Search

International audience; Due to the complexity of geological formations and limited subsurface data in volcanic islands, hydrogeological conceptual models can differ from one island to another and for the same island. The Galapagos Islands, like most inhabited volcanic islands, face important water resource problems which might have a major impact on their unique and pristine ecosystems, Endanger World Heritage list (June 2007). The scarcity of geological and hydrological data combined with the difficulty of access for field measurements lead to a Poor understanding of the island hydrogeology and unconvincing interpretation of traditional geophysical data. Here we present three dimensional (3-D) resistivity maps for the supposedly "waterless" Santa Cruz Island, obtained by using the SkyTEM device, a helicopter-borne transient electromagnetic method. The latter is non-invasive and measurements over inaccessible terrain were of vital importance. We show that even in high-relief terrain with extreme subsurface resistivity contrasts [1-6 000 ohm-m], the method is sensitive to low-resistivity layers of hydrogeological interest [50-200 ohm-m] to a depth of approximately 300 m. The unique spatial resolution and 3-D view of the subsurface resistivity structures allow identification of two zones of hydrogeological importance: a previously unknown extensive perched aquifer (50 km(2)) on the southern mountain side and the geometry of the salt-water wedge in the basal aquifer located with an accuracy of a few meters, This finding supports the existence of hidden perched aquifers on basaltic islands, until now only inferred from hydrogeological studies. It is seen to be affected by faulting. Beyond the vital implications for water resource and ecosystem management in the unique yet severely threatened Galapagos Islands, some conceptual flow models of volcanic islands have to be reconsidered. The presence of such structures in a volcanic edifice recognized as low strength layer may help to explain the origin of potentially catastrophic landslides. (C) 2008 Elsevier B.V. All rights reserved.

Geophysical and hydrogeological investigations have been carried out around Sawmills in Zimbabwe, Africa. The investigations are components of a larger investigation to assess the groundwater potential of the Karoo sedimentary basin with regards to supplying water to Bulawayo City. The Sawmills area was selected due to the availability of borehole logs indicating favourable stratigraphy for groundwater availability and due to the high yields from the aquifers measured from these boreholes. Data collected using two geophysical methods are presented here: transient electromagnetic (TEM) and continuous vertical electrical sounding (CVES) data. The data have also been processed using laterally constrained inversion (LCI). Because the CVES provides greater detail in the shallow subsurface, whereas TEM is more effective at depth, a more accurate image of the entire subsurface profile is provided based on using both methods. The results suggest that LCI of CVES and TEM data, in the subsurface at the required depths at Sawmills, is able to provide a substantially more accurate image of the subsurface than either method alone. The hydrogeological interpretation of the geophysical data is valuable for determining the depth to and thickness of the potential aquifer horizon(s) and for identifying the position of potential recharge zones.

We describe the challenges for the realization of a transient EM (TEM) system capable of measuring both the z- and x-components. This involves significant changes to instrumentation as well as to the processing and inversion algorithms. Geophysical mapping of groundwater resources using TEM involves both delineation of the aquifer structure and the aquifer protection. The demands for the instrumentation and the processing algorithms are quite different for the two tasks. Structural mapping requires large depth penetration and thereby a large transmitter moment while aquifer protection mapping requires early time measurements and extremely accurate determination of the system transfer function. Furthermore, the inversion software must be capable of modelling this system transfer function. For an airborne TEM system systematic analysis of the resolution capabilities of relatively thin low resistive layers in the upper 30 - 50 m of the subsurface has shown that the resolution is enhanced if not only the transient z-component but also the transient x-component of the induced fields in the Earth is measured and included in the inversion. The abstract will describe and discuss three major challenges related to the inclusion of the x-component in an airborne TEM system.

In this presentation, we discuss geophysical methods for general vulnerability mapping of aquifers. The vulnerability of aquifers is to a large degree determined by the content of clay minerals in the near surface geology. Clay minerals have a large cation exchange capacity (CEC) and clayey sediments reduce the infiltration speed thus reducing the aquifer vulnerability. In sedimentary areas, the near surface geology often shows large variations, and therefore the requirements for dense detailed geophysical data are required in order to achieve a satisfying geological resolution. The DC resistivity method is a good choice as it has a high near surface resolution. Alternatively, very early transient electromagnetic measurements can be used. The latter are cost effective and can often be made from the air. We discuss advantages and drawbacks of the two methods and show an example from a 100 km survey where we have used a newly developed variation of the helicopter-borne SkyTEM system for vulnerability mapping. Both the Zand the X-component of the decaying field were measured and the data were inverted using a laterally constrained inversion method.

Resistivity imaging in combination with borehole information is a powerful tool for site investigation. We show that the combination of 1D laterally constrained inversion (1D-LCI) with the use of a priori information from borehole data and 2D smooth inversion adds significant value to the interpretation of continuous vertical electrical sounding (CVES) data. The ID-LCI offers an analysis of the resolution of the model parameters. This is helpful when evaluating the integrity of the model. Furthermore, with the 1D-LCI it is possible to constrain model parameters with a priori information, e.g. depth-to-layer interfaces, based on borehole information. We show that 2D smooth inversion resolves lateral changes well, while 1D-LCI results in well-defined horizontal layer interfaces. In geological environments where the lateral variations are not too pronounced, the 1D-LCI contributes to a geological interpretation of the resistivity measurements. Depths to layers can be interpreted with greater certainty than if using results from 2D smooth inversion only. The inclusion of a priori information in the inversion reveals further details and enhances the geological interpretation significantly.

A033 Geostatistical estimation of structural vulnerability Summary 1 In this abstract we present a newly developed concept using geostatistical estimation and nonlinear inversion to optimize a function translating geophysical resistivity models to geophysical clay thickness. The algorithm uses information on reported clay thicknesses in drill holes. We call it geoStatistical estimation of Structural Vulnerability (SSV). The clay thickness in the upper part of the subsurface is an important factor for the water infiltration speed and thereby the vulnerability of underlying aquifers to pesticides nitrate etc. Drill hole information contains the most detailed information on clay thicknesses but most often drill

A035 INTERPRETATION OF A HYDROGEOPHYSICAL SURVEY – DATA FROM THE HIGH- RESOLUTION SKYTEM SYSTEM Abstract 1 During the last decade approximately 50 000 ground-based TEM soundings have been carried out in Denmark for groundwater investigations. This number will increase in the future and therefore we have developed a new time-domain electromagnetic (TEM) helicopter system SkyTEM. In the development and design of the SkyTEM system it has been an unchangeable demand that the data quality of the SkyTEM system should be the same or better than the data quality of ground-based systems e.g. the Protem 47 system (40 x 40 m

A033 INTEGRATED INVERSION OF CVES AND TEM DATA USING LATERAL CONSTRAINTS Summary 1 We present a mutually and laterally constrained inversion between transient electromagnetic (TEM) and geoelectric (DC) data sets. Although both methods measure in some sense the electrical resistivity or conductivity of the subsurface they sample different volumes and have different sensitivities. The different sampling volumes and different sensitivities are exploited by the mutually and laterally constrained inversion algorithm combining the coinciding profile data sets. The output model incorporates the information from both profile data sets to obtain the optimum layered 1D models fitting both data types. All data

2D laterally constrained inversion (2D-LCI) is a new approach to interpretation of continuous resistivity data. The 2D-LCI enables a layered 2D inversion with sharp layer boundaries as compared to traditional smooth inversion types. However, the full 2D solution is quite time consuming when calculating the derivatives in the Jacobian matrix. To reduce the computation times we suggest using 1D derivatives, Broyden's update formula (Quasi- Newton) or combinations of Broyden's update formula and 1D derivatives together with the full 2D solution. The reduction in computation time is between 7 and 30 times dependent on the combination of approximations.

In this paper we present how a 2D least squares algorithm, Res2Dinv (Loke and Dahlin, 2002), and the 2D Lateral Constrained Inversion (2D-LCI) algorithm (Christiansen et al., 2003) can be applied in geotechnical site investigations. The 2D-LCI algorithm performs a parameterized 2D Laterally Constrained Inversion (Auken and Christiansen, 2003) on resistivity data.

In Denmark, 99% of the water supply is based on groundwater abstraction. Increasing problems with the groundwater quality, due to urban growth and agricultural contamination, have caused the Danish Minister of Environment to approve a plan for improving ground water protection.

2D laterally constrained inversion (2D LCI) is a new approach to interpretation of continuous resistivity data using an inversion scheme similar to that used for 1D LCI. The only difference is that the forward code in the inversion scheme is 2D. This enables a layered 2D inversion in sedimentary areas where the layering has been disturbed in a degree, that a 1D model is inadequate, but a layered model is still applicable. The layered 2D inversion provides a full analysis of the model, and can include a priori information just as in the 1D case. Preliminary results on synthetic data show that a layered 2D approach enables a more detailed description of 2D structures, which the 1D formulation can not resolve. The model complexity is controlled and balanced by the lateral constraints on depths and resistivities.

Along North Eastern Coast of North Island (New Zealand) there is a true old-wood mine, which exploits 30.000 years old kauri trees, covered by very partieular terrains able to preserve them from insects, mushrooms and other parasites. In tact, the catastrophe, which collapsed them happened some 30.000 years BC, had to be really impressive this propaedeutical Note suggests some ideas to envisage reasonable geophysical models to picture a sudden event adequate to these exceptional evidences.

A field test and analysis method has been developed to estimate the vertical distribution of hydraulic conductivity in shallow unconsolidated aquifers. The field method uses fluid injection ports and pressure transducers in a hollow auger that measure the hydraulic head outside the auger at several distances from the injection point. A constant injection rate is maintained for a duration time sufficient for the system to become steady state. Exploiting the analogy between electrical resistivity in geophysics and hydraulic flow two methods are used to estimate conductivity with depth: a half-space model based on spherical flow from a point injection at each measurement site, and a one-dimensional inversion of an entire dataset. The injection methodology, conducted in three separate drilling operations, was investigated for repeatability, reproducibility, linearity, and for different injection sources. Repeatability tests, conducted at 10 levels, demonstrated standard deviations of generally less than 10%. Reproducibility tests conducted in three, closely spaced drilling operations generally showed a standard deviation of less than 20%, which is probably due to lateral variations in hydraulic conductivity. Linearity tests, made to determine dependency on flow rates, showed no indication of a flow rate bias. In order to obtain estimates of the hydraulic conductivity by an independent means, a series of measurements were made by injecting water through screens installed at two separate depths in a monitoring pipe near the measurement site. These estimates differed from the corresponding estimates obtained by injection in the hollow auger by a factor of less than 3.5, which can be attributed to variations in geology and the inaccurate estimates of the distance between the measurement and the injection sites at depth.

In time domain electromagnetic, TDEM, it is a wide spread practice to model the current waveform (e.g. Asten, 1987), where as the effect of low pass filters (LP) in the receiver system is neglected. In this paper we show that when the TDEM method is used in hydrogeological and environmental investigations it is mandatory to include the filters in the forward and inverse modeling scheme to obtain avalid geological model for the survey site. IDEM systems is normally band-limited as input signals from amplitude modulated radio transmitters operating in the target bandwidth of the system often dominate the noise characteristic of the system over several decades in time. Band-limiting the input signals significantly improves the signal-to-noise ratio and thereby the depth of exploration. However, the LP filters distort the early and medium part of the decay curve due to damping of the high frequency content. The level of distortion depends on several factors such as cut-off frequency and subsurface resistivities. Therefore, the LP filters must be incorporated in the inversion scheme such that the calculated forward response is subject to the same filtering as the measured response. In this paper we (1) quantify the distortion of responses due to band-limitation for a number of typical resistivity models. The band-limitations are based on cut-off frequencies typically used in commercial instruments. (2) We demonstrate that distortion due to band-limitation can be modeled in the inversion scheme by including the LP filters in the forward response. This approach is tested on theoretical data.

Since 1995, we have collected electromagnetic (EM) field data using a rotating azimuthal geometry analogous to that described for resistivity surveys by other workers. These EM azimuthal resistivity data are relatively easy to acquire and can be collected in areas with high electrode contact resistance. As an aid to interpreting field results, including apparent paradox of anisotropy conditions, we have used physical-scale modeling, and have run a suite of numerical models. Numerical and physical-scale modeling results indicate that the primary factor affecting the orientation of the apparent resistivity ellipse relative to a macroanisotropic linear conductor is the depth to the target (relative to the transmitter-receiver coil separation). At target depths less than about 10% of the coil separation, an electromagnetic paradox of anisotropy is observed in the modeling tank, in the numerical model, and in our field data (i.e., high values of apparent resistivity along the strike of the target). While physical-scale modeling is limited by the availability of materials of appropriate conductivity, numerical modeling provides a sufficiently extended parameter space for these analyses. The numerical model also indicates that the conductance of the target does not effect the orientation of the apparent resistivity ellipse.

Time domain electromagnetic soundings have been calculated over 2D / 3D structures in a layered halfspace and hereafter inverted using a 1D least squares time domain code. The forward calculations are done in the frequency domain and the subsequent transform to time domain is performed using fast digital filters. The study focuses on typical hydrogeological models. The investigated models include a resistive dyke, a resistive beam and resistive block structure with edges with varying slopes in a conductive layered halfspace with an overburden. The 1D interpretations of the 2D / 3D responses show that the vertical resolution is reduced and the resistivities are smeared out. The 2D / 3D responses over models with a resistive body in a conductive host is easily fitted with the 1D models whereas it is not possible to fit the responses from good conductors in a resistive host very well.

By the integrated use of geophysical methods a ltydrogeological investigation with a very high r~~solution of the geology has heen carried out in Denmark. llsing newly developed methods such as auger Ellog drilling with levrl specific Water sampling and pulled array contitttlous electric profiling together along with more Ouniliar techniques such as transient electromagnetic sounding we have perfortned a general exantinatiott of the vulnerability, the volumetric extent and the chemical properties of the aquifer itt the area. The protection of the aquifer was less than expected. There was a rorrelation between the water quality and the lithology of the aquifer.

In many applications the resolution of near surface geological layers are of great importance. A number of groundbased transient electromagnetic systems (TEM) provide sufficient early time data enabling such a resolution whereas this has not been possible with airborne TEM systems. Until now. In this abstract we describe the latest development of the SkyTEM system where early time data from as early as 8 microseconds (from begin of ramp) can be used even with a 500 m2 transmitter frame and a maximum magnetic moment of approx. 160,000 Am2. The system itself provides unbiased data from around 14 microseconds, but because the bias response at earlier times is 1) dependent of the transmitter coil geometry only, 2) it is constant from flight to flight and 3) it can be estimated from high altitude measurements, the amplitude of the bias response can be determined during the inversion process and thereby modeled. Using the Laterally Constrained Inversion algorithm to correlate the amplitude from sounding to sounding we show that data from as early as 8 microseconds from begin of ramp can be included in the inversion process. The inverted sections have been successfully validated against ERT measurements on the ground and 6 boreholes to a depth of 25 meters with both geological descriptions, resistivity logs and gamma logs. In the larger perspective the method can be used to accurately estimate near surface layers thus avoiding expensive and logistically challenging groundbased geophysical campaigns.

Groundwater in the Eyre Peninsula of South Australia is scarce with potable resources limited to the western coastal margin and the southern tip of the peninsula. Consequently an understanding of their extent has become increasingly important particularly with demand being close to current extraction limits. In September 2006, about 1000 line km of TEMPEST AEM data were acquired over the Southern Eyre Peninsula, in order to assist in the definition of freshwater lens systems and in particular aquifer bounds associated with them as part of a resource definition project. Following their acquisition, the TEMPEST data set was analysed for data quality and then transformed into conductivity depth images (CDI) using EMFLOW and subsequently using a smooth model inversion (see Fitzpatrick and Munday, 2007). In an effort to better define to better define the geometry of specific bounding surfaces of hydrogeological relevance the TEMPEST data were inverted through the application of the laterally constrained inversion (LCI) technique. This paper describes the initial results from the first application of the LCI to data from a fixed wing AEM system.

Since 2000 the Aarhus Workbench (aarhusgeo, 2008) is constantly developed to meet the research needs of the HydroGeophysics Group (HGG) at the University of Aarhus. It allows you to handle, process, invert and visualize electric and electromagnetic (EM) data on a common GIS platform. Geological data can be shown on the GIS map and on cross-sections for comparison with the geophysical results. The basic idea of the Aarhus Workbench was to develop a single and integrated software platform for handling a number of different data types. The Workbench uses an open-source client server database to manage data and settings. The benefits of using a databases compared to flat ASCII column files should not be underestimated. Firstly, user-handled input/output is nearly eliminated, thus minimizing the chance of human errors. Secondly, data are stored in a well described and documented format which is well suited for both exchange and storage of data. The Workbench allows the user to present the output of inversions as point themes or as color contoured thematic maps, such as mean resistivity slices, depth to a conductor etc. Models can also be shown on sections which are linked both to the GIS and to displays of data and forward data. The sections can contain numerous layers representing different data types. Over the years HGG has developed stable processing and inversion algorithms for airborne and ground-based EM data. The inversion is known as Laterally Constrained inversion (LCI) for quasi 2-D modeling and Spatial Constrained inversion (SCI) for quasi 3-D inversion. The Workbench implements a user friendly interface to these algorithms enabling non-geophysicists to carry out inversion of complicated airborne data sets without having in-depth knowledge about how the algorithm actually works. Just as important is an extensive system for evaluation of inversion results with plots of results like resistivity models, flight height, pitch, roll, residuals etc.

Over several decades, much research has been done to develop 3D electromagnetic inversion algorithms. Due to the computational complexity and the memory requirements for 3D time-domain electromagnetic (TEM) inversion algorithms, many real-world surveys are inverted within one dimension. To speed up calculations and manage memory for 3D inversions of TEM data, we have developed an approach using three uncoupled meshes: an inversion mesh, a forward-model mesh, and a mesh for Jacobian calculations. The inversion mesh is a coarse regular and structured mesh, such that constraints are easily enforced between the model parameters. Forward responses are calculated on a dense unstructured mesh to obtain accurate electromagnetic fields, whereas the Jacobian is calculated on a coarse unstructured mesh. We found that using a coarse mesh for the Jacobian is sufficient for the inversion to converge and, equally important, that it provides a significant speed boost in the overall inversion process, compared to calculating it on the forward-modeling mesh. The unstructured meshes are made of tetrahedral elements, and the electromagnetic fields are calculated using the finite-element method. The inversion optimization uses a standard Gauss-Newton formulation. For further speed up and memory optimizing of the inversion, we use domain decomposition for calculating the responses for each transmitter separately and parallelize the problem over domains using OpenMP. Compared to a 1D solution, the accuracy for the Jacobian is 1%–5% for the dense mesh and 2%–7% for the coarse mesh, but the calculation time is approximately five times faster for the coarse mesh. We also examined the algorithm on a small ground-based TEM data set acquired in an area where a 3D earth distorts the electromagnetic fields to such a degree that a 1D inversion is not feasible.

Groundwater in the Eyre Peninsula of South Australia is scarce with potable resources limited to the western coastal margin and the southern tip of the peninsula. Consequently an understanding of their extent has become increasingly important particularly with demand being close to current extraction limits. In September 2006, about 1000 line km of TEMPEST AEM data were acquired over the Southern Eyre Peninsula, in order to assist in the definition of freshwater lens systems and in particular aquifer bounds associated with them as part of a resource definition project. Following their acquisition, the TEMPEST data set was analysed for data quality and then transformed into conductivity depth images (CDI) using EMFLOW (see Fitzpatrick and Munday, 2007). In an effort to better define to better define the geometry of specific bounding surfaces of hydrogeological relevance the TEMPEST data were inverted through the application of the laterally constrained inversion (LCI) technique. This paper describes the applied inversion methodology and the initial results from the first application of the LCI to data from a fixed wing AEM system.

Summary Airborne Time Domain ElectroMagnetic (TDEM) and magnetic (MAG) surveys have successfully been used for environmental studies. In igneous provinces, TDEM and MAG data may be jointly interpreted for the characterization of the geological environment. In May 2015, a heliborne survey was conducted in New-Caledonia Island (south-west in the Pacific Ocean) over a small area nearby Kone for hydrogeological characterization. The area is hosted by pre-Late Cretaceous metamorphic unit, Cretaceous formations, two volcanic facies and peridotite massifs, relic of the ophiolite nappe. TDEM and MAG data were separately and jointly interpreted in order to characterize the Kone area. Separately, results obtained from the two methods provided new information on the hydrogeological environment and allow delineating the precise limits of the different geological formations. Jointly, the multilayered results were interpreted using a hierarchical ascendant clustering algorithm. Through a statistical classification of three MAG grids and three TDEM grids, it was possible to propose a predictive geological map for the Kone area. The input will help to better manage the groundwater resources and will be used to position new exploration wells.