The development of transient antennas for Ground Penetrating Radar would be difficult without numerical simulations of their performance. To this end FDTD code should be considered as a useful and powerful tool for transient antenna analysis. As an example of such analysis we describe simulations of the dielectric wedge antenna. Detailed investigations on such key model elements as the staircase approximation of 3D structures, the antenna feed model, the cell size and the size of the domain of computation were carried out. Very good agreement between theoretical predictions and experimental results was observed. Some of antenna characteristics (like the radial component of the electric field in the near zone, the radiated waveform and the antenna footprint in different grounds) cannot be measured easily, so the correct theoretical (numerical) model is essential for their determination. The numerical model of the antenna can be also used for optimization of the antenna.

To understand better how a borehole antenna radiates radar waves into a formation, this phenomenon is simulated numerically using the finite-difference, time-domain method. The simulations are of two different antenna models that include features like a driving point fed by a coaxial cable, resistive loading of the antenna, and a water-filled borehole. For each model, traces are calculated in the farfield region, and then, from these traces, radiation patterns are calculated. The radiation patterns show that the amplitude of the radar wave is strongly affected by its frequency, its propagation direction, and the resistive loading of the antenna.

Antenna is considered to be the most critical part of any OPR system. AithoLigh a lot of work had been done in the field of antenna modeling and design, GPR antenna measurements and their comparison still do not have reliable methodological basis. This work is devoted to the comparison of performance of different GPR antennas used in step-frequency radar. To this end, we employed technique based on radar calibration and subtraction of a dominant response usually represented by the ground return. The residue is further transformed into synthetic range profile and serves as a measure of the antenna non-ideality. A metal sheet shifted from the calibration position has been used as a test object. In this case, subtraction of the dominant reflection leaves only clutter components in the received signal. Discrimination of small shallow underground object can also provide means for the antenna comparison. The more ideal is the GPR antenna, the more effective is the calibration/subtraction procedure. Hence, small shallow target can be detected or not depending on the antenna performance and this is demonstrated in several experimental examples.

In this paper we present the electric field integral equation (EFIE) method for the analysis of transient metallic antennas above a lossy half-space. This formulation is suitable for the analysis of a wide class of metallic GPR antennas that does not contain any dielectric material. Ground influence is properly taken into account by the Green's function for layered media. The EFIE is solved numerically by the method of moments using the wellknown Rao-Wilton-Glisson basis function. For GPR antennas, the free-space Green's function is replaced by the dyadic Green's function for layered media. The time-domain solution for transient antennas is obtained using Fourier transform method and a time-window technique. By using time window to remove reflections from antenna open ends, the solution for transient antennas is obtained. To speed up computations, several approaches are used, which include the exploitation of antenna symmetry, a special treatment for integration over singularity, and an interpolation method to expedite the evaluation of Sommerfeld integrals. We have found that those approaches allow rapid and accurate computations for obtaining the time-domain solution. Good agreement between the computation and measurement of the input impedance of a wire dipole and a wire bow-tie antenna is obtained.

In this paper a comprehensive characterization of transient bow-tie antennas for GPR applications is presented. In particular the characteristic impedance of a transient bow-tie antenna above different types of ground, as a function of its flare angle is given. The time-domain reflection coefficient of the transient bow-tie antenna above the ground is derived and also plotted as a function of flare angle. This result shows that there exists an optimal flare angle which gives minimal reflection at the antenna terminal for various ground types. Additionally, here we also investigate the characteristic impedance as a function of antenna elevation above the ground. Moreover, in this paper the subsurface footprints of bow-tie antennas as a function of flare angle for different ground types are shown. This result shows that depending on the target's size, the size of footprint can be adjusted by varying the flare angle in order to reduce clutter from ground surface. The analysis is carried out in the frequency-domain using the mixed-potential integral equation (MPIE) method for problems in multi-layer media, which is solved numerically by the method of moments. The solution in the time domain for transient antennas is obtained using the Fourier transform method and a time window technique to remove open-end reflections.

In the paper we have derived the following equation 4 f0 Lequi = c/sqrt(?equi ) where f0 is the resonant frequency of the antenna input impedance, c is the speed of light, Lequi and equi are respectively the equivalent length of antenna and the equivalent dielectric constant of the medium formed by halfspace ground and half-space air. When measuring in air, the dielectric constant is known. One can obtain the antenna equivalent length from the measured fo, by using the above equation. When Lequi is obtained, one then can calculate ?equi from a measured f0, for the case when placing antenna on material interface. One can further estimate the ground dielectric constant by using the equation: ?equi = (?air + ?ground )/2. Bow-tie antennas with different flange angles are constructed for experiments, in order to discuss the application of the above equation. The lengths of those antennas are all 30cm (length of the antenna conductor: 15cm). But the flange angles of the antenna conductors are respectively: 15, 40, 60, 75, 90 degrees. The antenna input reflections are measured over the frequency band up to 500 MHz, for cases when the antennas are placed in air and on sand, stone and water surfaces.

The bowtie antenna provides a solution to the requirements of a broadband non-dispersive antenna for low frequency pulsed ground penetration radar (GPR) applications. This paper reports on a bowtie antenna numerically optimized with respect to bandwidth and radiation performance into the ground over the 10-100 MHz band. Rounding the edges of the bowtie was found to stabilize the antenna performance by reducing the internal angle dependence of the 1St resonant frequency. Half hemisphere radiation was achieved using a nonabsorptive cylindrical cavity placed over the rounded geometry bowtie. Late time ringing due to reflections was minimized by the use of edge termination resistors connected between the cavity wall and the bowtie. The antenna match to the ground was improved using an optimized low permittivity polypropylene slab. This restored the radiation pattern distortion due to the presence of the ground and stabilized the input impedance for fluctuations in ground parameters. This also permitted the antenna to be wheel mounted at various heights above the ground with minimal influence on the antenna performance.

Ground Penetrating Radar (GPR) images highly depend on the relative geometry existing between the transmitting and receiving antennas and the depth reflectors. The resulting variations are rain1y due to the directional properties of the antennas and also to the sensitivity of the considered reflector to the polarization of the incident electromagnetic wave. In the present study, GPR data sets have been recorded using several 100 MHz antenna configurations, including Transverse Electric (TE) and Transverse Magnetic (TM) modes. In addition, a polarization analysis has been conducted by surveying twice the studied profile in both modes (TE and TM), using parallel and perpendicular antennas. The obtained images display a high complementarity that provides more details on the studied geological structures when compared to an image derived from a single conventional acquisition configuration. These studies emphasize how multi-configuration antennas surveys have the potential to improve GPR imaging and interpretation. Furthermore, polarimetric surveys have been carried out in order to study the possible link between interface depolarization phenomenon and phase wave inversion and amplitude wave decrease observed for TE Common Mid Point gathers. The latter would induce disturbances of the GPR images when stacking procedures are used for multi-offsets surveys.

Arrays of phase-shifted antennas, made of several singleelement antennas, are commonly used in air radar and communication systems to obtain very directive radiation patterns. In this paper, we propose to apply this technology to GPR. Firstly, a theoretical study of linear arrays is presented, allowing to determine the influences of the element number, their spacing and phase-shift. Secondly, field experiments have been carned out, consisting of Common Receiver Gathers with two single 100 MHz antennas, in order to reconstruct sections acquired with arrays of 21 transmitters and a single receiver. The results lead to greater signal to noise ratios and penetration depths. Lastly, with the aim to build an actual array of transmitters, antennas of several kinds have been made and tested.

The use of bow-tie antennas with time domain GPR systems has become quite common in civil engineering for investigation of structural integrity and location of defects in concrete structures. The depth of a sub-surface feature is evaluated by measuring the time taken for a signal pulse to reflect back from the feature and return to the surface antenna. Optimum interpretation using this approach thus depends upon an accurate knowledge of the speed of propagation of the radar signal, which is very sensitive to the moisture condition of the concrete. The amount of moisture in concrete is usually unknown, variable and very difficult to assess rapidly & non-destructively for in-situ concrete. The paper discusses the development of a wide-band system at Liverpool University, which will be used to make measurements of the complex permittivity and conductivity of concrete structures at radar frequencies, thus enabling a detailed knowledge of the variation of moisture content with depth to be acquired. Instead of using a conventional bow-tie antenna, a much wider-band TEM horn antenna has been developed and can be used in conjunction with a battery-powered portable network analyser suitable for use in the field. Results are presented showing the performance of the new antenna over a wide frequency range. Laboratory tests are discussed in which a variable moisture profile in concrete can be reliably simulated using discrete layers of dielectric material to obtain a closely stepped profile, that closely approximates to the desired linearly varying dielectric model.

The importance of archaeological heritage justifies looking for new techniques and methods which allow their knowledge in a more exhaustive way. We are not only talking about detection, but also about remains geometry and construction details. Cultural heritage record documents should include all possible information and the collection of this non-destuctive techniques information is recommended (Neubauer, 2001). In this work we show the preliminary results obtained applying three techniques at an archaeological site in Galicia (Spain), in order to document the remains of a megalithic tomb. First of all, a full topographic total station survey was made to obtain a digital terrain model of the studied area. The GPR investigation was made with Zond- 12c equipment operating with a 900 MHz antenna, radargrams were corrected with the digital terrain data attained hefore. The results showed a very shallow reflector on the top of a small hummock (15 m diameter, 3 m high), very close to an emerging flagstone which could be a part of the tomb. Excavation makes evident the presence of some other flagstones of the tomb at this point. The full archaeological site was excavated and a close-range photogrammetric study was made to obtain a cultural heritage record document including all possible metric information of the remains. A calibrated digital camera was used to obtain the spatial representation of the tomb. This information may be used in the future to reconstruct the tomb in another place, because the contruction of a new highway crossing at this archaeological site is going to take place at some future stage.

A detailed GPR survey at the Mahram Bilqis was undertaken to investigate the architectural features buried beneath the desert sands. GPR used in combination with trenching and hand augering for depth verification and material identification, was found to be very effective at mapping the three dimensional location of buried buildings and other subsurface architectural and sedimentary structures. The unique characteristics of the sediment covering the site and the scale of the architectural elements (above and below the surface) presented a number of issues in designing the survey and interpreting the data. The GPR profiles revealed a variety of different structures to a depth of 8 m. The extensive survey and multitude of subsurface features enabled a radar stratigraphic analysis at the site. This was undertaken to classify features according to their geophysical character.

GPR surveys were conducted at two sites within the hillside town of Rennes-le-Château, France. The town is linked with many traditions and mysteries associated with the Templar movement and its treasure, including according to some documents, the possible location of the Holy Grail. Two locations were chosen for the initial project: the Tour Magdala and the Church of St. Mary Magdalen. The survey at the Tour Magdala was carried out to image any features that may be located beneath the tower floor or around its outer base. Results indicate the tower is built on the local bedrock with possible surface and subsurface disruptions in the local stratigraphy, while 3D cubes show a hyperbolic feature, which may indicate the possibility of a buried feature. The survey at the Church was carried out to image any features that may be located beneath the church floor. 2D and 3D images show a subsurface anomaly (hyperbolic feature) that extends along several parallel lines that may possibly indicate a burial crypt.

The Dead Sea Scrolls are one of the greatest manuscript discoveries of the twentieth century. Since 1947 the Qumran region, the site of the Scrolls discovery on the northwestern shores of the Dead Sea, Israel, has been subject to countless probes. In 2001, ground penetrating radar (GPR) was used in an expedition that was initiated to better understand the Qumran site. Primarily, GPR was used to aid in identifying unmarked graves in the Qumran cemetery, and secondly, to determine if there were more caves in the marl cliffs that might contain artifacts associated with Qumran. In regards to the first GPR objective, two patterns emerged as burial signatures - a hyperbolic feature and/or a "V" shape. An extensive GPR survey was conducted along the outer edges of the presently exposed cemetery as well as empty patches of ground within the present cemetery. Over 100 potential graves were located that did not show surficial expressions. The second objective was achieved by running GPR surveys along the cliff faces and tops. Two sites were then chosen for excavation based on GPR images that showed hyperbolic features between 0.5 and 1 .0 m depth. Artifacts were recovered at one site.

A GPR survey at Muweilah in the United Arab Emirates capably identified archaeological features such as walls and floors to a depth of several metres. Surveying conditions were favourable with sand cover allowing good radar penetration and strong reflections from targets. A feature of this work is that three-dimensional migrations of two dense data sets provided plan view slices at 10 cm intervals, in a manner which mirrors the excavation methodology employed by archaeologists. Presented also are problems and limitations encountered throughout the survey and subsequent data processing. It is conjectured that data quality is significantly improved when the area is flattened and de-vegetated prior to surveying. Although much of the interpretation is subject to ground-truthing, the preliminary analysis is extremely encouraging. It is observed however, that interpretation is complex and ambiguous if undertaken in isolation. Reliance on a conceptual archaeological model, a combination of other geophysical approaches, and ongoing ground-truthing are thus strongly recommended.

Barcombe Roman Villa, a site under threat of destruction from ploughing, has been investigated by use of resistivity, field walking and excavation. GPR survey was added to these techniques and a comparison is made between the composite picture provided by the former and time slices derived from the latter.

The villa of the Roman Emperor Marcus Ulpius Trajanus (AD52-117) was built in Arcinazzo Italy, approximately 55 km northeast of Rome. Today, the only remains left standing at the site are public building entrances comprising a small portion of the entire site. As part of an ongoing study to rescue this national archaeological treasure, an extensive grid system was laid out and high-resolution GPR surveys using sub-meter profile spacings were conducted. Amplitude time slice analysis indicates that many structural foundations of the villa are still well preserved below the ground surface. Time slices below 40 ns in one area reveal several large mushroom shaped structures enclosed within a large building over 100 meters in length. These rounded structures are believed to be dipping pools within the bathhouse to the villa. At a location west of the bathhouse, a large oval shaped anomaly 45 meters along its major axis was discovered. Several intermediate time slices show a very faint overlapping oval reflection with a different orientation. The fainter anomaly may indicate the initial construction geometry was adjusted soon after construction began on the site for a garden pond or outdoor pooi area. The data are also examined using fast animation of the radar time slices . In this dynamic display of the datasets, several other information such as overlying topsoil depths across the site, and the relationship of structural anomalies at different levels, can be easily visualized when compared to normal static displays of radar data.

We developed an interferometric borehole radar system, which has one transmitter and two receivers. A new analog optical link is equipped in this new borehole radar system. Field experiments were carried out and the system was evaluated. Clear reflections from subsurface fractures were observed. Two radar profiles can be obtained by one measurement. These two radar profiles are acquired by two receivers spacing 1 m apart, therefore these profiles can be used for interferometry. The difference in these profiles is caused only from the different of the wave paths from the radar target. By interferometric interpretation, much detailed information can be obtained from borehole radar compared to the conventional technique. For example, the direct coupling component of the received signal can be used for electromagnetic well logging.

We have carried out borehole radar (BHR) surveys at gold mines in the Witwatersrand Basin. South Africa in order to map the Ventersdorp Contact Reef (VCR). In one such survey 20 kW transmitter and receiver pairs, 32 mm in diameter, with a bandwidth of 10-125 Mllz were used to profile a ~300 metre long section of the reef from a borehole that intersected it at an angle of 26°. Structures on the VCR were visible to a distance of 8Om, before noise started to dominate the signal. We established that the V.C.R. is sufficiently reflective, and its host rocks are transparent enough to open not only the certainty of high resolution echo sounding along the nadir line, but also the possibility of mapping off-axis back-scatterers by applying modified SAR reconstruction techniques to VHF BHR data. One of the problems facing synthetic aperture borehole radar is that it is difficult to build thin, efficient, directional radar antennas. Thin borehole radars are cylindrically omnidirectional and cannot be used to distinguish left from right. In this paper we show that borehole curvature can be used to address the difficulty of determining on which side ofthe survey line a backscattering object might lie.

A new high-resolution directional borehole radar-logging tool (DBOR tool) was used to log three wells at the Idaho National Engineering and Environmental Laboratory (INEEL). The radar system uses identical directional cavitybacked monopole transmitting and receiving antennas that can be mechanically rotated while the tool is stationary or moving slowly in a borehole. Faster reconnaissance logging with no antenna rotation was also done to find zones of interest. The microprocessor-controlled motor/encoder in the tool can rotate the antennas azimuthally, to a commanded angle, accurate to a within few degrees. The three logged wells in the unsaturated zone at the INEEL had been cored with good core recovery through most zones. After coring, PVC casing was installed in the wells. The unsaturated zone consists of layered basalt flows that are interbedded with thin layers of coarse-to-fine grained sediments. Several zones were found that show distinctive signatures consistent with fractures in the basalt. These zones may correspond to suspected preferential flow paths. The DBOR data were compared to core, and other borehole log information to help provide better understanding of hydraulic flow and transport in preferential flow paths in the unsaturated zone basalts at the INEEL.

Analyses of traveltimes and amplitudes of crosshole georadar data provide estimates of the electromagnetic velocity and attenuation of the probed media. In contrast to inversions of traveltimes, which are well established and robust, ray-based inversions of amplitudes depend critically on a priori assumptions about the directive properties of the antennas. In particular, the influence of electric material property fluctuations on antenna performance may lead to serious distortions of the radiation pattern. Such distortions cannot be accounted for by currently employed ray-based amplitude inversion algorithms. To explore the problem of antenna coupling to local heterogeneities, we generate synthetic crosshole georadar data for a suite of stochastic models using a finite-difference time-domain (FDTD) solution of Maxwell's equations in cylindrical coordinates. Analyses of the radiation patterns extracted from the synthetic data indicate that distortions of the radiation patterns are primarily due to propagation effects and not to dipolecoupling effects. We do, however, find that the quality of the amplitude tomograms diminishes rapidly with increasing heterogeneity, probably because of inherent inadequacies in ray-based inversion methods. In contrast, the quality of ray-based traveltime tomograms is surprisingly high, even in the presence of strong heterogeneity.

In this paper, we demonstrate three-dimensional estimation of target positions with borehole radar in single borehole. We developed array type borehole radar having four dipole antennas with optical modulators. Experiment at a model site was done using the developed radar system. In the site, one borehole, in which there is no fluid, and one cave exist. Receiving dipole antennas and a transmitting antenna are set in a single borehole. The 3-D MUSIC algorithm modified for array type radar in a borehole is applied to the measured data to estimate location of the cave. 3-D estimation results from the experiments were reasonable. In the estimation, Method of Moment (MoM) was used to model the antennas in a borehole. Green's function used in MoM was modified to include influence of cylindrical boundary around antennas in a borehole. Finally, we did experiments in granite, where boreholes are filled with fresh water. Cross-hole measurement was carried out to investigate characteristic of the dipole array in a water-filled borehole. Cross-spectrum analysis was adopted to obtained data, and we found that the experimental results give fairly good agreement with the theoretical results. This implies that the experimental results are accurate enough for estimation of azimuth of arrival waves, even if the antennas are in a water-filled borehole.

We have designed and developed a borehole radar, which has a directional radiation pattern and fits in a single borehole. A 3D image of the subsurface is obtained by applying a linear inversion scheme on the data, in which we deconvolve for the computed radiation pattern.

Unexpectedly weak signal levels, in both radar and crosshole modes, were obtained during a recent VHF borehole radar experiment in an arid, alluvial An investigation revealed that the kaolinitic strata, which hosted the vertical boreholes, were moistened by drilling fluid during site preparation. It was therefore conjectured that the weak received signals were due to attenuation caused by wet, conductive annuli surrounding the boreholes in an otherwise dry medium. To examine this premise, the VHF electrical properties (permittivity and conductivity) of representative dry and wet samples from the different strata were measured. These data were then used in a 3D, electrodynamic Finite Difference Time Domain (FDTD) code to model different annular configurations of dry and wet strata in and around boreholes of different dimensions. In the model, a 1 .2 m center-fed dipole was placed coaxially in the various boreholes and excited by a Gaussian pulse. This allowed quantification of the reduction in electric field strength due to the wet drilling technique. The results correspond well with the measured attenuation observed during the borehole radar surveys. The paper concludes with recommendations for the preparation of borehole radar surveys of paleochannels covered by kaolinitic strata.

Borehole radar has shown its value in delineating geological features, especially orebody geometry, in deep level gold and platinum mines in South Africa. At the same time, the cost of the critical components of a radar system has decreased dramatically while their performance has increased. The time is ripe for the introduction of a 100 MHz bandwidth instantaneous sampling slimline borehole radar to fit into 38 mm boreholes. A new generation of low cost, low current, fast analogue to digital converters (ADCS) has made it possible to implement instantaneous sampling in the probe. Memory speed is still limited, with speeds beyond 100 MHz not readily available. The borehole radar design uses four 8 bit 100 MS/s analogue to digital converters (ADC5) in parallel, each feeding its own First-In First-Out (FIFO) memory. The four ADC5 are each clocked 2.5 ns apart, to achieve a sampling rate of 400 MS/s. Once the trace has been acquired, it is downloaded from the FIFOs to the onboard microprocessor, where it is stacked. The resultant trace is transmitted out of the borehole using a 1 1 5 kbaud serial link over an optical fibre cable and displayed in real time on the surface control unit Data is processed using conventional GPR processing techniques. Once the data has been filtered, it is transferred to a 3D visualization environment, where reflectors can be compared to inferred target horizons. More complex environments can only be interpreted by using data from multiple boreholes, or by using directional antennas.

Broadband VHF borehole radars can be used by miners as tactical tools to map orebodies, faults and other marker horizons; to identify hazards well in advance of mining and to stop unnecessary mine development or underdevelopment. In principle, fan scans and crosshole profiles of borehole pairs and triplets can be used to synthesize three-dimensional images interferometrically if their trajectories are accurately known, signal-to-noise ratios are adequate, the spatial sampling rate is sufficient and if the target space is sufficiently uncluttered to limit the formation of mirages during reconstruction. However automatic methods of projecting data into 3D image space make stringent demands upon rock homogeneity, translucence and the accuracy of borehole trajectories. These demands can be relaxed by kinematic mapping, using geologically plausible 3D primitives such as cylinders, planes and hollows. In this paper we show that a number of useful primitives derive from a common element — a hoop (with semi-circular shell cross-section) of finite radius, centered on the borehole. The parameters defining the hoop's plane, radius and thickness can be recovered sequentially from borehole radar data. Information about the location and shape of simply curved 3D reflectors can be recovered by finding common tangents to groups of kinematically derived rings.

Concrete inspection is a well-known use of GPR. The advent of modem GPR instruments combined with vastly increasing computing power and rapidly improving software permits more effective use of GPR. The impediment to wider use is now the ease-of-use of GPR technology for the average commercial user. The requirements for concrete inspection are many and varied. The most common is to clear areas prior to cutting and coring for the installation of utilities or renovations. The requirement is a quick means of knowing how to avoid critical elements such as posttension cables or embedded utilities. Structural applications address the integrity of the concrete itself such as the presence of voids/air pockets, chemical alteration, or cracking. Owing to the less well defined and site specific character of such features, GPR applicability is unpredictable and interpretation of results still depends on gaining experience with the specific site conditions. Intrusive investigation must augment GPR findings to draw definitive conclusions. The embedded object class of problems is readily tractable with GPR. In the last year we focused on development of a system which enables non-GPR specialists to image concrete. The system required integration of a GPR sensor with a positioning procedure and on site data processing in a single package. The output is depth slice maps that allow the operator to immediately select cutting and coring locations. Systematic ergonomic procedures to make operation easy for average concrete users represented a major challenge. In this paper we discuss the design considerations and present the final system embodiment. We use a number of data examples from real environments to illustrate the development.

This paper focuses on the accuracy of detection of the first defect from a ground penetrating radar (GPR) signal in a masonry wall. The main conclusions are drawn from a carefully executed piece of experimental research work based upon field work on a masonry wall on the Bell Tower at Cremona. From inspection of the field GPR records, the resolution of detection of the first target or defect was found to be related to the length of the first reflection from the surface of the masonry. Thus conventional geophysics guidelines with respect to target detection related to one-tenth of a wavelength were tested against field observations and found to be inapplicable in relation to the detectability of the first defect. The shallowest detectable target proved to be at a depth of one-third the centre frequency of the antenna.

An investigation into the detection of voids in posttensioned concrete beams by ground penetrating radar is presented. Both experimental and numerical modelling results suggest that the optimum orientation of the radar's antennas is perpendicular to the long axis of the ducts containing the post-tensioning tendons.

The application of impulse radar for structural investigation of concrete elements has been increased during the last years related to technical developments, i. e. of high frequency antennas. In this paper, case studies will be presented related to the location of reinforcing bars, tendon ducts and repaired concrete areas in concrete bridges, anchors and dowels in concrete highways and delaminations of layered structures as they are used for non-ballasted railway tracks. It is shown that impulse radar can be applied in case of regular inspection and for searching the cause of damages but also for quality assessment in civil engineering.

The inspection ofbridge decks is today a routine application of the GPR method. Pavement thickness, depth of re-bar, concrete damage and the position of tendons are some of the issues that are addressed with GPR. However, in most cases a verification of radar results is either not possible, restricted to laboratory specimens or to a few points on real objects. A detailed comparison between radar results and reality for bridge decks is not only of scientific interest but could also improve the acceptance of GPR in this field of work. In a research project, bridges designated for demolition are inspected with EMPA's mobile GPR unit. Radar results are reported before the bridge is taken down. After demolition radar results are verified through a comparison with actual bridge parts. Work on the first in a series of 6 bridges has been completed recently. The radar inspection was carried out along 2 lines over the whole length of the bridge. During demolition, the bridge was sawed along the lines where the radar inspection had been carried out and the parts were stored in a landfill. This enabled a one-to-one comparison between radar results and reality.

After the complete demolition of a prestressed concrete bridge in the south of France, a suspected weak posttensioned beam has been put aside for non-destructive testing (NDT). Ground Penetrating Radar (GPR), Ferroscan® (covermeter), gamma-ray radiography and impact-echo (IE) have been tested and their results are discussed after the autopsy of the beam by hydrodemolition. The paper describes how NDT surveys can answer structural engineers need, using complementary NDT approaches. The use of a second technique should provide a better level of answer, while eliminating certain ambiguities either by improving measurement reliability, or by focusing onto doubtful zones for more precise local measurements. Finally, problems not answered by these techniques, described herein are pin pointed.

One of the most crucial problem in roads rehabilitation regards the pavement damage. Usually it is easy to localize the damage, but it is always difficult to identify the causes. The rehabilitation can be compromised, if the cause is not removed. The GPR technique is used by many Agencies involved in roads management. It is nondestructive and it is promising for soil characteristics interpretation, such as moisture or density. A classification of pavement damage, based on GPR analysis, is already performed and experimentally validated. The causes of damage are often referable to water intrusion in subgrade or clay intrusion in sandy subgrade. This is why we principally investigate how the moisture and the soil density influence the dielectric constant. The outcomes of a laboratory experimental survey are here discussed. Different soils have been considered. The GPR responses of each soil have been investigated, considering different moistures. The dielectric properties are highly correlated with the water content and the water status in soil. A significant correlation between dielectric properties and soil density is shown; this correlation is not so relevant as the previous. Any generalization is not reliable, but the study proves that the GPR evaluation of subgrade characteristics is possible.

A series of experiments on asphalt pavement specimens with 1-GHz GPR system was conducted in the Connecticut Advanced Pavement (CAP) Laboratory. The objectives of these experiments are as follows. (1) Determining the dielectric property of the asphalt specimens; and (2) Correlating electromagnetic (EM) properties with void ratio and asphalt binder content ratio of asphalt pavement specimens. Results of this study can be used as the baseline to calibrate GPR field surveys for pavement assessment, as well as an indirect means to monitor the compaction process. The travel time for direct and reflected phases are the fundamental information for computing EM wave velocity. Supplemented by pre-existing information on geometry, void ratio, composition measured by other means, we were able to compute the dielectric constant of 30 pavement specimens. The major conclusions from these experiments are as follows. (1) In general EM wave velocity is highest in dry conditions, intermediate in frozen, and lowest in watersaturated conditions; Correspondingly, the dielectric constant is smallest in dry conditions, intermediate in frozen, and highest in water-saturated conditions. (2) EM wave velocity increases slightly with the increase of void ratio for dry samples. In contrast, it decreases significantly with a void ratio increase in water-saturated conditions. Correspondingly, the dielectric constant decreases noticeably with an increasing void ratio in dry conditions, and increases appreciably in saturated conditions. (3) The changes of EM velocity and dielectric constant for dry and saturated conditions can be predicted by the effective medium theory for porous media. (4) When the dielectric constant of the pore material is taken as the value of fresh water ice, the change of EM velocity and dielectric constant for the asphalt pavement specimens in frozen conditions cannot be correctly predicted with the effective medium theory. This implies that the pore water was not completely frozen when GPR measurement was taken. There are no significant changes in dielectric constant for dry and frozen conditions. (5) Variations in EM velocity and dielectric constant with asphalt binder ratio imply that a low asphalt ratio corresponds to a high void ratio so that in the lower end of the asphalt ratio, EM velocity has maximum fluctuation among the different conditions.

roadbeds using multi-channel GPR is described. GPR is a continuous and non-destructive method and its capabilities of providing information on soil and water content is well documented in the past. Most of these earlier described methods involve several time consuming measurements with a variety of antenna settings. A multi-channel radar system can however make more efficient measurements through the use of independently controlled transmitters and receivers in the antenna array. The results, from measurements with a 500 MHz GPR system, show a clear correlation between the GPR data and the soil water content in a Swedish roadbed. Further evaluation is, however, needed to compare different antenna frequencies and to calibrate the equipment together with reference velocity analysis.

The construction of motorways generally involves the removal of the loose clayey surficial deposits thus creating adequate conditions for the illumination of structures several meters deeep within resistive bedrocks. The present paper describes the acquisition, processing interpretation and drilling results associated with the construction of a motorway intersecting karstic limestones in southwestern France. The processing sequence was similar to to sequences routinely applied to reflection seismic data from gain recovery to migration. In addition a Georadar Index section was derived from the envelope of the analytic signal after migration. Drilling on a vertical anomaly about 1.5 m high at a depth of 1,5 m confirmed the presence of a void between 1.5 m and 3.0 m. Many other structural details such as joints enlarged by karstification, layering, faults, can be seen on the sections.

Although GPR is normally capable of detecting the responsefrom buried plant, accurate detection and mapping of extended geometrical features in 3-dimensional data is often a major problem faced by the radar operators and geophysicists. This paper presents a pattern recognition approach based on the 3-dimensional Hough Transform for the detection of extended linear targets. By transforming spatially extended patterns into spatially compact features in parameter space, a difficult global detection problem in data space becomes a more easily solved local peak detectionproblem in parameter space. This technique allows the combination of qualitative site information and ground truth in order to increase the accuracy of the final result. Improved freedom of movement and accuracy is achieved by logging the movement of the GPR unit using DGPS. The user is presented with a 3-dimensional site survey report detailing the length, depth and orientations (azimuth and zenith) of any pipes, cables or the like.

Manually picking the first anival of energy in a series of cross borehole GPR ray traces can be time consuming and subjective, especially when large data sets need to be processed. One possible remedy is the application of a back propagating neural network. Neural network applications have been used previously in seismic studies to pick the arrival of the P and S waves (Dai and MacBeth, 1997; McCormack et al. 1993; Murat et al. 1992). One particular method, which applied a moving window over the trace, is used here with slight modification. Noisy time-amplitude records were first normalized to range from —1 and 1 . These data were then filtered such that values between —1 and a negative threshold were set to —1 , values between 1 and a positive threshold were set to 1 and all other values were set to zero. The filtered wave was fed through a neural network that searched for a pattern related to a first arrival. Several filtering parameters were tested, including the size of the moving window, the values of the positive and negative thresholds, and neural network parameters pertaining to training and testing. With minimal training, the neural network performed very well compared to hand picking of arrival times on large data sets.

Continuing improvements in computer technology have made 3-D imaging a standard GPR interpretation technique. The most common data collection methodology for 3-D imaging involves collection of data along parallel profile lines. The data are then often migrated and concatenated into a 3-D file. A 3-D image generated from the file is manipulated to detect linear and finite-size targets. The detection of linear and finite-size targets can be enhanced by creating images generated from data collected along orthogonal profile lines. The fact that the minimum angle formed between the long axis of a linear target and one of the orthogonal profile lines is 45 degrees enhances the detection of a linear target because in at least one profile line direction the reflection from the linear target will form the familiar hyperbola and a series of hyperbolas concatenated from parallel profile lines are readily observed in the 3-D image. Perhaps the most beneficial aspect of using bi-directional data is the ability to perform spatial filtering operators to improve detection of linear targets. Background removal filters applied to parallel profile line data will generally erase reflections from pipes or rebar that trend parallel to the direction of the profile lines. Comparisons of the data visualization capabilities between one-direction and orthogonal profile line data collected on reinforced concrete and on a buried pipe test site clearly show the advantages of imaging using orthogonal profile line data on both small and large scales.

Accurate and consistent manual interpretation of the vast quantities of GPR data collected during a typical survey constitute an implementation bottleneck that often limits the practicality and cost-effectiveness of this tool for rapid site investigation. Automatic unsupervised interpretation of GPR data is achieved by training a neural network to discriminate between signals originating from different types of targets and other spurious sources of reflections such as clutter. This is achieved by computing a number of statistical data descriptors for feature extraction. The neural classifier is capable of returning 3-dimensional image outlining regions of extended targets (such as reinforced concrete, disturbed soil or storage tanks) and pinpointing the location of localised targets such as mines and pipes. These reports are accompanied by a written log detailing the depths and geometry of these targets. This classifier was applied to a variety of GPR data sets gathered from a number of sites. The obtained results were in close agreement with those obtained by a trained operator manually, but in a fraction of the time. Different targets have been successfully discriminated, with a consistency greater than that of the operator. Although the system is implemented in software, the rate at which classifications are rendered lends the system Authors would like to thank the Engineering and Physical Sciences Research Council (EPSRC) for funding this work as a part of a larger project regarding automatic data-processing of ground penetrating radar. Authors would like also to express their gratitude to Zetica (UK) Ltd. for supporting this work financially, and providing sites data and related software. favourably to near real-time on-site processing and interpretation.

Ground penetrating radar (GPR) has been widely used for the detection and location of buried objects. However, the detection method is often subjected to operator's interpretation due to large quantities of data and undesired clutter and noise. Such a detection method is neither reliable nor efficient.

The U.S. Geological Survey is developing a directional borehole radar (DBOR) tool for mapping fractures, lithologic changes, and underground utility and void detection. An important part of the development of the DBOR tool is data analysis and visualization, with the aim of making the software graphical user interface (GUI) intuitive and easy to use. The DBOR software system consists of a suite of signal and image processing routines written in Research Systems' Interactive Data Language (IDL). The software also serves as a front-end to many widely accepted Colorado School of Mines Center for Wave Phenomena (CWP) Seismic UNIX (SU) algorithms (Cohen and Stockwell, 2001). Although the SU collection runs natively in a UNIX environment, our system seamlessly emulates a UNIX session within a widely used PC operating system (MicroSoft Windows) using GNU tools (Noer, 1998). Examples are presented of laboratory data acquired with the prototype tool from two different experimental settings. The first experiment imaged plastic pipes in a macro-scale sand tank. The second experiment monitored the progress of an invasion front resulting from oil injection. Finally, challenges to further development and planned future work are discussed.

For the imaging of ground-penetrating radar (GPR) data, the form of the radar wavelet is usually not taken into account, because the behaviour of actual source and receiver antennas, that have a large influence on the wavelet, is rather complex. Nevertheless, knowledge of the wavelet has the potential to improve the imaging and interpretation of GPR data. An efficient way to deal with the wavelet is proposed by introducing an effective wavelet that incorporates the influence of the finite-length antennas. To obtain this effective wavelet, the impulse response for a point source-receiver antenna system is calculated using the medium properties, that are obtained from the isolated air- and ground-waves observed on the actual CMP data. The deconvolution of this impulse response with the actual CMP data, yields an effective wavelet. Together with the well-known radiation characteristics of dipole antennas in a dielectric halfspace, the propagation of the electromagnetic waves emitted a finite-length source-receiver antenna system can be effectively described. We demonstrate that propertiesof the shallow subsurface can be extracted from the ground-wave with reasonable accuracy. An effective wavelet determined from numerical data calculated for a finite-length source-receiver antenna system shows an effective wavelet that is less minimum phase than the effective wavelet calculated from the electric field generated by a point source-receiver antenna system.

Automatic detection and characterization of the signatures of solid reflecting targets in ground-penetrating radar data is achieved by a combination of signal and image processing stages. For the class of target under consideration, namely localized or extended linear reflecting targets such as landmines, pipes or cables, the reflections exhibit a broad hyperbolic anomaly in the region of the target. Detection and characterization of these distinctive signatures yields information about the location of the targets as well as the surrounding medium. Edge enhancement and edge processing techniques are developed to trace the envelope of the reflected wavefronts. By fitting hyperbolae to these detected edges, the location of the targets and the relative permittivity of the medium are estimated. This estimate enables the effective elimination of the background clutter that leads to spurious non-hyperbolic reflections. Thus automatic target detection and mapping is achieved without the heavy computational demands of techniques such as synthetic aperture radar processing, enabling on-site data interpretation.

The volume of image data generated in ground-penetrating radar surveys can severely restrict the practicality of this site investigation technique. This is particularly true in situations where automatic analysis or interpretation is required, as segmentation and classification tasks that utilise multivariate data are critically affected by the volume and dimensionality of the data. A general-purpose unsupervised image segmentation system is presented here for the automatic detection of image regions exhibiting different visual texture properties. A suboptimal feature selection procedure is proposed to automatically select the set of texture features best suited for the particular application. The reduction in the size of the feature set both reduces the computation time and improves the accuracy of the final classification.

Improvement of resolution is the challenging issue in Ground Penetrating Radar (GPR) and that is greatly desired to increase in order to get the clear imaging of very closely buried targets. GPR has been approved as very successful technology for various kinds of investigations & detection of buried targets. In this paper, the application possibility of super resolution technique MUSIC (Multiple Signal Classification) algorithm is examined because of its superior results. Moreover, the conventional FFT (Fast Fourier Transform) has been utilized to get higher precision receiving signal level. Combined Processing Method (CPM) oftime domain response of MUSIC and IFFT (Inverse FFT) has been proposed for the first time to get high resolution and high precision receiving signal level. Simulation and experiment result show that the proposed method has high resolution and high precision receiving signal level than other conventional signal processing approach.

Subsurface imaging techniques have been discussed with MUSIC algorithm to locate point reflectors. Generally, a radar signal received by a spatial array must be decorrelate as pre-processing for the MUSIC algorithm. Since most of decorrelation techniques are based on a plane wave incidence model, it is difficult to apply the decorrelation technique for the radar targets around the near field of the array. In this report, we introduce a new decorrelation technique with transmitting and receiving array for targets near the array and apply the MUSIC algorithm for estimation of the 2D position of the point targets in crosshole tomographic measurement. Simulation results show that this method has much higher resolution and accuracy than the conventional diffraction stack method. Moreover, we discuss the relationship between the S/N ratio and the geometrical separation.

Large quantities of non-aqueous phase liquids (NAPL), contaminate the near surface sediments at Operable Unit 1 (OU1), Hill Air Force Base (HAFB), Utah. In October 2000, a 3D, multi-offset GPR survey was acquired at OU1 with two objectives: 1) to image the aquifer/aquitard boundary at a depth of about 30 ft, and 2) to evaluate quantitative processing and interpretation methodologies for direct detection of NAPL. Using pre-stack depth migration, we map the aquitard boundary to about ± 1 ft throughout the survey area. An unusual reflection is identified within the vadose zone that does not correlate with known geology. The region below this reflection has anomalously high velocity, implying low electric permittivity, and the amplitude of the anomalous reflection deviates significantly from the background AVO trend. Fitting the Fresnel equation to the AVO data, we estimate the velocity contrast at the anomaly boundary and find that it is in good agreement with the migration velocity model. We interpret the anomaly as a previously unidentified NAPL rich zone. Subsequent coring and chemical analyses verify our interpretation. This exciting result implies that these methodologies may be useful for direct detection of NAPL at other HAFB locations and at sites with similar hydrogeology.

During 2000, Hager GeoScience, Inc. (HGI) performed a geophysical survey at a former paint factory site outside Boston, Massachusetts at which lead- and chromiumcontaining paint residue was known to be present in the subsurface. The goal was to determine the applicability of electromagnetic methods to mapping variations in the concentrations of pigment in fill. GPR and EM terrain conductivity techniques were tested in areas of known contamination and then applied to a nearby area under investigation. A GPR survey was performed along traverses at 2-foot spacing, enabling the use of 3D modeling techniques. Survey results were presented as GPR 3D and filled color contour EM plots. Slices at increasing depths through the GPR 3D plots revealed spatially localized high-amplitude zones of pigment concentration and attenuated zones of normal fill. The shallow depth slices showed that the pigment was most concentrated between 10 and 15 feet. Slices through the 3D plots at greater depths showed that in some areas the attenuated zones of normal fill reached depths of over 30 feet, consistent with fill depths in boring logs provided by the client.

By its very nature, ground penetrating radar (GPR) is an ultra-wideband device (UWB), requiring a large range in frequency to penetrate the ground and image with sufficient resolution to solve practical problems. The increasing scarcity of electromagnetic spectrum and the proposed use of other UWB devices in 1998 caused the U.S. Federal Communications Commission (FCC) to initiate an inquiry (NOl ET Docket 98-1 53) to investigate permitting the operation of ultra-wideband devices (including ground penetrating radar) on an unlicensed basis under Part 1 5 of he FCC rules. Through 14 February 2002, the FCC had received over 9 1 0 comments in the inquiry (NOl 98-153) and on the Notice of Proposed Rulemaking (NPRM 00-163) issued in June, 2000. The First Report and Order (R&O 02-8) was issued on 14 February 2002. Before this, ground penetrating radar use was only officially permitted for those who had received waivers from the FCC or NTIA (National Telecommunications and Information Administration), who jointly regulate radio spectrum use in the United States. Legal waivers to manufacture ground penetrating radar were issued to U.S. Radar Inc., Time Domain Inc. and Zircon Corp. in June, 1999, and expiring with the R&O. Waivers to build or use ground penetrating radar were issued to the U.S. Geological Survey, U. S. Army and U. S. DOE by NTIA (or predecessors) from about 1976. The R&O issued on 14 February 2002 ". . provides for the operation of GPRs and other imaging devices under Part 15 of the Commission's rules subject to certain frequency and power limitations. The operators of imaging devices must be eligible for licensing under Part 90 of our rules. . ." "At the request of NTIA, the FCC will notify or coordinate with NTIA prior to the operation of all imaging systems." "GPRs must be operated below 960 MHz or in the frequency band 3.1-10.6 GHz." The FCC ". . intends within the next six to twelve months to review the standards for UWB devices and issue a further notice of proposed rule making to explore more flexible standards and address the operation of additional types of UWB operations and technology." This FCC rule impacts ground penetrating radar manufacture, sale and use in the United States (and in other countries whose rules are linked to FCC regulations).

This study tested the applicability of utilizing groundpenetrating radar (GPR) in Florida to detect buried bodies. Twenty-four burials were constructed with pig cadavers and divided equally into two groups of average weights (24.49 and 63.5 kg) and buried at one of two depths (50-60 or 100-1 10 cm). Two soils were also utilized in this study to represent two of the most common soil orders in Florida: Entisols and Ultisols. Graves were monitored on a monthly basis for time periods up to 21 months with two GPR systems. During this study, grave anomalies became less distinctive over time due to decomposition of the body and settling of the disturbed soil (backfill) as it compacted. Soil type was a major factor. Grave anomalies became more difficult to recognize over time for deep targets that were within clay. Forensic targets that were in sandy soil were recognized for the duration of this study. Pig size was not a factor. The anomaly that was produced from a child size pig cadaver had the same general characteristics and was detected for the same duration of time as a larger pig cadaver.

Locating clandestine burials of human remains has longchallenged law-enforcement officials investigating criminal activity, and continues to confront scientific disciplines in finding well-defined procedures. Forensic specialists and law enforcement agencies have noted that multidisciplinary search efforts are becoming more of a necessity in searching for buried remains. Collaborative research at The University of Tennessee's Anthropological Research Facility (ARF) in Knoxville supports this concept. We are correlating groundpenetrating radar (GPR) imaging with postmortem processes. Decompositional stages and rate imagery are presented that utilize sweep-frequency radar and timeelapsed imaging. Greater accuracy in predicting clandestine burials using dynamic GPR anomaly detection will reduce widespread excavations and may better assist lawenforcement personnel in obtaining site-specific search warrants.

Ground-penetrating radar (GPR) technology has supplied invaluable assistance in numerous criminal investigations. However, field personnel desire further development such that the technology is rapidly deployable, and it provides both a simple user interface and sophisticated target identification. To assist in the development of target identification algorithms, our efforts involve gathering background GPR data for the various site conditions and circumstances that often typify clandestine burials. For this study, forensic anthropologists established burial plots at The University of Tennessee Anthropological Research Facility (ARF). These plots contain donated human cadavers lying in various configurations and depths. Each plot includes a fleshed cadaver with varying combinations of human skeletal remains, construction material, and backfill. We scanned the plots using two GPR systems. The first system is a multi-frequency synthetic-aperture unit (GPR-X) developed by the Department of Energy's Special Technologies Laboratory (STL), Bechtel Nevada (Koppenjan et al., 2000). The impulse radar system is a newly released commercial unit (SIR-20) manufactured by Geophysical Survey Systems, Inc. (GSSI). This paper provides example scans from each system, and a discussion of the survey protocol and general performance.

Ground-penetratinragd ar( GPR)a ndd ownholes oil moisturer eadinghs aveb eenc onductedo ver af ormer landfill to the north ofAdelaide, South Australia, in order to determine and map the thickness and lateral extents ofthe landfill capping layer over the underlying domestic waste. The study site, Cavan Landfill, is a former landfill undergoing a remediation program including landfill gas extraction and capping. Through the integration ofthe two data sets, GPR and soil-moisture, accurate information on the thickness and continuity of the capping layer on the landfill has been obtained. Six GPR profiles were collected using a Noggin 500 MHz unit. A time window of 50 nanoseconds was set to give maximum resolution between 0.8 metres and 2.4 metres. Ten shallow wells were installed for characterisation by the 'Diviner 2000' capacitance soil moisture probe along the periphery of the area. Moisture readings were collected in mmH2O over time at 10 centimetrei ntervalsu p to 1.6m etres in depth. Moisturep rofiles were superimposedo n the GPR profiles for interpretation. The results give accurate information on the capping layer, the degree of heterogeneity and moisture level fluctuations.

The performance of ground-penetrating radar (GPR) is dependent upon the electrical conductivity of soils. Soils having high electrical conductivity rapidly attenuate radar energy, restrict penetration depths, and severely limit the effectiveness of GPR. Factors influencing the electrical conductivity of soils include the amount and type of salts in solution and the clay content. A thematic map of the conterminous United States showing the relative suitability of soils for GPR applications has been developed. This map is based on field experience and soil attribute data contained in the State Soil Geographic (STATSGO) database developed by the Natural Resources Conservation Service. Attribute data used to determine the suitability of soils include clay content, electrical conductivity, sodium absorption ratio, and calcium carbonate content. Attribute data were generalized from soil surveys. Each sOil attribute was rated and assigned a value ranging from 1 to 10. The most limiting (maximum) indices within depths of 1 .0 or 1 .25 m were selected for electrical conductivity, sodium adsorption ratio, calcium carbonate content, and total clay for each soil component. Indices were summed for each soil component and for each map unit. The dominant value was selected as the GPR suitability index for each map unit. The GPR indices are displayed in a graduated color map. This map can be used to assess the relative appropriateness of GPR for soil investigations within comparatively large areas of the conterminous United States.

Active oolitic sand bars like the modern ones in the Bahamas as well as those in the ancient exhibit a complex internal architecture with a multitude of stacked sedimentary structures. As a result, the internal anatomy of these sand bars is usually too complex to be captured with twodimensional outcrop and one-dimensional well information. In order to improve fluid flow modeling used in water and hydrocarbon resource management, closely spaced threedimensional (3-D) data is needed to accurately map sedimentary structures. To improve our understanding of ancient oolitic sand bar anatomy we collected a 3-D 100 MHz Ground-Penetrating Radar (GPR) data cube covering an area of 24x46 m with 7m penetration depth. A grid spacing of 0.1x0.2 meters is needed to fully exploit the imaging capability of the 100 MHz antennae. The high-resolution 3-D GPR data enable: . Volume maging of oolitic sand bar architecture. . Understanding of spatial relationships of sedimentary structures. S Reconstruction of depositional environment and assessment of paleocurrent and sandwave migration directions.

To ease geological investigations with GPR (Ground Penetrating Radar) in rough terrain, the GPR equipment can be combined with a carrier-phase DGPS (Differential GPS). As traditional communication with GPS (with e.g., the NMEA protocol) involves an inherent time delay, a new method has been constructed to synchronize the measuring with GPS and GPR units in which the GPS is transmitting a trig signal to the GPR. The results of this case study clearly shows that combining GPR- and DGPS-measurements gives an investigation without need of initial land surveying or clearing of obstacles and vegetation. The investigation also gives information of the topography, which allows for a volume estimation of the investigated target, in this case a limestone layer.

Turbidite outcrops of the on shore portion of Almada Basin are made of conglomeratic sandstones and shales of the Urucutuca Formation. Those rocks are the unique Turbidite examples of the marine transgression mega-sequence in the Brazilian passive margin, which were sedimented during late Cretaceous. They are analogues to important Campos Basin reservoirs, offshore of Rio de Janeiro State, which is the largest Brazilian oil producer basin. A combined GPR, resistivity and well log survey was planned in the region to provide information regarding rock properties, such as lithology, geometry and permeability. The survey's goal was to predict the variability on the characteristics normally encountered in this kind of deposits. The GPR survey was done in order to provide a stratigraphic interpretation.

We study the capabilities of low frequency radar systems to sound the subsurface in arid countries. This approach is based on the coupling between two complementary radar techniques: the airborne Synthetic Aperture Radar (SAR) used in L-band (1.2 GHz) for imaging large scale subsurface structures, and the Ground-Penetrating Radar (GPR) used between 500 and 900 MHz for sounding soils at a local scale, from the surface down to several meters. In this paper,. we first recall the results obtained on the Pyla dune (France). This site is a bare sandy area presenting large subsurface structures (paleosoils) at varying depths. A polarimetric analysis of airborne SAR data, as well as the GPR sounding experiment, shows that subsurface scattering occurs at several places. The SAR penetration depth is estimated by inverting a simple scattering model for which the subsurface structure, i.e. geometric and dielectric properties, is determined by the GPR data analysis. The recent results obtained on the well-known site of Bir Safsaf (southern Egypt) are then presented. The comparison between L-band SAR and GPR sections shows that penetration effects occur in many places, revealing rich subsurface structures. These results suggest that airborne radar systems in a lower frequency range (P-L band) should be able to detect soil structures down to several meters, leading to innovative Earth observation systems for geological and hydrogeological mapping in arid regions.

Urban areas built on old lacustrine basins in central Mexico show subsidence related to creep-fault processes. These processes are highly dynamic and enhanced by aquifer water extraction and El Ni?o events. The affected cities are Morelia, Salamanca, Silao, Celaya, Aguascalientes, and Querétaro, with a global population of around 2.5 million, and considerable industrial assets. The first three cities show NE-SW creep-fault trends, while in the remaining the trend is N-S. The subsidence rate varies: 2-3 cm/year at Salamanca, 4-6 cm/year at Morelia and Querdtaro, and 6-8 cm/year at Celaya. In order to preliminarily evaluate the effects of the creep-faults, we performed a series of nondestructive ground penetrating radar surveys in the city of Morelia, that complement concurrent geological studies of the area. We report herein results in three locations where the surface expression of the faults is well defined, including substantial damage to homes and buildings. One of the objectives was to determine the length of the disturbance, perpendicular to the faults' trends. Using 50 Mhz antennas we reached an exploration depth of around 10 m in sandstone-bearing lacustrine deposits and lava flows. In most cases the faults show low-angle dips that vary with depth in the main fault. Additional, synthetic and antithetic faults are developed in the disturbed area of the main fault, which we define as the influence zone of the main disturbance. We find that this zone varies from 15 to 50 m on both sides of the fault, although it tends to be larger on the down thrown block. Faulting appears to be controlled by geological formations within the first 10-15 m in depth.

Good drainage is important for healthy turf and proper playing surface. Because ofthe aesthetic and playability considerations, most ofthe drainage systems in a golf course were installed underground. With time golf green drainage systems can fail or become plugged up due to improper construction and/or management. Unfortunately, many golf green drainage maps are either unavailable or inaccurate. Locating a drainage system in a green is a very time consuming and difficultjob. Many golf course superintendents invested many hours in locating these pipes when drainage problems occured. Correcting the drainage problems can be destructive to the green and expensive when location of the present system is unknown. The objective of the study was to locate and map the tile drainage system under a green using a ground penetrating radar. In the study, a SIR system 2000, with a 400 MHz antenna, ground penetrating radar (GPR) was used to scan a USGA green and a California style green. The experiment was conducted at the Stone Creek golf course(SCGC) Green No. 3 (a USGA green located at Makanda, IL) and Hickory Ridge golf course (BRGC) Green No. 2 (a California green, located in Carbondale, IL). The green at the SCGC was about 500 m2 and the one at HRGC was close to 200 m2. Since sprinkler heads are fixed objects around the green, they were used as permanent reference points. The golf greens were divided to form a grid system and marked with flags 1 meter apart. The green was marked and scanned. In the measurement, the scanning was completed within fifteen minutes, but it took up to 45 minutes for laying out the grid of a 500 m2 green. Results indicated that GPR could accurately locate the rooting zone thickness, depth of gravel layer, and drainage tiles in a golf green with minimum time and disturbance.

Information extracted from crosshole georadar data has been used to characterize a gravel- and sand-dominated aquifer. Inversions of direct arrival traveltimes and amplitudes have provided electromagnetic velocity and attenuation tomograms that have allowed critical hydrological structures and parameters to be determined. An integrated interpretation of the velocity and attenuation tomograms was performed via a k-means cluster analysis. As a result of this multivariate statistical analysis, major trends in the relationship between velocity and attenuation were identified, thus enabling us to outline the major hydrostratigraphicu nits ofthe surveyedd eposit.

Radar reflections for a layered medium are dependant on the dielectric constants ofthe layers, which is closely linked to saturated porosity, and more loosely to hydraulic conductivity. Radar data have been obtained at a site where hydraulic conductivity has been measured in great detail. The radar cross section from the site clearly shows layering within the section, and it is tantalizing to predict that the hydraulic conductivities also persist along the bedding surfaces. The radar trace may be converted to a band limited pseudo-dielectric constant log by the same methods used to estimate an acoustic velocity log in seismic work. Thus, the resulting dielectric constant section can be converted to pseudo-porosity and pseudohydraulic conductivity displays. But, because of the limited bandwidth of the radar signal, it is tricky to invert the radar traces to yield dielectric constant and ultimately hydraulic conductivity. The main computations are 1 . deconvolutionw ith Seismic Unix routines and 2. conversion to dielectric constant including filtering to minimize numerical instabilities.

Borehole ground penetrating radar (GPR) was used to measure the temporal and spatial variability of soil water content under uniform wetting and drying conditions. Zero Offset Gather (ZOG) surveys were conducted before and during the uniform infiltration and the subsequent drainage experiments using the PuLseEKKO 100 borehole system with 200 MHz antennas in horizontal access tubes. Time domain reflectometry (TDR) data were collected from 10 vertical probes installed at 0. 1 m increments from 0.1 to 1.0 m below the ground surface. The TDR data were used as standard measures of soil water content to compare with the GPR estimated water content. The electromagnetic wave velocity along the survey profile at about 1 .0 m below the ground surface was estimated using ZOG data by picking the arrival time of the first event. Volumetric water content was calculated using a standard empirical relationship between velocity and water content for each ZOG location. Measured higher soil water content zones are potentially preferential flow areas and were observed in consistent locations throughout both the wetting and drying experiments. The radius of influence of the borehole GPR measurements was about 0.5 m determined theoretically and by comparing GPR and TDR data.

A controlled water production was performed at a water source of Ulaanbaatar city, Mongolia to evaluate the effectiveness of ground penetrating radar for detecting and monitoring dynamic groundwater movements in the subsurface. The groundwater level in Ulaanbaatar city area is between 2m-10m. This relatively shallow depth makes it a suitable target for detection by GPR. Field experiments in Ulaanbaatar were carried out in 2001 with 100 MHz antennae. We measured the groundwater level around a pumping well, then we monitored the change of the level after the pump was stopped. The groundwater level was about 5m. We could detect the groundwater change about 50cm, when the water level in the production well changed by 65cm. If we acquire the data by locating the antenna positions very accurately, we can obtain radar profiles with very high coherency, and we could see the groundwater migration clearly. The CMP was also used in the same site in order to determine the reflection from the groundwater surface. It was found that CMP and velocity analysis give a good information about the depth variation of the groundwater saturation in soil.

Two infiltration experiments were conducted to monitor an advancing wetting front at 2.25 m below ground surface with cross-borehole ground penetrating radar (XBGPR). The focus of the experiment was to understand how XBGPR responds to dielectric permittivities that vary on a scale that is smaller than the antenna length. To test this response, a sharp wetting front was formed by applying water evenly over a 5 m by 5 m area at a rate of 5x10-4 cm/s through porous hoses. The center of XBGPR antennae were placed at a depth of 2.25 m in a pair of vertical, PVC lined access tubes located within the irrigated area. The velocity of the first arrival was converted to moisture content using a standard calibration. The measured water content increased linearly with time during the advance of the wetting front. Through comparison with modeled results of flow in unsaturated media, we demonstrate how water contents are "averaged" along the antennae.

The soil water content distribution at two field sites was measured with the air launched surface reflectivity method using a standard GPR system elevated ~1 m above the surface. Time domain reflectometry (TDR) measurements of water content were also acquired at these sites. At one site, water was applied to the surface in two separate experiments and the water content was measured during drainage. At the other site, a water content profile was acquired across two terrain types, a flat grass field and a corn field. Although the GPR surface reflectivity method was able to map the water content distribution at both sites there were substantial differences between these measurements and those acquired with TDR. The main contributors to these differences are likely scattering and changes in the nature of the gradational air/ground interface, related to spatial variability in water content.

Densely sampled GPR data can supplement hydrogeological data for estimating the spatial distribution of porosity and hydraulic conductivity over an aquifer. However, most of the time, the GPR surveys are performed along profiles (2D) and, when 3D models are seeked, the third spatial dimension is assessed by acquiring data over parallel profiles. This configuration is possible when the ground surface is clean and free of any obstacle but these conditions are seldom met in the field. Here, we show the results of a pseudo-3D acquisition protocol when spatial sampling cannot be done on a regular grid. The GPR reflection times are correlated with piezometric and stratigraphic information; cokriging of both data after some mathematical manipulation yields to an estimate of the hydraulic conductivity.

Continuous monitoring by time-lapse and repetitive measurements using cross-hole geo-radar was conducted to investigate soil moisture distribution and migration beneath infiltration pit for artificial groundwater recharge. This monitoring enabled us to clarify the infiltration process from the infiltration pit into the vadose zone in a quantitative, nondestructive, and noninvasive way. The infiltration pit was 2.0 x 2.0 m square and 2.3 m deep between 2 boreholes in gravel soil. The groundwater table was at about —10 m. We monitored the veitical distribution of electromagnetic wave traveltime beneath the infiltration pit by repetitive measurements using cross-hole geo-radar profiling with zero-offset gathering. Traveltime was distinctly retarded from the upper layer to the deeper one after ponding of the pit. The downward retardation velocity of the infiltration rate into the soil was estimated at8 x 10-2cm/s. The estimated values for water content and water seepage rate in the soil were almost coincident with the directly measured values. In our case of an infiltration pit test, cross-hole geo-radar monitoring was an efficient, noninvasive method for visualizing the infiltration process and estimating water migration properties of the soil on a macro scale.

A research project is being conducted to identify methods of using ground penetrating radar (GPR) to improve railway track condition assessment and enhance track inspections and safety. The safety of passing traffic can be improved if better indicators of problematic track conditions can be developed and utilized to better inspect the track for safety and to guide maintenance. The research effort has included evaluation of data collection techniques including testing a variety of GPR systems, identification of data interpretation techniques, and comparison of GPR data to track condition information. One limitation that has been identified is a lack of information on the electrical properties of track materials. Although data from geologic materials is well documented, the specific characteristics of railway track materials are different. For example, granite is documented widely as an intact rock mass and is used as ballast for track. However, in the railway ballast application, granite is used as crushed stone. During this research, tests were conducted to measure the dielectric permittivity of a variety of track materials to verify and supplement field measurements and to provide reference data for data interpretation. This paper describes the research project and the results of the testing and analysis.

GPR applications to historical buildings have been cautiously increasing in recent years. The investigation of the wooden structural elements in historical buildings is essential to plan the restoration works. We discuss three case histories where 2D and 3D GPR surveys were executed to solve problems posed by restorers. The first survey was carried out to locate the beams of a wooden floor in a two hundred years old house in Pescate (Italy). The second survey was carried out in a stone masonry house of the 19th century in Lecco (Italy) to investigate the beam-wall connection. The third survey was carried out in a five hundred years old church in Busto Arsizio (Italy) to detect all the wooden elements. In this case some beams were totally hidden inside the brick and stone walls.

A three-dimensional (3-D) georadar survey has been conducted across a 41.2 x 34.5m area with moderate topographic relief (dips: 4 — 16°) near Randa in southwestem Switzerland. For this survey, we employed a semiautomated acquisition system that combined a standard georadar unit with a self-tracking theodolite. This system recorded georadar data and coordinates simultaneously. Subsequently, an accurate topographic model of the acquisition surface was determined from the measured coordinates. With the aid of this topographic model, a provisional static correction for each georadar trace was determined. Application of the static corrections removed the most significant distortions of the major reflections and diffractions, which likely originated from the soil-rock interface and/or major fractures.

The objective ofthe present GPR survey was to investigate a segment of underground railway in Paris where subsidences had been observed. The railway sleepers are made of concrete reinforced by 7 rebars. The sleepers are regularly spaced every 60±2cm. With 400 MHz antennas part of the transmiued energy penetrates the ground and is reflected back to the receiving antenna by the subsurface mirrors, unfortunately backscattering from the regularly sleepers makes it difficult to see reflections from the subsurface. Taking advantage of the uniform spacing between sleepers a wavenumber notch filter was designed and applied. The resulting image made reflections due to a decompressed volume clearly visible. Drilling confirmed the presence of a decompressed volume likely to be a precursor of an underground collapse progressing towards the surface.

This paper presents a non-destructive procedure to determine the surface hardness of the soil using an elevated radar system equipped with a focused-beam antenna. The complete system consists ofthe focused beam antenna, a network analyzer, a computer, and control software that process the data. The surface dielectric constant calculated from the radar data based on reflection measurement was compared with the surface hardness measured directly from a homemade hardness meter. A sand pit was used as test bed that was frozen by liquid nitrogen. It was found that there is a simple and direct relationship between the surface hardness and the measured dielectric constant in our case.

railroad track substructure condition on a continuous top-ofrail nondestructive basis. In this study, 1 GHz radar data were acquired between concrete and wood ties as well as from the ballast shoulders beyond the ends of the ties, and with multiple antenna orientations and polarizations. Automatic processing of the data was developed to quickly generate hard copy sections of radar images and for input into railroad track performance monitoring software such as ORIM. Substructure conditions were observed such as thickness of the ballast and sub ballast layers, variations in layer thickness along the track, pockets of water trapped in the ballast, and soft subgrade from high water content. In addition, locations and depths of subsurface drainage pipes, trenches, and utilities were quickly and continuously mapped. GPR data were acquired and processed from a hirail vehicle moving continuously at 10 miles per hour with radar resolution of a few inches horizontally and a fraction of an inch vertically to depths of more than six feet. The largest errors resulted from the positioning system used to locate the antennas along and across the track. Automatic modeling to determine density and water content is being developed but the uneven and rough (at radar wavelengths) air-ballast interface is a major problem in modeling the data.

Railway is nowadays considered as the most ecological and economical alternative to cars and planes; on the other hand, in order to be really competitive, trains have to go faster and faster on safe tracks that allows an high velocity. From this point of view, speedy track construction is just as important as periodic maintenance and methods of the past (such as coring and visual inspection) are not longer able to provide helpful information especially because too slow and, often, too expensive. Therefore, the use of a Ground Penetrating Radar can be thought as a suitable and economical alternative to the other survey methods. In this paper is presented a GPR equipment especially designed for inspecting and verifying railroad tracks in a reliable and effective way, without interruption of regular traffic.

Tunnel management system (TMS) is a software system that stores data on current tunnel conditions. The system designs repair strategy necessary to keep waterway tunnels in proper condition. The TMS database includes eye observation records, boring data, GPR-acquired information, etc. GPR has been widely used in Japan to disclose unseen objects behind the lining of waterway tunnels. Application has been limited, however, to comparatively larger diameter tunnels. A GPR rapid-survey system for small-diameter tunnels has been developed to fulfill the need for database completion of the TMS. Through application to many tunnels, GPR has been found effective in terms of obtaining knowledge objects behind the lining. Through theoretical speculation and experience, the interpreting effort has revealed many unseen events such as void, collapsed soil, and water accumulation. Such GPR information gave firm basis to learning the reason of faults (such as cracks, which appeared on the surface), evaluating the degree of risk for collapse, and determining repair methods.

In the last century ice bodies have changed significantly worldwide in response to climatic changes, apparently enhanced during the last decades, presumably owing to anthropogenic influence. Shrinkage of glaciers in the northern latitudes has been extensively documented; however, the behavior of glaciers in tropical or inter-tropical latitudes has been difficult to document. In Mexico, small glaciers are present in the highest mountains: Citlaltdpetl (5675 m), Popocatépetl (5452 m) and Iztaccthuatl (5286 m), all of them volcanoes. The glaciers of Mexico offer important clues about climate change at this latitude. Current inventory and monitoring of Mexican glaciers includes determination of glaciated areas and volumes. A survey has been carried out using ground-penetrating radar in order to get a figure of thickness and a preliminary volumetric estimate of an ice body of IztaccIhuatl volcano named Ayoloco glacier. A radar profile 415 m in length was obtained at a mean altitude of 5000 m with an azimuth of 190°, in a relatively flat area known as La Panza on top of the mountain. 100 Mhz antennas were used in the reflection mode with a separation of 2.5 m and a step-size of 2.5 m. A common mid-point survey was performed comprising 12 stations, at 0.5 m steps, to determine the wave velocity propagation in the ice-rock mixture that constitutes the glacier. Results yield a value of 0. 17 m/ns, which is close to the reported value for ice of 0. 1 6 m/ns. We find a layered structure perturbed by faulting at various inclinations, and an interface between the glacier and the underlying rock along a concave surface, shallow at the profile's extremes. There appears to be a transition at 50 m depth, from brittle to ductile ice. Since Iztaccihuatl volcano has many craters near the summit, we infer that this profile reveals the shape of one of the craters, which has been filled by the glacier's ice and a mixture of rock fragments and volcanic debris. The estimated upper volume of the Ayoloco glacier is 10.04 x 106 m3; this figure will have to be revised when more, perpendicular GPR transects are performed.

The advantage ofusing high-energy pulses makes classical pulse radar the favoured system for surveying very thick glaciers compared to other radar techniques. A high vertical resolution is necessary when using groundpenetrating radar for near-surface investigations at a depth range down to 50 m, like permafrost or internal layering in glaciers. Stepped-frequency radar technology offers an attractive alternative to the classical pulse radar systems. In order to compare the performances of stepped frequency instruments with a 150 MHz pulse radar during an airborne survey, both systems were used over the same area during a test flight for an ice thickness survey. The possibilities and limitations of the pulse radar and steppedfrequency technology are discussed with special emphasise on near-surface airborne investigations. Starting with theoretical considerations, an overview of the measurement campaign is presented, results are discussed, and conclusions are drawn.

The firn regime of Antarctica extends 60—100 m deep and contains at least the last 200 years of snow deposition. As part of a program to study the influence of industrialization on antarctic climate and glacial mass balance, I have acquired about 3000 km ofprofiles ofthe stratigraphy of this regime in West Antarctica using 400-MHz GPR. Here, sections of these deeper profiles are presented to show the performance that can be achieved with commercial-grade radar in polar fim and to show some of the horizon characteristics. The profiles show 1) distinct reflections that are probably thin layer responses, 2) deepest penetration to about 130 m, 3) long-distance horizon continuity (as great as 500 km), and 4) folding caused by tectonic compression. The continuity, the increase of reflection strength with depth, and the slow variation of strength with distance suggest that the reflections are caused by chemical impurities and not density contrasts. Continuity will improve with faster trace acquisition. Resolution may not improve at higher frequencies because horizons may blur from antenna motion and horizon irregularities.

A field survey organized in the framework of the Mammuthus project was recently carried out in the Taimyr Peninsula (Siberia, 26 April - 14 May 2001) in order to extract a mammoth carcass from the permafrost. The ground-penetrating radar (GPR) technique was firstly used to accurately locate the bones and tissue of the carcass. A 1 m3 block containing such remains was then extracted and transported to Khatanga City, where it is currently stored in a cave at a temperature of -12 °C awaiting to be studied by paleontologists. The second aspect of GPR application consisted in sounding the block in order to define the detailed distribution of the bones and tissue, valuable information for researchers in charge of excavation work.

The detection of discrete anomalies, such as cavities and tunnels, is an important application of crosshole radar tomography. However, tomographic inversion results are frequently ambiguous showing smearing effects and artifacts. This leads to uncertainties during interpretation and, hence, the size and shape of discrete anomalies can be interpreted only with limited accuracy and reliability. In this study, we present an adapted inversion strategy for the detection of discrete anomalies with crosshole tomography. For tomographic inversion, we use various partial data sets of specified angular aperture. The resulting tomograms contain different information with respect to the vertical and horizontal resolution of discrete anomalies. Ambiguities, such as smearing and artifacts, can be recognized and considered sidered during interpretation.F rom this, an adapted starting model is derived combining all additional information. Although the tomographic inversion results for different starting models differ significantly regarding the resolution characteristics of anomalies, the rms residuals are equivalent. Therefore, we additionally investigate the angular contribution of the residuals to the rms values, and propose another optimization criterion, the relative data misfit. It is shown, that the angular contribution of the residuals reflects the resolutionc haracteristicso ft he tomograms.

In this paper we apply a tomographic approach to the reconstruction of dielectric objects embedded in a layered medium, showing its reconstruction capability and robustness against noise. The problem is tackled with reference to a two-dimensional geometry and within the framework of the linear Distorted Born Approximation (DBA). Relying on the Singular Value Decomposition (SVD) of the operator describing the problem, the favorable effect of the increase in the band of adopted frequencies is also outlined. Numerical examples are provided to assess effectiveness robustness of the proposed approach against noise on data.

We apply linear and azimuthal multi-fold GPR techniques to study waste disposal sites and areas scheduled for quarry and civil engineering operations. Linear (single azimuth, multiple offsets) and azimuthal (multiple azimuths and offsets) multi-fold GPR data were obtained at four sites with maximum 6000% and 24000% fold for the linear and azimuthal datasets respectively. Pre-stack depth imaging of waste disposal sites provided detailed 3-D reconstructions of soil-waste and waste bedrock contacts. Post-stack migration algorithms successfully imaged quarry and building sites in limestone. Synthetic datasets, obtained by considering homogeneous and isotropic half-spaces, frequency dependent electrical properties of the media and by taking into account attenuation and dispersion effects, were compared with Common Mid Point (CMP) gathers to evaluate properties of subsurface materials. Soil-waste and rock-waste contacts, bedding j oints, fractures, cavities and filling materials in limestone were successfully imaged by migrated multi-fold data over 70% of the surveyed area. Results of processing and inversion of GPR data were calibrated by boreholes, excavations and inspections of outcrops and rock faces. An average 2% accuracy was attained in joint mapping at quarry and building sites, where field-synthetic data comparison was exploited to discriminate air/sediments- filled joints and cavities

Ground-penetrating radar (GPR) is a remote sensing technique used to obtain information on subsurface features from data collected over the surface. In this paper, we propose an automatic algorithm for estimating objects depth using f- k migration and velocity scanning method for the homogeneous medium. To improve the accuracy of the algorithm, the formula used to calculate the GPR valid lateral aperture is presented. Experimental results show that the relative estimating error of depth is as low as 5%.

Polarimetric borehole radar experiment was carried out in 2000 in Korea. Two boreholes separated by 20m were used. The host rock is granite. The cavity is located at about 80m depth. Single-hole and cross-hole radar profiles were acquired. We could clearly detect a subsurface cavity filled with air in the raw data. They have shown that cross-hole signal shows "double-dip" attenuation caused by scattering from an air-filled cavity. Although it is a simple technique, we found that it is suitable for detection of subsurface anomaly. Then we checked the attenuation between two boreholes, and showed that we can detect anomalous zone by a ray-based technique. In order to have vertical 2-D image between the boreholes, we developed a reverse time migration technique. In this analysis, we could assume two horizontal layers having different velocities, and we could image the cavity. The location of the cavity could clearly be determined by these signal interpretation.

Typical ground-pentrating radar (GPR) transmitters and receivers consist of dipole-type antennas. These antennas have pronounced directive properties and exhibit strong coupling to interfaces across which there are changes in electric material properties. Whereas coupling of antennas to smooth interfaces has been the subject of intense research for several decades, the behaviour of antennas in the vicinity of realistic small-scale heterogeneities is largely unexplored. To address this issue, we simulate the responses of a typical surface GPR antenna to a suite of scale-invariant earth models of increasing complexity. Finite-difference time-domain (FDTD) simulations demonstrate that roughness of the air-soil interface has a pronounced effect on radiation patterns. By comparison, small-scale fluctuations of permittivity only cause relatively minor local distortions of the radiation patterns.

Ground penetrating radar (GPR) was used to map the threedimensional extent and large-scale architecture of diamond and gold bearing fluvial deposits in the Potaro region of Guyana. The 100 MHz antennas achieved over 30-meter depth penetration and provided continuous, high-resolution, subsurface data. Eight radar facies were identified and used to interpret the GPR data. Bedrock surface and the bedrock-fluvial sediment interface were clearly recognized on most radar profiles. Analysis of over seven kilometers of continuous GPR data allowed the reconstruction of the 3D extent of the buried. diamond-bearing, palaeochannel, as well as that of the overburden. GPR data proved vital for cutting down exploration costs, speeding up exploration and putting the property into operation. The 3D model was used for estimating resource potential, overburden volume and for designing the layout and operation of the mine.

This paper presents a geophysical survey, which is being carried out with the use of the Ground Penetrating Radar (GPR) in an ornamental rock quarry located at Santo Antonio de Pádua area, northwest of Rio de Janeiro State, Brazil. The quarry is settled in a milonitic rock belt. It is exploited the "olho de pombo" rock, which is in commercial demand, in Brazil and abroad. The purpose of the investigation is to test the applicability of the GPR technique in order to determine the thickness of the sterile material, and the fractures within the rock massif from where the rock blocks are taken. However it will be discussed the results of two radargrams obtained on the rock without altered cap rock The study is developed with the partnership of the Centro de Tecnologia (Mineral Technology Center - CETEM), which has a great experience in deals with ornamental rock research. The acquired data were processed with the use of the software GRADIX® (version 1 . 1 1 — Interpex) in order to enhance the anomalies and the stratigraphy it was generated originated sections, with average depth of 20 meters. In the surveys it was utilized 100 MHz frequency antennae. It is intended to build a work methodology for the local mining operator in order to contribute to reduce the environmental impact problems, which in general occur within this kind of activity. It was possible to identify the fractures and also the weathering cover thickness. Presently, the requirements of govemmental institutions, the environmental impact caused by ornamental rock mining, occupy increasing distinction within mineral exploiting industry. The survey presented higer quality results.

The exploitation of limestone and gypsum quarries is often complicated by the presence of waste-filled karstic zones. The capabilities of GPR to quickly delineate the economic material are studied. Surface radar data collected in a gypsum quarry are presented. These data provide limited information due to poor radar penetration and the presence of abundant diffraction patterns of ambiguous origin. Borehole GPR surveys are considered to delineate the karstic zone. Surface and borehole radar data are first synthesized numerically to seek optimal acquisition parameters. We use Carcione and Schoenberg (2000) modeling code, in which the 3D staggered grid pseudospectral operator is implemented. Modeling results show that minimally processed borehole and borehole-surfaced ata provide valuablea nd otherwise inaccessible information. The numerical simulations have also revealed limitations of the modeling code which requires further improvement.

In this paper the mathematical models and results on processing the experimental single-frequency microwave holograms received by scanning subsurface radar with sine wave signal are submitted. The holograms reconstruction method with the use of support functions, which take into account the near field of the aperture antenna with round cylindrical waveguide, is analysed. The models consider both known and unknown phase shift of the signal reflected from the point object. It is theoretically and experimentally shown that single- frequency holograms reconstruction allows to estimate depth of shallowly buried objects and improve the resolution on the probing surface with the growth of objects depths.

For electromagnetic imaging of ground-penetrating radar ( GPR) data, the vectorial character of the electromagnetic field and the radiation characteristics of the source and the receiver play an important role. A new multicomponent imaging algorithm has been developed that takes into account these properties. This imaging algorithm employs four source-receiver configurations to obtain a bounded imaging operator. Measurements with the four possible source-receiver configurations, two copolarized and two cross-polarized, have been made across a sandbox in which variously oriented linear objects and metal spheres were buried. Essentially, electric wavefields with two electromagnetic polarizations were emitted and electric fields with the two possible scattered polarizations were measured. An automated measurement frame was used to acquire data with high positioning accuracy. Analysis of the images derived for each of the four individual data sets showed that the cross-polarized data contained useful information. The images of a horizontal pipe oriented 45° with respect to the survey lines, were almost identical for the two co-polarized configurations, as well as for the two cross-polarized configurations. A 25% increase in amplitude and a slightly improved horizontal resolution were obtained by adding the cross-polarized images to the co-polarized ones.

Recent theoretical studies have shown the sensitivities of the amplitude and waveform of a GPR reflection to the contrast in electromagnetic properties across the reflecting contact. In geological applications, however, extracting information about contrasting properties across a contact from a reflected pulse is generally difficult. Uncertainties stem largely from lack of information about the form of the outgoing pulse and lack of knowledge of the electromagnetic properties of media on both sides of the contact. Here we show that in a setting with repetitive layering, these uncertainties can be significantly reduced. An example is shown for 50, 100, and 200 MHz profiles that image subparallel dipping bedding planes in gravelly deltaic foreset facies on a Lake Bonneville delta, Utah, USA. Strongly reflecting horizons with 1-2 m spacings bound packages with finer internal layering. From finite-difference time-domain simulations of radar wave propagation, constraints can be placed on the variations in permittivity across the primary and finer-scale layering. Modeling the relative amplitudes of reflections demonstrates that the finer-scale permittivity contrasts are ~0.4-0.8 times that of the 1-2 m layering. Amplitude-versusoffset (AVO) analysis yields an upper bound of ~3.5 for the contrast in permittivity at larger-scale boundaries. Overall signal attenuation indicates the average conductivity is ~0.7-0.8 mS/m.

Local Plane-Wave processing is used to image synthetic GPR data. Numerically calculated Green's functions are used to back propagate the data down to a Green's function node, in the center of the region to be imaged. From the Green's function node a Plane-Wave approximation is used to back propagate the data over a local image. The data are correctly positioned in the wavenumber space and are re-sampled to a Cartesian grid. The data are then inverted using an IFFT.

Plane- and cylindrical-wave expansions are used to calculate the scattered electromagnetic field from a buried circular cylinder illuminated by an arbitrary antenna. With the expressions cast in a special form, the fast Fourier transform (FFT) can be used to efficiently calculate the scattered fields for fixed-offset GPR configurations.

Stimulated by the inquiry of using high frequency electromagnetic ground waves to communicate information among unattended ground sensors, numerical simulations of GPR performance in different near-surface geological settings were conducted. Two sets of 400 MHz GPR field data, one from Fort Richardson, Alaska, and the other one from Hanover, New Hampshire, were used to be the 'ground truth' to compare with numerical simulations. The numerical simulation algorithm we used adapts the finite difference time domain method, with a perfectly matched layer as the absorption boundary condition to truncate outbound waves. The signal impulse has a central frequency of 400 MHz, and the time step is 0.067 ns. We have simulated four cases: a combination of two radiation polarizations (TM and TE), and two geological settings, i.e., a sandy/gravelly half-space overlain by a silty/clayey layer (the case of Fort Richardson, AK), and a silty/clayey half-space overlain by sandy/gravelly layer (the case of Hanover, NH). The results depict the following implication. (1) More EM energy is radiated into the air as an air wave for the TM mode, and more EM energy will be sent into the ground when the TE mode is used, regardless of the geological setting. (2) Where a gravelly sandy half-space overlain by a silty/clayey layer, more EM energy will be trapped in the silty/clayey layer as a ground wave guide in the TE mode with almost no air radiation, when compared with the same radiation mode in the case of silty/clayey halfspace overlain by a layer of gravelly sandy. (3) For the geological setting of a sandy/gravelly half-space overlain by a silty/clayey layer, the TE mode only contains ground wave and the TM mode only contains air wave energy. This implies that for this case a far more complete separation of the air wave and the ground wave can be reached. These simulation results imply that transmission mode should consider the on-site geological setting when attempt to use the ground wave as a communication carrier.

The clutter caused by scattering from a rough air-ground interface is analyzed numerically. The simulations have been done using Monte-Carlo approach. An ensemble of surface profiles with the desired probability distribution and autocorrelation spectrum is simulated numerically. For each realization of the interface profile the scattered field has been calculated by a deterministic approach. Averaging of the scattered filed over ensemble of the surface realizations has been done numerically. Statistical properties of the scattered field have been analyzed. It has been found that the magnitude of the surface clutter caused by reasonably smooth surfaces can exceed easily the mean value of the ground reflection. Furthermore it is demonstrated that if the magnitude of the surface clutter is considerably less than the mean value of the ground reflection then the correlation function of the reflected field coincides with the correlation function of the rough surface.

Finite Differencing Time Domain (FDTD) modeling technique was developed as a tool to study GPR problem that could be very complex due to the antenna design, inhomogeneous soil and varieties of target types. The broadband, fully polarimetric horn-fed bowtie (HFB) antenna design (Chen, 1 997) was modeled as an example. Feeding cables, 3D antenna structure and tapered resistive loading were included. Calculated characteristics of the electrical properties of the HFB antenna in the entire 10 - 800 MHz range was obtained. Various technical issues involved in numerical modeling will be discussed.

The method of computational diagnostics has been recently developed to determine the depth oflayers and their electrophysical parameters related to structural diagnostics (monitoring of road pavements, runways, etc.). The method of computational diagnostics is based on minimizing a certain smoothing functional, which includes a functional of discrepancy between the resulting measurements of scattered electromagnetic field and the results of a direct model problem, and also a stabilizing functional accounting for a priori data on the electrophysical and geometrical parameters of a sounded object (medium). The solution minimizing this functional is found both by one of the iterative techniques (simple iteration, steepest descent, conjugate gradients, etc.) and direct techniques. The problem discussed is related to the complexity of searching for a global minimum ofthe finite dimensional spaces of the layer parameters, when this complexity is brought about by the multiple-extremality of the functional. The examples are presented of the actually restored electrophysical and geometrical medium parameters by using a video pulse ultra-widebandg roundp enetrating radar.

Characteristic features of ultra wide band (UWB) radio pulses reflection from a flat boundary between two lossy media with minimum contrast of dielectric permittivities are examined. The task has a high-priority for radar probing of upper soil layer and is connected with ranging of boundary between two media with close electrical parameters. This problem was solved many times for quasi-coherent signals and for transparent media. It was analyzed in [1] for UWB signals and monostatic GPR when difference of touching media electrical properties was pronounced (including a case of continuous vary of electrical properties near of boundary). Here we restrict ourselves in sharp border of homogeneous media in order to determine a generality of UWB signals reflection from lowcontrast boundaries.

The so-calledS uper-resolutionte chniques have often been used to improve the resolution of frequency domain Ground Penetrating Radar systems — with little understanding of algorithms. Most of these techniques are statistically sub-optimal Least Squares Estimators that require knowledge of the model describing the data. This paper develops a data model for a Stepped Frequency Continuous Wave Ground Penetrating Radar using simple scattering models and approximations to the wave-number in the medium. The Matrix Pencil of Function Method is used to estimate the model parameters from the measured data. The algorithm is tested for both simulated and captured data.

Polarimetric measurements provide additional information to aid in determining the geometric and physical properties of sub-surface targets. In this paper algorithms are derived to extract polarimetric target descriptors from a Stepped Frequency Continuous Wave Ground Penetrating Radar (SFCW GPR) data. The algorithm uses the multi-snapshot Matrix Pencil-of-function Method, to estimate the parameters of a fully polarimatric SFCW GPR model developed in this paper. The processing is applied to both simulated and laboratory measurements, and the results demonstrate the validity of this technique.

In heterogeneous sub-surface environments, the evaluation of GPR sections is complicated by the influence of nearfield effects, antenna radiation patterns, velocity variations and surveying inconsistencies. Section interpretation can be exceedingly difficult, even with advanced processing methods, and therefore mathematical modelling has become an increasingly popular addition to traditional techniques. The Finite-Difference Time-Domain method (FDTD) is the most common, but to be of practical use the modelling scheme must incorporate realistic antenna configurations, complex sub-surface geometries and accurate material property descriptions. These additional components add computational complexity to the models and, at present, most single processor FDTD schemes are only capable of modelling relatively basic three-dimensional data sets in practical time scales. Modern parallel computing techniques have the potential to overcome these limitations by spreading the computational demand across a number of processors (or individual PC's). These PC 'cluster' machines provide the necessary computational power required to model more complex GPR problems in realistic time-scales. Consequently, the scope and run-time of current GPR FDTD modelling applications can be improved making them an accessible and affordable aid to GPR interpretation.

Performances of low frequency sounding radars, in term of penetration depth and signal to noise ratio, are mainly function of electric and magnetic losses, volume scattering, and interface roughness. In order to evaluate experimentally the impact of each of those parameters on the future sounding radar missions foreseen for Mars exploration, we conducted series of measurements of the electromagnetic properties of volcanic and sedimentary materials that may be present in the near Martian subsurface layers. Results were used to construct some representatives geoelectrical profiles of the Martian subsurface for the 1-20 MHz frequency range. We considered in particular a simple 3- layered model, which is a primary model to describe terrains where recently observed fluvial-like features raise the possibility that liquid water may exist at shallow depths (100 to 500 meters). We used expected subsurface geophysical conditions such as temperature gradient, rock porosity that may exist for such sites (Clifford, 1993) to construct representative laboratory samples for each layer. We integrated the measured EM characteristics in a geoelectrical model including interface roughness and volume scatterers. We then used the Finite Difference Time Domain (FDTD) algorithm to simulate the radar backscattered echo and evaluate the ability of future sounders to detect the probable presence of ground ice — wateri nterfacei n such a geological configuration.

We present in the first part the state ofdevelopment of the laboratory prototype of the GPR which will allow to check the performances of all the sub-systems. Then some results obtained from numerical simulation are shown to demonstrate the radar capabilities and the anticipated characteristics of the detected signal. Simulated data have been used to study the algorithms which will be employed to analyse the observations and some examples of initial results are presented. Initial field measurements are reported.

The GPR data acquisition software is arguably one of the most important components of a Ground Penetrating Radar system. This software needs to ensure the capture of reliable and repeatable data by a diverse user community. Yet, too many commercial and academic systems provide a singleuser highly configurable interface for all applications. his paper describes the development of a configurable data acquisition and display pplication for use with ground penetrating radar. It is proposed that the standard GUI-based application that offers access to all controls and settings is distracting to many GPR end-users. It is therefore suggested that an application should present only the necessary controls to the end-user, and all other options should be inaccessible. This paper describes the development of such an application from a user requirements analysis to the design and implementation of an object-oriented, Java-based solution.

We have developed a miniature range-gated step-frequency (RGSF) ground penetrating radar (GPR) for use in mapping subsurface structures from planetary rovers such as the rover associated with the Mars 2009 Smart Lander Mission. At each stepped frequency the antenna is used to both transmit and receive RF radiation. Radiation scattered back from the subsurface is amplified, mixed with a frequency reference, detected as an in-phase mplitude, and transformed from the frequency domain to the time/distance domain for interpretation. When fully developed, the frequency range will be from approximately 10 MHz to 500 MHz. Compared to conventional impulse GPRs, the RGSF system has a similar total operational bandwidth, but at any instant of operation, the transmitter/receiver system is tuned to a narrowband configuration. It can provide high average transmitter power per frequency and allow low sampling rate. This capability provides both significantly enhanceddepth penetration and delineation of fine-scale interfaces. Our prototype GPR has miniaturized radar electronics (3 x 12 x 3 cm, 3W power, 30g), with a battery (14 V, 760 g)and a resistively loaded dipole antenna (2.3 m length, currently operating from 30 MHz — 150 MHz). Field testing of the prototype has been conducted over alluvium and a basaltic lava flow near the Lunar Crater Volcanic Field, Nevada, along a breakout channel north of the Silver Lake Playa, Mojave Desert, California, and over till and ash on Mauna Kea, Hawaii. Comparisons to known subsurface structures for the test sites demonstrate that the system is capable of resolving fine-scale variation in lava flow thickness, the depth to bedrock beneath the aeolian fill and alluvium in the break-out channel, and variations in ill and ash thickness, from depths from meters to approximately 10 meters.

This paper illustrates IDS dedicated tools for the high resolution inspection of walls and structures. In particular is presented here a novel Ground Penetrating Radar system based upon an array offour IDS/TR 1600 antennae, capable to acquire up to seven channels at the same time. After a brief system overview, is described the theory and are reported the results of measurements and calculations performed to obtain the transfer function of each IDS 1.6 GHz antenna. Then are reported the results of field test carried out in the "Salone dci Cinquecento", Palazzo Vecchio, Florence.

Existing positioning technologies used in conjunction with Ground Penetrating Radar (GPR) are generally too time-consuming or insufficiently accurate for high resolution, high frequency, 3-d structural investigations. In this paper we present an optical positioning system for use in GPR surveys. This system uses a camera mounted on the GPR antenna that takes video of the surface beneath it and calculates the relative motion of the antenna based on the differences between successive frames of video. Positioning using this technology can provide positioning accuracy to within several millimeters. Because the antenna can be moved free hand the procedure is orders of magnitude faster than surveying a grid of data points or laying out parallel lines and surveying each line with an antenna and odometer wheel. Time domain synthetic aperture radar algorithms reconstruct an image of the subsurface using this data. This is a new technology, but one which has potential for future research, improvements, and practical use.

In this on-going study, a combination of Ground Penetrating Radar (GPR) and electrical surveys is used to characterize the physico-chemical properties of soil, in particular those concerning lithology variations, water content, porosity and salinity in agricultural zones. These investigations will be completed by independent chemical analyses of soil and water. The experiments are conducted on fields of intensive corn culture (4 x 2 km) where the water table is 1.5 to 5 m deep, in a geological context of alluvial deposits. Analyses of water from various wells in the study area show high concentrations of nitrate coming from the chemical fertilizers spread on the soil surface. Globally, values increase from south to north and can be explained by the general direction of flow in the water table aquifer. In this ongoing study, GPR and electrical data are recorded at two sites located along a north-south line and seasonal variations will be monitored for twelve months. Preliminary results which are presented here, give structural information concerning the subsurface and show vertical flow of mineralized water (from the surface to the aquifer) providing evidence of transport of dissolved contaminants to the water table aquifer. This study demonstrates the potential of non-destructive geophysical methods for providing information on hydric and solute transfer in the ground and monitoring soil contamination.

Ground penetrating radar (GPR) has been used characterising sandy till deposits in Denmark. 2D GPR surveys are carried out at several field sites, known from the geological mapping to be located on sandy till deposits. Some field sites turned out to have a high geological variability with several different glacial deposits. Although, it is possible to bound areas with sandy till in the uppermost layer, mainly on penetration depth, but also on reflection pattern and amplitude strength. Results from the GPR surveys show that there are similarities between sandy till deposits from different localities.

Ground-penetrating radar (GPR) is potentially rich in information about both subsurface structure and water saturation, though non-unique interpretations of GPR data are possible since the dielectric constant, often inferred from GPR travel times, is a function of both porosity and water content. Further, water is often non-uniformly distributed in the vadose zone, its distribution being influenced by the spatial structure of hydraulic parameters and transient conditions at the ground surface. Synthetic time-lapsed GPR surveys are shown useftil in this work for improving overall characterization of the Vadose zone and for giving valuable information about flow processes. Variably saturated flow and GPR are simulated simultaneously for a highly heterogeneous vadose zone (outcrop) model in order to obtain synthetic GPR images before and during infiltration experiments. Aside from delineating the advancing water front, GPR crosshole surveys obtained during redistribution of water (after the infiltration front passes) provide additional information. Transients in water content, and therefore in electrical parameters, are seen to be related to soil type, and even more so to the spatial structure of fluid permeability. Instead of requiring a direct relationship between the dielectric constant of soil and its permeability, the goal of this work is to take advantage of the sensitivity of GPR to changes in water content and ultimately relate estimated changes in water content to fluid permeability through inverse modeling.

Three-dimensional modeling of Ground-Penetrating-Radar (GPR) surveys is increasing in popularity and demand. Combining GPR data with other georeferenced data using Global Positioning Systems (GPS) and Geographic Information Systems (GIS) would allow for more detailed and efficient site analysis and planning. Integrating GPR and GPS data into a geographic information system (GIS) has not been investigated in Florida. Thus, the objectives ofthis research are to (i) identify subsurface soil horizons using GPR, (ii) create a three-dimensional (3D) subsurface model based on imported GPR data, and (iii) develop a procedure to import GPR data into a GIS. The soils investigated are located in Marion County, Florida, and are characterized by Pleistocene-age sands over the clayey, marine deposited Miocene-age Hawthorn Formation, which drapes the Eocene-age Ocala Limestone. GPR transects of different lengths and spacing were collected and georeferenced using GPS to determine the subsurface variation of the Hawthorn Formation and location of karstic features. The GPR and GPS information was incorporated into ArcView and ArcGIS software. The information generated by integrating GPR, GIS, and GPR may assist researchers in determining best management practices including experimental plot placement, irrigation management, fertilizer treatments, and pesticides applications.

Recognizing the importance of quantitative estimates of GPR velocities, we have implemented several standard approaches as public domain software. A field observation that a rainfall event of 3.5 cm was associated with a decrease in the velocity of a shallow reflected phase from 0.104 to 0.085 mlns, along with antecedent literature on the effects of soil water on GPR velocities, underscored the potential hydrological applications of such results. The algorithms we favor are robust in implementation, provide accurate results, yet are based on methodologies readily grasped by nonspecialists and beginning students. Some applications involve such housekeeping elements as format conversion; filtering; gain options; and resampling using FFT interpolation. More useful to the general user are analysis packages for research and teaching: NMO/LMO velocity scans; time-base stretching; event picking; windowed scaling; optimal stacking; and cross-correlation analysis. Testing our procedures on an air wave phase, its velocity from zero-break picks was 0.27 1 m/ns, and from thresholdbreak picks was 0.284 m/ns. These are inconsistent and biased low. However, a cross-correlation procedure with an optimized native wavelet provided an estimate of 0.304 mlns, much closer to the expected value. Such routine checks corroborate the accuracy of particular procedures, and identify potential timing problems.

The Pleistocene push moraines in the Netherlands were formed during the penultimate glaciation. The data presented were collected at the eastern Veluwe ridge. Based on extensive 50 MHz ground-penetrating radar surveying, three architectural styles are distinguished. In order to deduce relationships between these structural styles, sediment facies and radar facies, the GPR surveys were complimented by data from continuously cored boreholes. Additional borehole logging served the same purpose. In this paper GPR facies of two glaciotectonic styles are analyzed: gently folded parallel fluvial sequences and imbricated thrusts in fluvial sequences. It is shown that coherent reflectors, using 50 MHz central frequencies, can be related to relatively thin (8-20 cm) fine-grained sedimentary layers. These layers are not necessarily clay, but more often silty fine sand. The presented radar facies show strong relationships with sedimentary facies as evidenced by lithology in the core and the accompanying gamma-ray borehole logs The result leads to the conclusion that ground-penetrating radar is an excellent tool for mapping the complex architecture of Pleistocene push moraines and that a combination of techniques is particularly powerful.

A study using GPR has been carried out at Itaipuaçu beach,Marica, Rio de Janeiro with the aim of reconstructing and analysing the environment of depositing sediments. Fifty profiles were made using a 200 MHz bistatics antennas (nominal frequency). These profiles generated a 3D cube. In the same way, we obtained an estimation of propagation velocity of the signal in the subsoil (by CMP technique) using two 80 Mhz antennas. The analysis of the radargrams allows to recognize the stratigraphic sequence and its progradation. Through core drill sampling in the investigation site, it has been possible to establish a correlation between some reflections and the presence of organic matter, and to make evident that the salt content of water below the water table increases with depth. With these profiles we can see the progradations, and under the water table, a reflector which seems to be caused by the combination of organic matter, together with a sharp increase in the salt content of the water. This interpretation agrees with the model "Ghyben-Herzberg fresh-water lens" Hagrey & Müller,( 2000), as well as other results obtained by these authors.

GPR is considered ineffectual in penetrating silty soils. Capillary tension appears to move the interstitial water dielectric relaxation frequency near the GPR range. If hydrocarbons relieve the tension, then GPR attenuation rates should decrease. We profiled these rates across a partly contaminated quartz silt diamicton at an inactive fuel terminal in Haines, Alaska. We calculated the rates from the signal amplitude decay within a series of 400-MHz moveout surveys recorded along a single transect. We sampled the hydrocarbon content along the transect and on the wall of a trench we excavated beside it, along with water content, density and resistivity. The silt was saturated and generally at 50—80 ohm-m. The ground impedance loading brought the antenna pulse center frequency down to near 200 MHz. The rates ranged from 5 to 8 dB/m in the contaminated zone and about 10 to 15 dB/m in the uncontaminated zone. The resistivities account for the contamination rates, but a dielectric relaxation centered near 1—2 GHz, using a simple mixing model, is required to account for those in the uncontaminated zone. Hydrocarbon soil analysis suggests that the threshold level for the attenuation effect is not higher than about 1000 ppm. We suggest GPR offset surveys for assessing contamination zones.

Ground-penetrating imaging radar ("GPiR") combines standard GPR with accurate positioning and advanced signal processing to create three-dimensional (3D) images of the shallow subsurface. These images can reveal soil conditions and buried infrastructure typically down to depths of about 2-3m with high resolution. A commercial GPiR called the CART Imaging System, which was designed for mapping urban infrastructure, has been developed. The CART system uses a radar array consisting of 17 antennas (9 transmitters and 8 receivers) that cover a 2m swath on the ground and can collect data while moving at speeds up to about 1 km/h. A laser theodolite tracks the position of the array during operation. The system collects enough data in a single pass to form a 3D image beneath its track; side-by-side passes are stitched together to create a seamless image of the subsurface. GPiR was first tested on a large scale in a project that mapped an area of approximately 12,000m2 in the south Bronx in four nights. Positions of surface features were also surveyed with the theodolite to provide a local reference grid. Final images were visualized with large-scale maps and electronic movies that scroll through the 3D data volume and show the enormous complexity of the subsurface in large cities.

Subsurface mapping using a single-antenna GPR is a time-consuming operation especially when large areas are to be covered. The data acquisition can be performed much more efficiently using an electronically scanned antenna array. The stepped-frequency 3D GPR that is developed at the Norwegian University of Science and Technology uses a 1 meter wide antenna array that consists of 31 transmitlreceive antenna pairs. The main application of the system is utility mapping of streets where the underground consists of a jungle of cables and pipes. The data from the 1 meter wide swath is focused into a 3D image cube using 3D wavenumber migration. By combining several parallel swaths, it is possible to generate underground maps of the whole area at different depths. The radar has successfully been used for mapping of pipes, cables and old tramlines in Trondheim during 2001. The wide bandwidth (100 MHz — 1 .6 GHz) gives high enough resolution to map shallow features such as asphalt thickness and base layer structure as well as deep structures such as ground composition and utility infrastructure . In this way, the data from a survey can serve more than one customer and be stored in databases for future retrieval. The results from the field tests demonstrate the much higher user friendliness and quality of data for true 3D imaging as compared to standard 2D GPR profiling.

This paper analyzes the early-time radar response of buried penetrable targets such as plastic landmines. The Born approximation is used to derive simple analytical expressions relating target and soil properties to the earlytime response. Understanding these dependencies is crucial for target identification under varying soil conditions. The derived expressions include the transfer function and the impulse response of a penetrable target embedded in an unbounded homogeneous lossy medium and illuminated by a uniform plane wave. Using a truncated circular cylinder having the dimensions of a PMA-3 mine as an example, the early time responses predicted by the Born approximation are compared against responses obtained by threedimensional finite-difference time-domain (FDTD) simulations. The results demonstrate that with the Born approximation it is possible to predict the general shape of the target response, i.e. the number of amplitude peaks, as well as the amplitudes of those peaks that relate to backscatter from the top of the example target. To improve the fit between the predicted and simulated responses, two phenomenologically motivated modifications to the earlytime response expressions are proposed. The modified expressions are able to accurately predict not just the general shape of the early-time response, but also the influence of the host medium conductivity on the target impulse response.

A prototype of an ultra-high frequency radar system (2- 60Hz) has been developed at the RST laboratory within the framework of the HOPE project funded by the European Community and aimed at integrating three sensors (metaldetector, GPR and microwave radiometer) into a unique portable system for humanitarian demining. An advanced prototype of the GPR sensor assembled with the dual metal coil MD sensor has been recently tested at the outdoor facilities of the Joint Research Center in Ispra (Italy). The test field specifically prepared by JRC consists of a unique target scenario that is recreated under different type of soils and surface conditions. The target scenario includes different type of mines and false alarm targets like stone, wood, metallic or plastic objects. The dataset collected during this test are quite interesting for planning the future improvements of both the hardware and the software solutions. The data has been processed with a 3D imaging software specifically developed by the authors for the HOPE project. The preliminary results are encouraging for some scenarios whereas some others seem to be really demanding for the GPR sensor.

A full-polarimetric GPR front-end has been developed. The front-end comprises a generator section, a multi-static antenna system and a receiver based on a multi-channel sampling converter. The front-end allows to perform subsurface imaging with resolution and accuracy sufficient for antipersonnel mine recognition. In comparison with commercially available video impulse GPR systems the key advantages of the front-end are considerably larger bandwidth, very high precision of measurements of scattered field and ability to measure polarimetric structure of the scattered field.

Optimal frequency bands for radar detection of buried anti-tank (AT) and anti-personal (AP) mines are substantiated. Mines geometry, permittivity of explosives and soils are taken into account. In that case we can obtain maximum number of parameters for identification of mines against a background of noises and false signals. Analysis of used antennas of ground penetrating radar (GPR) is completed. For improvement of mine detection probability we propose to use scanning antennas and focused scanning antennas.

The OSU/ESL GPR systems have been applied to the detection and classification of buried unexploded ordnance (UXO) for years. It has evolved from an impulse and single-polarization (cross-polarization) system utilizing complex natural resonance (CNR) feature to the recent step-frequency and fully polarimetric system utilizing CNR, polarization and scattering features. Significant progresses in measurement techniques, feature extraction algorithms and classification rules have been made during the past three years under the support US DoD ESTCP program. These important progresses were motivated by field data collected at government test sites such as Tyndall AFB (1999), Blossom Point (2000) and Jefferson Proving Ground (2001). This paper briefly describes these progresses and the motivations behind them.

A new multistatic phase-based GPR technique for precise characterization of rough surfaces in three-dimensional space and for ingenious detection of weak-contrast scattering objects buried beneath the air-ground interface is proposed. The technique has been tested by the data measured using wideband stepped-frequency GPR over the compound surface, which consists of flat surface, several small isolated areas of rough surfaces and areas with shallowly buried plastic objects as surrogate landmines. The maximum height of rough surfaces and the maximum depth of buried dielectric objects are made less than one practical range resolution cell. The buried objects have low dielectric contrast against their surrounding medium in the frequency range used in the measurements. The antennas adopted are broadband, have low gain and wide beamwidth. The results show good agreement with the actual distribution of the discrete scattering objects such as buried plastic M14 type mine (5 cm in diameter and 4.2 cm in height) and with the real shape of continuous scatterers (rough surfaces). Moreover a new spatial-based processing method is described for noise reduction and the measurement errors. The processed results show better accuracy and focusing patterns than those previously obtained. This technique neither requires a priori knowledge about the background medium nor needs any pure background measurement.

Based on the strengthening example of the historical building, the Shanghai Huifeng Bank Mansion, the methods and efficiency of GPR in non-destructive testing on concrete buildings are discussed in detail. Its successful application into the durability diagnosis of historical buildings is also demonstrated.

The objective of this paper is to examine the structure of the image reconstruction algorithm for synthetic-aperture GPR systems operating with pulse-echo and step-FMCW illumination schemes. The main structure of the image formation algorithms is based on the framework of the backward propagation image formation technique. Mathematical modeling, theoretical analysis, and results from full-scale experiments are included.

In this paper a high-frequency large-bandwidth synthetic - aperture penetrating radar for inspecting masonry structures is described. A Continuous Wave Step Frequency (CW-SF) radar operating at 10 GHz centre frequency with 4 GHz bandwidth has been designed and built. The system is operated in a non-contact manner by mechanically moving the transmitting and receiving antennas in order to synthesize a N-dimensional aperture up to 3 m in length and 1 m in height. In spite of the fact that penetration depth decreases dramatically with increasing frequency, a penetration depth up to several tens of centimetres in masonry, that can be satisfactory in a number of applications, was observed. The high central frequency, making available very large bandwidths, provides high resolution images of the investigated structures. Laboratory tests in order to assess the performances of the system are reported.

Bow-tie antennas are often used for investigating subsurface features in concrete structures. It is important that the antenna selected for a survey has suitable detection capabilities and is also physically small enough to be usable within the confines of the structure. It should also be noted that the interaction of the antenna and material under inspection may change the characteristics of the radar signal being transmitted and this may have implications for interpretation ofresults. Higher frequency antennas normally give a better resolution but poorer depth penetration capability to lower frequency antennas. An experimental programme is described in which the resolution capabilities of two commercially available 900MHz & 1GHz antennas are compared in air. Studies are also carried out using an antenna in contact with concrete and the practical results are compared with theoretical predictions of resolution expectation. These results are considered in terms of the practical implications for field-testing using radar

The study of root growth and development in soil has been intellectually and technically challenging. In response to concern about increasing levels of atmospheric carbon dioxide (CO2), resulting from increase in global energy use, the cycling of carbon has become the object of many intensive investigations.. Terrestrial ecosystems are a huge, natural biological scrubber for CO2 currently sequestering, directly from the atmosphere, about 22% of annual anthropogenic carbon emissions. It is assumed that a significant fraction of this carbon uptake goes into roots. Presently, there are no means by which root morphology, distribution, and mass can be measured without serious sampling artifacts that alter these properties. This is because the current methods are destructive and labor intensive. A non-invasive, imaging procedure for examining roots in situ would be a powerful tool quantifying subsurface storage, as well as for documenting changes in root structure. Preliminary results using a high frequency, 1.5 Ghz, impulse Ground Penetrating Radar (GPR) for nondestructive imaging of tree root systems in situ are presented. Two 3D reconstructed images taking advantage ofthe polarization effect are used to assess root morphology and dimensions. The constraints, limitations, and potential solutions for using GPR for tree root systems imaging and analysis are discussed.