cover about 20% of the Earth’s surface, it is difficult to locate steady sources for groundwater due to inadequate understanding of the fracture networks. A comprehensive knowledge of fracture distribution at the regional scale is necessary to delineate sustainable aquifers and manage them efficiently. The resistivity maps derived from the airborne electromagnetic (AEM) survey over the Ankasandra watershed in Karnataka, India, reveal sharp and deep zones of low formation resistivity, which indicate groundwater-bearing zones. It is found that some of these zones are hydrogeologically connected through fracture networks resulting in augmented yield. AEM results in combination with an in-depth understanding of the geological structures successfully map these groundwater-saturated fracture networks (or hydrogeological lineaments) that we term as ‘Hydrolins’. As groundwater occurrence is generally associated with lineaments, we analyzed the drilling and geophysical logs from 21 wells within a 380 sq.km area to study the relationships of various lineaments with ‘Hydrolins’, particularly in respect of their groundwater potential. AEM results, though calibrated and correlated with a limited number of well data, revealed a threshold groundwater horizon (TGWH), found to be at 80 m depth for Ankasandra watershed, beyond which a strong correlation exists between the depth of a well and its yield. While the TGWH may differ for different watersheds, the approach presented here can be readily adopted to map sustainable groundwater sources in hardrocks worldwide.

In the present study, we show how persistent management and collection of hydrological and geophysical data at a national scale can be combined with innovative analysis methods to generate decision support tools for groundwater and surface water managers. This is exemplified by setting up a regional scale groundwater model in an area with geophysical data of varying age, type, and quality. The structure for the regional model is derived from a newly developed resistivity clay-fraction cluster analysis. This modelling strategy can be used in combination with local detailed geological modelling thus utilizing the detailed expertise locally, while securing a cost-effective (price vs. performance) solution to the numerical simulations of the regional scale water balance. In this way we avoid unwanted boundary effects on the local model simulations due to the presence of artificial numerical boundaries located proximate to the areas of interest. In this application, it is particularly important that boundary conditions are remote, due to the presence of a dense network of buried valley structures. Simulated impacts of groundwater abstraction from two existing well-fields spread through the valley system far beyond the local focus areas of the study.

Surface nuclear magnetic resonance (surface-NMR) has the potential to be an important geophysical method for groundwater investigations, but the technique suffers from poor signal-to-noise ratio (SNR) and long measurement times. We present a new wireless, multichannel surface-NMR receiver system (called Apsu) designed to improve SNR, field deployability and minimize instrument dead time. It is a distributed wireless system consisting of a central unit and independently operated data acquisition boxes each with three channels that measure either the NMR signal or noise for reference noise cancellation. Communication between the central unit and the data acquisition boxes is done through long distance WiFi and recordings are retrieved in real time. The receiver system employs differential coils with low-noise pre-amplifiers and high-resolution wide dynamic range acquisition boards. Each channel contains multi-stage amplifiers, short settling-time filters and two 24-bit analog-to-digital converters in dual-gain mode sampling at 31.25 kHz. The system timing is controlled by GPS clock and sample jitter between channels is less than 12 ns. Separated transmitter/receiver coils and continuous acquisition allow NMR signals to be measured with zero instrument dead time. In processed data, analog and digital filters causes an effective dead time of 4 ms. Synchronization with an independently operated transmitter system is done with a current probe monitoring the NMR excitation pulses. The noise density measured in a shorted-input test is 1.8 nV/√(Hz). We verify the accuracy of the receiver system with measurements of a magnetic dipole source and by comparing our NMR data with data obtained using an existing commercial instrument. The applicability of the system for reference noise cancellation is validated with field data.

Surface nuclear magnetic resonance (surface NMR) has the potential to be an important geophysical method for groundwater investigations, but the technique suffers from a poor signal-to-noise ratio (SNR) and long measurement times. We present a new wireless, multichannel surface-NMR receiver system (called Apsu) designed to improve field deployability and minimize instrument dead time. It is a distributed wireless system consisting of a central unit and independently operated data acquisition boxes each with three channels that measure either the NMR signal or noise for reference noise cancellation. Communication between the central unit and the data acquisition boxes is done through long-distance Wi-Fi and recordings are retrieved in real time. The receiver system employs differential coils with low-noise preamplifiers and high-resolution wide dynamic-range acquisition boards. Each channel contains multistage amplifiers, short settling-time filters, and two 24 bit analog-to-digital converters in dual-gain mode sampling at 31.25 kHz. The system timing is controlled by GPS clock, and sample jitter between channels is less than 12 ns. Separated transmitter/receiver coils and continuous acquisition allow NMR signals to be measured with zero instrument dead time. In processed data, analog and digital filters cause an effective dead time of 5.8 ms including excitation current decay. Synchronization with an independently operated transmitter system is done with a current probe monitoring the NMR excitation pulses. The noise density measured in a shorted-input test is 1.8 nV Hz-1/2. We verify the accuracy of the receiver system with measurements of a magnetic dipole source and by comparing our NMR data with data obtained using an existing commercial instrument. The applicability of the system for reference noise cancellation is validated with field data.

Envelope detection is an integral step of the surface NMR data processing workflow to estimate these parameters. We present a new method for retrieving NMR signal envelopes using spectral analysis and the subsequent inversion scheme with the new data space. By exploiting the fact that the spectral magnitude at the Larmor frequency is proportional to the product of the initial amplitude and relaxation time, a high-SNR complex envelope can be extracted by Fourier transform for a number of sliding windows. However, SA estimated envelopes are weighted by the NMR relaxation time during processing and the envelopes have units of volt-seconds. We propose to modify the surface NMR forward model such that it predicts data directly in the voltage-time data space. Field data inversions are presented to demonstrate advantages of pairing the SA envelope detection scheme with a forward model that works in the voltage-time data space.

With the computer power available today full 3D inversions of transient electromagnetic data (TEM) is no longer a dream of the past. Although it is possible to perform these inversions, the problems still scale in three dimensions making large datasets slow to invert. We here propose a new triple mesh method for inverting TEM datasets with multiple transmitters and multiple receivers per transmitter. The code is relative fast and with a manageable memory consumption. In this new approach we show that by using a decoupled regular structured model mesh and two finite element forward meshes, one with a coarse discretization and one with a fine discretization, we get a substantial speed up in calculations times without sacrificing much in terms of how well we fit the data. We show that we can invert large datasets by decomposing our domain and applying this triple mesh method on each domain separately.

Summary The top 0–50 m of the subsurface are critical for infrastructure, water supply, artificial infiltration, farming, waste deposits, construction, etc. Yet, the tools for imaging this zone are limited to boreholes or geophysical methods such as electrical resistivity imagining (ERT) with a limited applicability for area-covering surveys larger than a few hectares. We present a new highly efficient towed transient electromagnetic system, tTEM, which bridges the gap in coverage and resolution between point or line measurements and airborne electromagnetics. The system yields images in full 3D with a lateral resolution down to 10 × 10 m. The system is towed by and All-Terrain Vehicle, uses a 2 × 4 m2 transmitter coil and has a z-component receiver 9 m offset from the transmitter. The first bias-free gate is as early as 4 µs from ramp beginning (1.4 µs after ramp-end). The turn-off time is 2.6 µs. Data are processed and inverted using methods directly adopted from airborne electromagnetics. We discuss the system design and present a case study where the system has been used to map complex subsurface geology in large detail to build 3D geological/hydrogeological models.

In this study we present the prediction of hydraulic permeability (k) from time-domain spectral induced polarization (IP) data, measured in boreholes and along 2D surface profiles. The borehole data were collected with the El-log technique, which provides undisturbed “while drilling” measurements of the direct current (DC) resistivity, time-domain IP data and gamma radiation. Data were measured on unconsolidated formations at a landfill site in South Denmark, in three boreholes and along 16 2D profiles. DC and full-decay IP data were inverted in terms of the BIC re-parameterization of the Cole-Cole model, which present smaller parameter correlations and disentangles bulk and surface conduction. Permeability values were computed from bulk conductivity and the maximum imaginary conductivity, using the empirically-derived formulae presented in a recent study without any calibration. The IP-derived k estimates were compared to those estimated using grain size analysis and slug tests, for a total of 157 comparisons. A good correlation, on average within one decade, was found between the k estimates over four orders of magnitude, with similar depth-trends. In conclusion, IP can be reliably used for estimating hydraulic permeability on unconsolidated formations at the field scale, using the relations found in the laboratory without any further calibration.

Summary We present new developments of the versatile towed transient electromagnetic system (tTEM). The system now meets the demand of near surface mapping down to 100m and is suitable for application not only ground-based but also in aquatic (rivers, lakes etc.) and snow environments. The receiver technology of the system has been upgraded and it has a 4 times lower noise level compared to previous system, increasing the depth of investigation down to 100 m. The latter two system are called FloaTEM and SnowTEM, respectively. FloaTEM surveys were conducted in several parts of United States including the Mississippi delta region for mapping and characterizing the alluvial aquifer system below the river beds. The results show that the FloaTEM system produces detailed subsurface resistivity images in quick time compared to other geophysical methods. We conducted over 300 km of survey along several rivers within a week in the Mississippi delta region. The SnowTEM system uses a 4 turn, 4 m x 4 m transmitter coil and a specially designed low noise receiver coil. It has been put on skies and is towed behind a snow scooter or a snow robot.

Surface nuclear magnetic resonance (NMR) data are sensitive to key hydrogeological parameters including water content and pore size. The measured data are modelled as a complex sinusoidal exponential decay where the phase is a function of the physics of the experiment and instrumental factors, parameters that are difficult to decouple. When inverting surface NMR data, practitioners typically account for these phases either by considering only the amplitudes of the complex signals, thus eliminating the influence of the phase, or by iteratively rotating the complex data during inversion so the data phase matches the theoretical phase generated during forward modelling. Each of these approaches assumes the user has an accurate forward model; if not, the data will be incorrectly rotated and forced to fit an erroneous forward model. Additionally, this rotation will artificially reduce the total data misfit, thus masking the effect of the erroneous forward model. We demonstrate the pitfalls of using inverse methods that correct for the phase by inverting synthetic data with three types of deliberate modelling errors that may occur during a surface NMR experiment: errors in the offset between the Larmor and transmit frequencies, errors in the subsurface resistivity model, and errors in the relative positioning of a separated transmitter-receiver coil pair. The inverted water content profiles show that the modelling errors can introduce inversion artifacts. However, the amplitude inversions and complex inversions with iterative phase correction frequently produce χ2 misfit values close to unity, showing that these inverse methods will fail to “raise the alarm” when an incorrect forward model is implemented.

In Denmark and several other countries drinking water originates from groundwater treated in open sand filters. Due to iron oxidation, the upper 30 cm of the filter clogs over time, so for optimal performance, the open sand filters must be backwashed periodically. However, often the waterworks has insufficient knowledge about the clogging process and the backwash is not performed efficiently. In this project, we try to optimize the backwash process by monitoring the clogging process, by time-domain induced polarization (TDIP) measurements. The setup consisted of 14 electrodes installed at various depth along a 2D profile within the filter, and full-waveform IP data, sampled at 3750 Hz, was recorded on a daily basis. The environment inside the water-filter proved to be very noisy, so special care in the signal processing was necessary. In particular, harmonic de-noising was essential to retrieve reliable IP data. We present the preliminary results from this study.

Summary We have developed a new towed magnetic gradiometer array for environmental, engineering, and archaeological applications. With 16 sets of vector magnetometers plus accelerometers on a low-noise platform we can construct an estimate of the full magnetic gradient tensor. A high sample rate and suspension system allows for rapid collection of data—up to 70 hectares a day at full sensor coverage, while careful design of the platform has removed the majority of noise sources from proximity to the sensors. The application of vector magnetic gradiometry allows for advanced processing techniques such as calculation of tensor invariants, principal component analyses, wavelet decomposition, and more. In this abstract, we describe the design of the gradient array and introduce an initial processing workflow. We also discuss system noise characteristics and show an example from a clay tile drain mapping application demonstrating the efficacy of the system for accurate detection of low-signal sources.

Conventional regularized nonlinear inversion methods for estimating electrical conductivity from observed electromagnetic data seek to find a single model that fits the data while minimizing a user-imposed model regularization norm. By contrast, Bayesian sampling techniques produce a large suite of models, all of which fit the data adequately, providing a wealth of statistical information about the model parameters. Importantly, this includes quantitative uncertainty estimates as well as any statistical property of interest. In this work, we apply a Bayesian trans-dimensional Markov chain Monte Carlo scheme to recover subsurface conductivity from airborne transient electromagnetic (TEM) data collected over Taylor Glacier, Antarctica, to image subglacial hydrologic structure. We provide a synthetic model study, followed by inversions of real soundings. Our results identify a zone of conductive, wet sediments beneath the glacier, corroborating interpretations from previous studies that used regularized, smooth inversions. Our results provide, however, the opportunity to examine a rich suite of additional information, including uncertainty estimates on the conductivity within the conductive subglacial layer as well as quantitative estimates of its total conductance. We apply principles of Bayesian information theory for estimating the depth of investigation of the airborne TEM data and apply it to this data set. Additionally, we use themodel ensemble to derive estimates of pore fluid resistivity within the conductive layer, with associated uncertainties. Finally, we use Bayesian model studies to explore the range of ice thicknesses and conductive layer thicknesses that could be resolved with ground or airborne TEM data if they had one to two orders of magnitude lower noise levels.

In many cases, inversion in 2D gives a better description of the subsurface compared with 1D inversion, but, computationally, 2D inversion is expensive, and it can be hard to use for large-scale surveys. We have developed an efficient hybrid 2D airborne frequency-domain electromagnetic inversion algorithm. Our hybrid scheme combines 1D and 2D inversions in a three-stage process, in which each step is progressively more accurate and computationally more expensive than the previous one. This results in an approximately [Formula: see text] speedup compared with full 2D inversions, and with only minor changes to the inversion results. Our inversion structure is based on a regular grid, in which each sounding is discretized individually. The 1D modeling code uses layered models with derivatives derived through the finite-difference method, whereas our 2D modeling code uses an adaptive finite-element mesh, and it uses the adjoint-state method to calculate the derivatives. By incorporating the inversion grid structure into the 2D finite-element mesh, interpolation between the different meshes becomes trivial. Large surveys are handled by using local meshing to split large surveys into small sections, which retains the 2D information. The algorithm is heavily optimized and parallelized over the frequencies and sections, with good scalability even on nonuniform memory architecture systems, on which it is generally hard to achieve a satisfactory scaling. The algorithm has been tested successfully with various synthetic studies as well as field examples, of which results from two synthetic studies and a field example are shown.

One of the primary shortcomings of the surface nuclear magnetic resonance (NMR) free-induction decay (FID) measurement is the uncertainty surrounding which mechanism controls the signal's time dependence. Ideally, the FID-estimated relaxation time T2∗ that describes the signal's decay carries an intimate link to the geometry of the pore space. In this limit the parameter T2∗ is closely linked to a related parameter T2, which is more closely linked to pore-geometry. If T2∗ T2 the FID can provide valuable insight into relative pore-size and can be used to make quantitative permeability estimates. However, given only FID measurements it is difficult to determine whether T2∗ is linked to pore geometry or whether it has been strongly influenced by background magnetic field inhomogeneity. If the link between an observed T2∗ and the underlying T2 could be further constrained the utility of the standard surface NMR FID measurement would be greatly improved. We hypothesize that an approach employing an updated surface NMR forward model that solves the full Bloch equations with appropriately weighted relaxation terms can be used to help constrain the T2∗-T2 relationship. Weighting the relaxation terms requires estimating the poorly constrained parameters T2 and T1; to deal with this uncertainty we propose to conduct a parameter search involving multiple inversions that employ a suite of forward models each describing a distinct but plausible T2∗-T2 relationship. We hypothesize that forward models given poor T2 estimates will produce poor data fits when using the complex-inversion, while forward models given reliable T2 estimates will produce satisfactory data fits. By examining the data fits produced by the suite of plausible forward models, the likely T2∗-T2 can be constrained by identifying the range of T2 estimates that produce reliable data fits. Synthetic and field results are presented to investigate the feasibility of the proposed technique.

Surface nuclear magnetic resonance is a technique capable of providing insight into subsurface aquifer properties. To produce estimates of aquifer properties (such as the spatial distribution of water content and parameters controlling the duration of the nuclear magnetic resonance signal), an inversion is required. Essential to the reliable interpretation of the estimated subsurface models is an understanding of the uncertainty and correlation between the parameters in the estimated models. To quantify parameter uncertainty and correlation in the surface nuclear magnetic resonance inversion, a Markov chain Monte Carlo approach is demonstrated. Markov chain Monte Carlo approaches have been previously employed to invert surface nuclear magnetic resonance data, but the primary focus has been on quantifying parameter uncertainty. The focus of this paper is to further investigate whether the parameters in the estimated models exhibit correlation with one another; equally important to building a reliable interpretation of the subsurface is an understanding of the parameter uncertainty. The utility of the Markov chain Monte Carlo approach is demonstrated through the investigation of three questions. The first question investigates whether the parameters describing the water content and thickness of a layer exhibit a strong correlation. This question stems from applying concepts known to electromagnetic surveys (that the layer thickness and layer resistivity parameters are strongly correlated) to the surface nuclear magnetic resonance inversion. A water content–layer thickness correlation in surface nuclear magnetic resonance would not have large effects for quantifying total water content but would affect the ability to identify layer boundaries. The second question examines whether the parameter controlling the duration of the nuclear magnetic resonance signal exhibits a correlation with the water content and layer thickness parameters. The resolution of surface nuclear magnetic resonance typically does not consider the duration of the signal and focuses primarily on the distribution of current amplitudes that form the suite of transmit pulses. It is common to treat regions with short-duration signal with greater uncertainty, but it is important to understand whether the signal duration controls resolution for medium to long duration signals as well. The third question explores if the parameter uncertainty produced by the Markov chain Monte Carlo approach is consistent with that produced by an alternative approach based upon the posterior covariance matrix (for the linearised inversion). The ability of the Markov chain Monte Carlo approach to more thoroughly explore the model space provides a means to improve the reliability of surface nuclear magnetic resonance aquifer characterisations by quantifying parameter uncertainty and correlation.

We have investigated the 3D topographic effects on controlled-source audio-frequency magnetotelluric data. Two 3D topographic models are considered: a trapezoidal-hill model and a trapezoidal-valley model. Different responses are generated, including the amplitude of the electric field, the amplitude of the magnetic field, the apparent resistivity, and phase data. The responses distorted by the 3D topography are simulated for the source located next to and on the hill/valley. Our study indicates that all electric field, magnetic field, apparent resistivity, and phase data are influenced by 3D topography, but to different extents. These topographic effects depend on the transmission-receiver-topography geometry, the transmission frequency, earth resistivity, and the roughness of the surface. The effects in the near-field generated by topography in the survey area are quite different from those in the far-field because of the existence of the source. Compared with those in the far-field zone, the magnetic field and phase data in the near-field zone are less distorted, but more distortions can be found on the electric field and apparent resistivity data over the hill and valley models. Our results also indicate that not only can the 3D topography in the receiver area lead to strong distortions, but also that at the source position can lead to strong distortions. We concluded our study by quantifying the roughness with which the topographic distortion can be ignored, setting the accepted data distortion to a maximum of 10%.

Combined resistivity and time-domain direct current induced polarization (DCIP) measurements are traditionally carried out with a 50\% duty cycle current waveform, taking the resistivity measurements during the on-time and the IP measurements during the off-time. One drawback with this method is that only half of the acquisition time is available for resistivity and IP measurements, respectively. In this paper, this limitation is solved by using a current injection with 100\% duty cycle and also taking the IP measurements in the on-time. With numerical modelling of current waveforms with 50\% and 100\% duty cycles we show that the waveforms have comparable sensitivity for the spectral Cole–Cole parameters and that signal level is increased up to a factor of 2 if the 100\% duty cycle waveform is used. The inversion of field data acquired with both waveforms confirms the modelling results and shows that it is possible to retrieve similar inversion models with either of the waveforms when inverting for the spectral Cole–Cole parameters with the waveform of the injected current included in the forward computations. Consequently, our results show that on-time measurements of IP can reduce the acquisition time by up to 50\% and increase the signal-to-noise ratio by up to 100\% almost without information loss. Our findings can contribute and have a large impact for DCIP surveys in general and especially for surveys where time and reliable data quality are important factors. Specifically, the findings are of value for DCIP surveys conducted in urban areas where anthropogenic noise is an issue and the heterogeneous subsurface demands time-consuming 3D acquisitions. (Less)

SUMMARY We present an automatic method for parameterization of a 3D model of the subsurface, integrating lithological information from boreholes with resistivity models through an inverse optimization, with the objective of creating a direct input to groundwater models. The parameter of interest is the clay fraction, expressed as the relative length of clay-units in a depth interval. The clay fraction is obtained from lithological logs and the clay fraction from the resistivity is obtained by establishing a simple petrophysical relationship, a translator function, between resistivity and the clay fraction. Through inversion we use the lithological data and the resistivity data to determine the optimum spatially distributed translator function. Applying the translator function we get a 3D clay fraction model, which holds information from the resistivity dataset and the borehole dataset in one variable. Finally, we use kmeans clustering to generate a 3D model of the subsurface structures, which we then use as direct input in a groundwater model. We apply the concept to the Norsminde survey in Denmark integrating approximately 700 boreholes and more than 100,000 resistivity models from an airborne survey in the parameterization of the 3D model covering 156 km2. The final five-cluster 3D model is input to a groundwater model and it performs equally well or slightly better than traditional groundwater models from the area.

Geophysical Journal International, 203 (2), ISSN:0956-540X, ISSN:1365-246X

Modern airborne transient electromagnetic surveys typically produce datasets of thousands of line kilometres, requiring careful data processing in order to extract as much and as reliable information as possible. When surveys are flown in populated areas, data processing becomes particularly time consuming since the acquired data are contaminated by couplings to man-made conductors (power lines, fences, pipes, etc.). Coupled soundings must be removed from the dataset prior to inversion, and this is a process that is difficult to automate. The signature of couplings can be both subtle and difficult to describe in mathematical terms, rendering removal of couplings mostly an expensive manual task for an experienced geophysicist. Here, we try to automate the process of removing couplings by means of an artificial neural network. We train an artificial neural network to recognize coupled soundings in manually processed reference data, and we use this network to identify couplings in other data. The approach provides a significant reduction in the time required for data processing since one can directly apply the network to the raw data. We describe the neural network put to use and present the inputs and normalizations required for maximizing its effectiveness. We further demonstrate and assess the training state and performance of the network before finally comparing inversions based on unprocessed data, manually processed data, and artificial neural network automatically processed data. The results show that a well-trained network can produce high-quality processing of airborne transient electromagnetic data, which is either ready for inversion or in need of minimal manual processing. We conclude that the use of artificial neural network scan significantly reduce the processing time and its costs by as much as 50%.

The increased application of airborne electromagnetic surveys to hydrogeological studies is driving a demand for data that can consistently be inverted for accurate subsurface resistivity structure from the near surface to depths of several hundred metres. We present an evaluation of three commercial airborne electromagnetic systems over two test blocks in western Nebraska, USA. The selected test blocks are representative of shallow and deep alluvial aquifer systems with low groundwater salinity and an electrically conductive base of aquifer. The aquifer units show significant lithologic heterogeneity and include both modern and ancient river systems. We compared the various data sets to one another and inverted resistivity models to borehole lithology and to ground geophysical models. We find distinct differences among the airborne electromagnetic systems as regards the spatial resolution of models, the depth of investigation, and the ability to recover near-surface resistivity variations. We further identify systematic biases in some data sets, which we attribute to incomplete or inexact calibration or compensation procedures.

Airborne electromagnetic (AEM) data is used throughout the world for mapping of mineral targets and groundwater resources. The development of technology and inversion algorithms has been tremendously over the last decade and results from these surveys are high-resolution images of the subsurface. In this keynote talk, we discuss an effective inversion algorithm, which is both subjected to intense research and development as well as production. This is the well know Laterally Constrained Inversion (LCI) and Spatial Constrained Inversion algorithm. The same algorithm is also used in a voxel setup (3D model) and for sheet inversions. An integral part of these different model discretization is an accurate modelling of the system transfer function and of auxiliary parameters like flight altitude, bird pitch, etc.

The free-induction decay (FID) is the most widely implemented surface nuclear magnetic resonance (NMR) experiment. One shortcoming of the FID is that there is often much uncertainty surrounding the meaning of the FID's time dependence - can the FID provide valuable insight into pore-size and permeability or has the signal been contaminated by background magnetic field inhomogeneity? To address this uncertainty, we investigate if an updated surface NMR forward modelling scheme, which involves solution of the full Bloch-equation, can provide the ability to better constrain the link between FIDs and pore-sizes. That is, can the link between the relaxation times T2* and T2 be constrained from FID data alone? A numerical analysis of whether an inversion based on the updated forward model can produce depth profiles of both T2* and T2 from only FID data is presented.

The top 30 - 50 m of the subsurface are critical for infrastructure, water supply, artificial infiltration, farming, waste deposit, construction, etc.. Yet, the tools for imaging this zone are limited to boreholes, electrical resistivity imagining (ERT) or other relatively expensive methods with a limited applicability for areas larger than a few hectares. We present a new highly efficient towed transient electromagnetic system, tTEM, which bridges the gap in coverage and resolution between point measurements and airborne electromagnetics. The system yields images in full 3D with a resolution down to 10 x 10 m. The system is towed by and All-Terrain Vehicle (ATV), uses a 2 x 4 m transmitter coil and has a z-component receiver 9 m offset from the transmitter. The first bias free gate is as early as 4 μs from ramp beginning . The turn-off time is 2.6 μs. Data are processed and inverted using methods adopted from airborne electromagnetics. We discuss the system design and present a case study where the system has been used to evaluate the potential contamination risk from a pesticide point source contamination to four water extraction wells.

With recent advances in groundbased and airborn electromagnetic methods, groundwater modellers can be equiped unique datasets holding high resolution information on the subsurface structures. These data often reveal a highly complex reality. It is hard to argue that data gaps can be modelled deterministically. Here, we would like to show, how multiple point geostatistical methods can be used in combination with 3D EM data to generate subsurface realizations, and how these realizations can be used efficiently in groundwater models.

In this study, we present the prediction of permeability from time domain spectral induced polarization (IP) data, measured in boreholes on undisturbed formations using the El-log logging-while-drilling technique. We collected El-log data and hydraulic properties on unconsolidated Quaternary and Miocene deposits in boreholes at three locations at a field site in Denmark, characterized by different electrical water conductivity and chemistry. The high vertical resolution of the El-log technique matches the lithological variability at the site, minimizing ambiguity in the interpretation originating from resolution issues. The permeability values were computed from IP data using a laboratory-derived empirical relationship presented in a recent study for saturated unconsolidated sediments, without any further calibration. A very good correlation, within 1 order of magnitude, was found between the IP-derived permeability estimates and those derived using grain size analyses and slug tests, with similar depth trends and permeability contrasts. Furthermore, the effect of water conductivity on the IP-derived permeability estimations was found negligible in comparison to the permeability uncertainties estimated from the inversion and the laboratory-derived empirical relationship.

Det danske samfund bruger hvert år mange midler på at undersøge, risikovurdere og oprense forurenede grunde. Det er derfor vigtigt, at vi konstant forbedrer de undersøgelser, der ligger til grund for, hvorledes vi risikovurderer og prioriterer de mange forurenede grunde til oprensning.

The nuclear magnetic resonance (NMR) technique has become popular in groundwater studies because it responds directly to the presence and mobility of water in a porous medium. There is a need to conduct laboratory experiments to aid in the development of NMR-hydraulic conductivity models, as is typically done in the petroleum industry. However, the challenge has been obtaining high-quality laboratory samples from unconsolidated aquifers. At a study site in Denmark, we employed sonic drilling, which minimizes the disturbance of the surrounding material, and extracted twelve 7.6-cm diameter samples for laboratory measurements. We present a detailed comparison of the acquired laboratory- and logging-NMR data. The agreement observed between the laboratory and logging data suggests that the methodologies proposed in this study provide good conditions for studying NMR measurements of unconsolidated near-surface aquifers. Finally, we show how laboratory sample size and condition impact the NMR measurements.

Surface nuclear magnetic resonance (NMR) is a geophysical technique providing noninvasive insight into aquifer properties. To ensure that reliable water content estimates are produced, accurate modeling of the excitation process is necessary. This requires that relaxation during pulse (RDP) effects be accounted for because they may lead to biased water content estimates if neglected. In surface NMR, RDP is not directly included into the excitation modeling, rather it is accounted for by adjusting the time at which the initial amplitude of the signal is calculated. Previous work has demonstrated that estimating the initial amplitude of the signal as the value obtained by extrapolating the observed signal to the middle of the pulse can greatly improve performance for the on-resonance pulse. To better understand the reliability of these types of approaches (which do not directly include RDP in the modeling), the performance of these approaches is tested using numerical simulations for a broad range of conditions, including for multiple excitation pulse types. Hardware advances that now allow the routine measurement of much faster relaxation times (where these types of approaches may lead to poor water content estimates) and a recent desire to use alternative transmit schemes demand a flexible protocol to account for RDP effects in the presence of fast relaxation times for arbitrary excitation pulses. To facilitate such a protocol, an approach involving direct modeling of RDP effects using estimates of the subsurface relaxation times is presented to provide more robust and accurate water content estimates under conditions representative of surface NMR.

We present a 2D inversion code for frequency-domain HEM data designed for inverting field scale surveys on normal desktop computers. The fundamental algorithm is a 2.5D algorithm with field separation into primary and secondary fields. Due to limited memory, as well as performance concerns, sectioning is introduced for splitting large surveys into smaller sections. Sectioning is only done when calculating the 2D forward and derivatives, and it is done with a sufficient overlap, such that vital 2D information is preserved. The algorithm uses a hybrid scheme which i) starts with 1D forward and inverse calculations, ii) then switches to 2D forward calculations and 1D derivatives, and iii) finally ends with full 2D calculations. The result of this is a code which produces results like a 2D code, but with a substantially shorter computational time.

Summary Knowledge of the sensitivity function of an electromagnetic (EM) system helps to assess the advantages and shortcomings of the system set-up. We have computed 3D sensitivity functions of three different EM systems – one airborne (SkyTEM) and two ground-based systems (tTEM and GCM) – in order to compare the sensitivity distribution of the different transmitter-receiver set-ups. The results show that the GCM has the highest sensitivity to near-surface structures, but also has a very limited footprint depth. The SkyTEM and tTEM systems have comparable sensitivities vertically in the ground with a 90% footprint depth around 5 m for early time measurements. However, the sensitivity of the SkyTEM system is distributed must wider horizontally than the tTEM.