- Home
- Conferences
- Conference Proceedings
- Conferences
Fifth International Conferention on Ground Penetrating Radar
- Conference date: 12 Jun 1994 - 16 Jun 1994
- Location: Kitchener, Canada
- Published: 12 June 1994
41 - 60 of 95 results
-
-
Evaluation and three-dimensional representation of ground-probing radar measurements
More LessGPR measurements are made in both boreholes and drifts for exploration. Measurements with directional antennae not only allow the distance to the reflection points to be determined but also the angle with respect to the azimuth of the plane defined by the reflection point and the two antennae. If a directional antenna cannot be used, the location of the reflection surface can be determined by special placement of the antenna in a vertical plane perpendicular to the axis of the drift. The migration method we use makes it possible to determine the point of reflection from the traveltime and the angle defined above. A planar element perpendicular to the normal to the reflection surface can be assumed through each point of reflection. If these elements are connected with each other, a zone is obtained representing the location of the reflector. The calculation of the angle, the filter processes, signal attenuation compensation, and other kinds of data processing, including determination of the traveltime ("picking"), is done using an interactive computer program we have developed. A software interface has been developed to transfer the reflection points to a 3-D graphics system so that projections, cross sections, and spatial representations can be prepared.
-
-
-
Processing ground penetrating radar data
Authors Steven C. Fischer, Robert R. Stewart and Harry M. JolGround penetrating radar (GPR) data are processed using seismic analysis techniques. Processing operations performed include signal saturation, spherical divergence gain recovery, spiking deconvolution, bandpass filtering, velocity analysis, elevation corrections, NMO correction, and f-k migration. As examples we first use 0.5 km lines of 50 MHz and 100 MHz, single-fold data, with a CMP gather from William River delta, Lake Athabaska, Saskatchewan. Next, a 120m, 100 MHz, single-fold line from Ft. Smith, N.W.T. is processed using similar steps. Finally, 100 m of 100 MHz, single-fold data, with a CMP gather, from Brigham City, Utah, are analyzed. Filter bandwidths chosen for the 50 MHz and 100 MHz measurements were 10/20-75/100 MHz and 20/30-150/180 MHz, respectively. Velocity spectra for NMO corrections were determined for William River delta and Brigham City based on CMP gather velocity semblances and interpreted arrivals. Velocities range from 0.3 m/ns to 0.07 ms/ns (air to wet sand respectively). Higher velocity surface statics, caused by dry topography, were corrected for after interpreting shallow arrivals. The Slave River velocities are similar to William River delta suggesting similar r'ock type with higher water saturation and/or lower porosity. Dry sands at Brigham City resulted in radar velocities of 0.14m/ns. Application of seismic processing techniques significantly improved these GPR sections.
-
-
-
A SFCW Polarimetruc Ground Penetration Radar
Authors Alan Langman and Michael R. InggsExtraction of target information from the measurement of the state of polarization of the scattered wave has been a promising area of research for a number of years. This paper discusses the theory and implementation of a stepped frequency CW polarimetric subsurface radar. The extended Prony method is used to obtain high resolution range profiles from which the scattering matrices are extracted. The results of laboratory trials are presented. These include measurements of the scattering matrix for a pipe and plate conducting targets, buried in a large sand box. Polarization responses for these targets are, plotted to emphasize the use of the scattering matrix for target discrimination.
-
-
-
Resampling in frequency domain, a method for interpolation of time series
Authors Rudolf Fruhwirth, Roswitha Müller and Rupert SchmöllerResampling is a process often used in geophysical applications. Especially in the field of seismic and radar data processing it is applied mainly to time series. In most cases the resampling process influences the contents of time series both, in the time domain and in the frequency domain. If a time series is resampled to a larger sampling interval the frequency spectrum will vary since the Nyquist frequency ifN) decreases and the frequency components higher then the new Nyquist frequency fold into the area below. Vice versa, by resampling to a smaller sampling interval, fNy is increasing and usually an unknown spectrum, depending on the used method will be added. Resampling is performed by interpolation. The simplest way is to use a linear function. On the other hand polynomial functions or cubic or rational splines lead to more satisfying results, but all of these methods influence the frequency spectrum. If the interpolation is performed in the frequency domain, the spectrum of the resampled data set can be controlled in such a way that it is unchanged. The principle of this method is based on the fact that the sampling interval of a time series in the time domain automatically defines the Nyquist frequency. But vice versa, in the frequency domain full control over the sampling interval is given by alteration Of/Ny' This work presents the above method and demonstrates that a discrete time series can be transformed into a continuous time series by moving/Ny to infinity and that it is very easy for that reason to resample to larger or smaller sampling intervals which are not integral multiples of the original sampling interval. The problem of aliasing is also discussed and practical examples for application of this method to GPR data are mentioned.
-
-
-
Image synthesis and perception of ground probing radar
Authors Ameen Maluf, Laeeque Daneshmend, Moe Momayez and Ferri HassaniA Ground Probing Radio Detection And Ranging (RADAR) system is used to extract subsurface scans. The principles and practices employed in this research relate to a Ground Probing Radar system operating with high frequency (lGHz) electromagnetic waves with adjacent transmitter and receiver antennas. Based on the theory of Kalman Filters, multi-sensor fusion is used to establish probabilistic models of individual sensor estimates and embed theses descriptions in a team-theoretic framework to finally describe the interactions between different sensors. Applied iteratively, consistent sensor estimates will converge to a qualitative image reconstruction. We investigate the possibility of deriving a functional model of image processing and representation to enhance the visual perception of subsurface images. The research is directed towards qualitative image integration and synthesis based on artificial neural networks pattern classification methods and front-end multi-sensor fusion techniques.
-
-
-
Signal and image classification in georadargrams
Authors Klaus-Peter Nick, Thomas Cimiotti, Uew Papzinger, Ljudmilla Kleinmann and Jrgen LaugksThe given tasks in our joint project are the detection and location of hyperbolas generated by pipes and cables in the subsurface when illuminated with a common pulse radar, and to find classification indications of the object's nature. Classification not only includes informations about the material and cross-section of an object, but also its three-dimensional extension in the subsurface. Thus, the processing of radar data is divided into several successive steps. First, the hyperbola has to be detected in the radargram and the approximate location of the pixels belonging to the hyperbola has to be detected. We perform this by means of the Hough transform algorithm. The result of the Hough transform and its degree of automation depend greatly on the preprocessing of the data transformed. To apply the correct preprocessing and processing strategy, classification of the radargrams by texture arialysis is investigated. After the detection of the hyperbola, an analysis of the reflected wavelets can provide some clues about the object's nature. Although absolute classification, which can be derived straightforwardly from simulations, will be successful only in a few exceptional cases, signal analysis is very useful in three-dimensional tracking of objects. We were able to demonstrate this by using different statistical correlation methods to identify whether hyberbolas in parallel transects were generated from the target or not. A simple but often successful method of detecting and following long objects is the presentation of parallel transects as a time slice.
-
-
-
Eccentricity-migration: A method to improve the imaging of pipes in radar reflection data
Authors Christian Stolte and Klaus-Peter NickHyperbolas are a widely observed feature in radar reflection data. Zero-offset migration of these hyperbolas rarely improves the image. Often energy is sme~red out and weak reflections disappear. The reasons are perturbations in the local geology influencing the velocity field, uncertainties introduced by uneven movement of the antennas, the bistatic character of the antenna as well as the geometry of the buried objects. In this work the target geometry is incorporated into migration to improve the subsurface image. An algorithm is presented, which automatically determines the diameters of buried pipes without prior knowledge of the medium-velocity for electromagnetic propagation. In a second step the algorithm utilizes these determined diameters of the buried pipes to enhance the radar image through migration. Conventional migration (KIRCHHOFF, f-k, FD) reconstructs the geometry of subsurface features from one parameter alone: the velocity field. Eccentricity-migration incorporates another parameter into migration: the deviating eccentricity of the hyperbola. Therefore, we chose the name "eccentricity-migration". The reflection image of a line diffractor in a radargram is uniquely characterized by its numerical eccentricity N E = V2, the ratio of the focal length of the hyperbola to the apexto- origin distance. For a pipe, N E is smaller than V2. Therefore, if the temporal and spatial resolution of the data are sufficient and changes in medium-velocity are small, then the pipe-diameter can be uniquely extracted from the shape of the reflection hyperbola alone. A further advantage of E-migration is its increased sensitivity to changes in velocity. Based on a known pipe-geometry, either a priori or extracted through E-migration, a more detailed velocity field can be derived and the subsurface image in the vicinity can be furthermore improved.
-
-
-
Recognition and removal of surface scattering in GPR data
Authors Roger Young and Jingsheng SunGround Penetrating Radar (GPR) data may show strong noise events due to surface scattering by obstacles such as trees and boulders. The surface scattering can be very strong due to low attenuation in the air. A procedure is developed for identification and removal of these events. It is based on the fact that the velocity of radar waves in the air is .3 mlns, but the velocity in the earth is much lower, approximately .1 mlns.
-
-
-
GPR frequency selection
Authors A.P. Annan and S.W. CoswayOne of the most important variables in GPR is selection of the operating frequency. Proper frequency selection can make the difference between survey success or failure. The most important aspect for a user is to understand the basic principles involved so that adequate survey planning can be carried out. Frequency selection depends on clearly defining the survey objective. This requires that the site and the survey conditions be evaluated prior to the survey. If not done, survey failure or success will be impossible to predict. Factors to be considered are site geology, surface obstructions, host geology, electrical properties, target size and depth of exploration. Given the above information, various constraints on frequency can be developed. From these frequency estimates, upper and lower bounds are placed on the frequency likely to be effective for the particular problem. In addition, the expectations of the end user of the survey results can be established thereby avoiding surprises.
-
-
-
Ground penetrating radar considerations for optimizing the data collection scenario
More LessAs research utilizing ground penetrating radar (GPR) systems is undertaken a wide range of problems and conditions is undertaken. Typically, before research of this type is funded, the client needs some degree of assurance that the research in question has a "good" chance of success. In this paper we address the topic in three steps. First, we discuss what one can reasonably determine from the client's description of the problem, utilizing pictures, sketches, etc., which indicate the chances of successfully solving the problem. Second, we discuss the minimum measurements that can be made on-site in order to improve the probability of success without actually performing the survey. Finally, we discuss the procedures/methods which we have developed which directly relate to the problem being addressed so that success will be assured while keeping the data collection time, the data processing time, and the data analysis time all listed below, to a minimum.
-
-
-
Detecting and mapping recent faults with a ground penetrating radar in the alluvial fans of the Arava valley, Israel
Authors Uri Basson, Yehouda Enzel, Rivka Amit and Zvi Ben-AvrahamGround Penetrating Radar (GPR) field experiments were conducted to detect recent faults that may be connected to the Dead-Sea Rift, on several all uvial fans in the southern Arava Valley, Israel. The existence of fault-scarps and lineaments interpreted as surficial faults there, indicated the Late Pleistocene and Holocene seismic activity of the rift. High resolution GPR profiles were carried out to map the shallow stratigraphy of the alluvial deposits and to detect subsurface recent faults. The GPR experiments were a part of integrated neotectonic research, which combines geophysical and geomorphological complementary methodsto investigate recent faulting events. During the study we used GPR and trench excavations to detect and study faults which rapture the surface, and faults with no visible surficial expression. The profiles that were conducted to cross the faults perpendicularly have produced reflections that were not sufficiently different from reflections of stratigraphic discontinuities that are common in the alluvial fans. On the other hand, a dramatic indication interpreted as a fault has been obtained in the profiles that were conducted along a wide faulted zone discontinuity, and along a strike-line of a fault, which is significantly wider then other lateral discontinuities. Consequent to locating a suspected faulted zone, high-resolution GPRmapping of its typical stratigraphy can be made. In spite of the complex stratigraphy of the alluvial fans, which produces abundance of pitfall reflection anomalies, we introduce a powerful way to identify the faults with GPR. The results of the GPRmapping can help focusing the trench excavations that are essential for understanding the details ofthe paleo seismic faulting events. Prospecting offaulted zones with GPR can reduce the excavation activities to a minimum, save time, effort and damage to the landscape.
-
-
-
Erosion of ejecta at meteor crater: constraints from ground penetrating radar
Authors John A. Grant and Peter H. SchultzGround penetrating radar (GPR) provides a rapid, non-destructive means for investigating the shallow subsurface around a variety of terrestrial landforms also found on other planets. Impact craters represent one of the most ubiquitous planetary features and GPR data at Meteor Crater, Arizona, delineate stratigraphic relationships between the ejecta and associated alluvial deposits in the shallowest 1-3 m ~f the subsurface. Analyses of these data indicate that the continuous ejecta remain better preserved than was defined previously via surficial mapping. Radar penetrates a thin veneer of alluvium and colluvium that masks the actual distribution. Returns from uniform, finegrained alluvium filling regional drainages blocked by the crater confirm their generally low transport capacity through time. GPR data and sampling further reveal that some subtle ridges extending from the crater are not the result of debris flow deposition of eroded ejecta. Instead these ridges are formed by preserved ejecta deposits draping pre-impacttopography. Collectively, these results establish that the continuous crater ejecta are in a surprisingly pristine state of preservation in contrast with visual inspection. More generally, this study demonstrates the high potential of GPR in future planetary applications. Ease of miniaturization makes GPR ideally suited for planetary rover missions to the terrestrial planets and Galilean satellites where it could help resolve a range of questions relating to their histories of surface processes.
-
-
-
Three-dimensional geometry of fluvial gravel deposits from GPR reflection patterns; a comparison of results of three different antenna frequencies
Authors Peter Huggenberger, Edi Meier and Beres MilanGround-probing radar (GPR) proftles have been recorded in fluvial gravel deposits to investigate the two and three dimensional aspects of inhomogeneities. Because reflection patterns depend strongly on the frequency of the input signal, the measurements have been perfonned with a variety of antenna frequencies: an OYO Georadar 1 system with a pair of 250 MHz antennae and a Sensors & Software Inc. pulseEKKO IV system with pairs of 100 MHz and 200 MHz antennae. The GPR profiles which have been recorded with different frequencies are compared. The reflection patterns on the GPR proftles illustrate different aspects of subsurface inhomogeneities. It is demonstrated that the 250 MHz antennae of OYO and the 200 MHz antennae of pulseEKKO IV portray the small scale variability of the gravel deposit, whereas the reflection proftles recorded with the pulse EKKO IV 100 MHz system show reflection patterns which can be assigned to the principal set-boundaries of the sedimentary units. Recently developed software packages provide some good processing opportunities such as filtering, trace editing and migrating in the time domain. The 3-D representation of GPR sections enables the recognition of the geometry of sedimentary units. This information could be useful in the process of quantification of geophysical data as input in hydraulic models.
-
-
-
Modelling GPR reflections from magnetic susceptibility and electrical conductivity variations
Authors O. Lzaro-Mancilla and E. Gmez-TrevioTypical GPR sections show a multiplicity of reflection signals, whose origin is usually considered to be related solely to variations in the electrical pennitivity of soil materials. The possibility that some of these reflections could be related to variations in electrical conductivity, is commonly ruled out on the basis of the operating frequency of current GPR instruments. Magnetic susceptibility variations , on the other hand, are commonly neglected because in most situations they affect velocity much less than electrical permitivity . In general, however, neither GPR operate at their optimum frequency nor can magnetic susceptibility be neglected at large. The objective of the presented work is to asses the relative importance of the last two properties in GPR studies. In general, actual field conditions usually fall between two extremes . One is the case of the very conductive ground where GPR is not applicable; the other is a completely transparent medium with perfect reflections. The intermediate cases may include very resistive media but with high enough conductivity contrasts to produce measurable reflections. We test this hypothesis along with the possibility that some of the observed reflections in GPR sections could also be due to natural- magnetic susceptibility variations. We consider a plane-wave source in the presence of a horizontally layered model whose three electromagnetic properties vary from layer to layer. Our results indicate that reflections due to electrical conductivity and magnetic susceptibility variations, can under natural conditions be of the same order as those associated with electrical pennitivity.
-
-
-
Ground penetrating radar of lakeshore spits in nothwestern Saskatchewan, Canada: variable internal structure
Authors Harry M. Jol, Derald G. Smith and Richard MeyersGround penetrating radar (GPR) field experiments were carried out on three different lakeshore spits located in remote northwestern Saskatchewan, Canada. The objective was to test present ideas regarding the origin of spits, particularly thickness, depositional migration direction and lake level stability. To determine depositional processes and direction of the spits, we first identified radar facies (groups of reflection with similar patterns) which from previous experiments have shown correspond with internal sedimentary structures and facies. From the results we have identified two different depositional processes which form spits not yet reported in the literature. Big Buffalo spit (4 km long) in Peter Pond Lake shows dipping reflections near the spit end indicating active longshore processes and overwash deposits. Ted Island spit, in Churchill Lake, is a complex 350 m long spit with internal reversed cross stratification sequences which infer alternating depositional trends from two different directions. The Big Island spit in Lac Ile-a-la-Crosse forms on the south side of a dumlinoid-shaped island and is 650 m long. Reflections show a deep infilled basin between the island and spit end, as well as dipping reflections (5_9) within the recurved spit end. Loss of signal return from below the spit strata in all three spits is attributed to a lithofacies change from sand to mud or diamicton.
-
-
-
Ground penetrating radar investigation of a wetlands replacement site
Authors John Madsen, Susan McGeary, David Krantz, Samuel McIntire, William Daniels and Curtis StorlazziGround penetrating radar (GPR) techniques were used with conventional lithostratigraphic information and water table elevations to develop a model for the hydrogeologic setting of a -0.5 km2 wetlands replacement site in southern Delaware, USA. The GPR profiles are a valuable component in the modeling because they provide a continuous, high-resolution image of the subsurface that cannot be provided by conventional coring and piezometric approaches. GPR data was gathered using a Sensors & Software, Inc. pulseEKKO IV unit using 50 and 100 MHz antennas. Over 2.3 km of common offset profiles were collected using the 100 MHz antennas with a transmitter-receiver separation of -1 m (3 ft) and a station spacing of -0.3 m (1 ft). 50 MHz data was also gathered to compare subsurface images using lower frequency antennas. Commonmid- point records were collected at -60 m (200 ft) intervals along the profiles. The quality of data collected with the GPR is excellent. The upper portions of the profiles contain a water table reflection (-1 m depth) that can be traced throughout the site.. Below the water table coherent reflections were obtained to depths greater than 10m. We have collected 7 vibra-cores to depths ranging from 2 m to 6 m along the profiles to correlate GPR reflections with changes in subsurface lithologies. Sediments encountered during vibra-coring consist primarily of fine to medium sands, silty clay sand, silty clay to clay, and a thin layer or quartz granules to pebbles. In addition to the vibra-cores conventional subsurface hydrogeologic information at the site includes sediment samples from borings, hand-auger cores, and water table elevations from piezometers.
-
-
-
Internal structure of a pacific coast barrier spit using ground penetrating radar
Authors Richard Meyers, Derald Smith, Harry Jol and M.B. HayGround penetrating radar (GPR) field experiments were carried out on the modem barrier spit at Willapa Bay, Washington (north ofthe Columbia River mouth). The barrier spit is influenced by a 3.6 m tidal range (Spring), as well as high wave energy and longshore transport depositional processes. The spit is 36 kIn long by 2-5 kIn wide and is composed of fine-grained beach and eolian sand. The spit has a freshwater aquif~r recharged by rainfall. To determine paleodepositional processes of the barrier spit, radar facies (reflection patterns) which correspond with sedimentary facies and depositional' processes were identified. Next, interpretations of depositional processes using the geomorphic association between facies and process were made. The objective was to reconstruct the overall sequence of paleo-depositional patterns and processes which have formed the Willapa barrier spit during the late Holocene (last 5000 years). A puiseEKKO™ IV radar system was used with 25, 50, 100 and 200 MHz antennae. GPR transects were carried out along beaches and roadways across the spit and where necessary were corrected for topography. Processing and plotting of the radar profiles was carried out using puiseEKKO™ IV software packages. GPR surveys indicate a shingle-like accretionary depositional pattern of beach and shoreface reflections which dip towards the ocean at about 1 degree. At the distal portion of the barrier reflections dip in the direction of longshore transport (north). Nearby radiocarbon-dated wood in beach sediment suggests that the spit began to form 4000 years ago. The loss ofradar signal from below 10-15 m deep may indicate a major lithofacies change. Very difficult penetration by vibracoring of spit sediment suggests that GPR reflections may be partially attributed to the tight packing of the individual sediment particles along nearly straight inclined reflections, as well as slight changes in sediment grain size at bedding planes deposited by major winter storm events.
-
-
-
Ground penetrating radar in a hydreological investigation of the Oak Ridges Moraine, Ontario
Authors J.A. Pilon, H.A.J. Russel, T.A. Brennand, D.R. Shapre and P.J. BarnettThis paper presents a summary ofthe results obtained with Ground Penetrating Radar at three ofthe eleven sites surveyed during the summers of 1992 and 1993 as part of the hydrogeological investigation ofthe Oak Ridge Moraine. After topographic correction ofthe data, we were able to accurately determine the location ofthe water table. Areal surveys at the St. John and Bolton Fann sites allowed comprehension of changes in water table topography. In addition, we obtained substantial subsurface structural information which improves our understanding of sediment architecture and genesis. The data from the Brighton and Bolton Fann sites were acquired near aggregate quarries which allowed direct correlation between the radar records and the sedimentary exposure. The results presented here highlight the level of geological details detected by the GPR surveys. Even though further studies will be conducted in 1994 and 1995, we are already able to demonstrate that GPR is an excellent geophysical technique for geologic (textural and structural), and hydrogeologic studies in appropriately resistive surficial sediments.
-
-
-
Application of ground penetrating radar in coastal stratigraphic studies
Authors Sytze van Heteren, Duncan FitzGerald and Paul McKinlayIn recent years, ground-penetrating radar (GPR) has been applied in a rapidly increasing number of stratigraphic studies. Its ability to produce continuous subsurface profiles (under favorable circumstances) has resulted in a better understanding of the shape, extent, and internal structures of facies belonging to several different depositional environments. Thus far, however, GPR has been used in few coastal regions, mainly because of the widespread occurrence of signal-attenuating units. Facies producing signal attenuation include fine-grained estuarine and lagoonal clays and coarser-grained units that contain saltwater. The present study, utilizing a Geophysical Survey Systems Inc. System 3 GPR, focuses on a number of barrier beaches and spits along the New England coast. Most of these barriers are sandy to gravelly; they include a reservoir of fresh groundwater that increases in thickness toward the center of the barrier. To a certain degree, the volume of fresh water is related to the size of the barrier. The occurrence of fresh groundwater in a mainly sandy matrix has turned out to be favorable for GPR application in several instances, with the best results being obtained in the centermost portions of wide and high barriers. Signature quality decreases rapidly toward both the sea and the back-barrier marshes and tidal flats. In both directions, the freshwater-saltwater boundary is increasingly closer to the surface. GPR has proven to be useful in several aspects of the study of coastal barriers. It is an important tool in identifying and characterizing different stratigraphic units that make up a barrier. GPR is also highly useful in determining sediment budgets of coastal systems. Finally, GPR can be applied to determine the occurrence and the maximum extent of peat layers underneath a barrier. Interpretation of GPR profiles, in conjunction with more conventional methods of barrier study, provides a data base that can be used to analyze a range of physical parameters which may have left their mark in a barrier, such as relative sea-level changes, paleotidal range, sediment source and supply, and the effect of storms.
-
-
-
Trial of tunnel radar for cavity and ore detection in the sudbury mining camp, Ontario
Authors P.K. Fullagar and D. LivelybrooksThe utility of ground penetrating radar (GPR) in metalliferous mines for cavity detection and orebody delineation is under investigation at Ecole Polytechnique. As a first stage in this investigation, fixed offset Pulse Ekko IV radar profiles were read in underground mine drifts at Sudbury in August 1993. At Stobie Mine a 3.7m diameter ventilation shaft was detected unequivocally using unshielded 100MHz antennas, at ranges between 12 and 16m, and an orepass was located at a range of 10m. This demonstrated that GPR is effective in 5m x 5m mine drifts, provided target travel times are greater than about 80ns, corresponding to a range of approximately 5m in Stobie metasediments and greenstones. Reverberation within the drift could obscure reflections from closer targets. Attempts to map massive sulphide ore contacts within 10m of drift walls at Lower Coleman Mine were unsuccessful. This outcome is attributed to sulphide blebs and stringers in the host rock which increase the intensity and longevity of reverberations within the drift (by enhancing the reflectivity of the walls) and which also scatter and attenuate any radar signals propagating to and from the contact. "Geologically minor" sulphide concentrations can thus seriously compromise the effectiveness of radar. Further ore delineation work should be focussed on massive sulphide bodies which make sharp contact with barren host rocks.
-