Ground Penetrating Radar Profiles Research Articles

An Improved Synchrosqueezing S-Transform and Its Application in a GPR Detection Task

The S-transform is a fundamental time–frequency (T-F) domain analysis method in ground penetrating radar (GPR) data processing and can be used for identifying targets, denoising, extracting thin layers, and high-resolution imaging. However, the S-transform spectrum experiences energy leakage near the instantaneous frequency. This phenomenon causes frequency components to erroneously spread over a wider range, impacting the accuracy and precision of GPR data processing. Synchrosqueezing is an effective method to prevent spectrum leakage. In this work, we introduce the synchrosqueezing generalized phase-shifting S-transform (SS-GPST). Initially, it resolves the compatibility issue between the S-transform and the synchrosqueezing strategy through phase-shifting. Subsequently, the SS-GPST accomplishes spectral energy focusing and resolution enhancement via a generalized parameter and synchrosqueezing. A synthetic signal test shows that the SS-GPST excels over other methods at focusing degree, spectral resolution, and signal reconstruction accuracy and speed. In actual GPR tunnel detection data processing, we assess the adaptability of the SS-GPST from three aspects: spectral energy distribution, thin layer identification, and data denoising. The results indicate: (1) compared to other methods, the SS-GPST accurately expresses spectral components with a strong focusing degree and fewer interference components; (2) high-frequency slices of the SS-GPST accurately detect the top and bottom interfaces of a 3.0–3.5 cm reinforcement protection layer; and (3) due to fewer interference components in the SS-GPST spectrum, reconstructing GPR profiles through the SS-GPST inverse transform is an efficient denoising technique. The SS-GPST demonstrates adaptability to different data processing purposes, offers high-resolution T-F spectra, and shows potential to supersede the S-transform.

Joint imaging of ERT datasets and its application in seepage characterization at Nanshan Dam, southeast China

Estimating resistivity distributions within dam structures accurately using electrical resistivity tomography (ERT) poses a significant challenge due to the limited 2D acquisition scheme, the 3D heterogeneity of near-surface materials, and the dynamic nature of fluid transport. To address this challenge, we present a new strategy by simultaneously inverting multiple collinear ERT datasets acquired from various standard electrode arrays. This strategy incorporates the coupling of the intrinsic parameter relationship into the objective function, constraining the inversion solution space by balancing the distinct resolution characteristics associated with the electrode arrays involved in the joint inversion process. Using synthetic Karst and infiltration models, we demonstrate that our strategy can yield accurate inversion results for both the Wenner and dipole-dipole datasets. Application of this strategy to a field case at Nanshan Dam in southeast China, which involves Wenner, Wenner-Schlumberger, and dipole-dipole surveys, successfully delineates preferential fluid seepage pathways. These findings are further corroborated by consistent inferred reflectors in Ground Penetrating Radar (GPR) profiles. The synthetic and field examples highlight the effectiveness of our strategy in achieving accurate and unified resistivity estimates by integrating multiple ERT datasets.

A multiproxy analysis of modern environmental change within a cypress swamp forest, Collier County, FL

The Florida Everglades are a vast subtropical wetland that historically spanned over 1,000,000 hectares, but much of the Everglades has changed in the last 100 years due to anthropogenic activity. Collier County, situated in southwest FL and bordering the Gulf of Mexico, was subject to alteration in the form of logging, road building, and canal digging. These actions disrupted the natural sheet flow of water and had large environmental impacts on the region, impacts which are slowly being addressed by Everglades restoration efforts. The aim of this work was to observe environmental change at a cypress swamp forest in Collier County within the Big Cypress National Preserve. Using sediment core data including charcoal analysis, loss on ignition, and peat humification, as well as remote sensing techniques, this project uses a novel approach to assess local environmental conditions in the modern era for the region. While this analysis was based on a single core located at the study site, additional ground-penetrating radar profiles at the study site showed a consistent, laterally continuous distribution of subsurface interfaces, therefore suggesting that the core is representative of the conditions at this specific location. Historical records and contemporary data are used to evaluate environmental change over time, and satellite imagery is used to quantify vegetation health. Environmental change related to anthropogenic activity is noted, and evidence of progress from restoration efforts is observed from the last two decades in our study’s data.

The last remnants of the Italian Apennine glaciers: 3-D modeling of the Lower Calderone Glacieret by the IDW spatial interpolation of GPR data

Abstract We present a quantitative ground-penetrating radar (GPR) analysis of the Lower Calderone Glacieret to highlight the recent evolution of one of the southernmost glacial systems in Europe. The Upper and Lower Calderone Glacierets are the last two perennial ice bodies in the Apennine Mountains (Italy), and their continuous monitoring is important for glaciology, hydrology and climate science. We applied a proprietary auto-picking algorithm to track reflections accurately and objectively within three pseudo-3-D GPR data sets that were acquired over the Lower Glacieret in different years. After the time-to-depth conversion, the basal reflections were projected onto the normal versors of a 3-D topographic model of the survey area, at the different GPR trace positions. We then applied an Adjusted Inverse-Distance Weighted Spatial Interpolation method to extrapolate the ice-bedrock interface within the areas not directly covered by the GPR profiles and compare it with the topographic surface to recover the glacieret volume. In this paper, we critically examine the accuracy of the reconstructed models, including possible issues related to the challenging survey areas, such as local artifacts in the interpolated interface caused by irregular GPR coverage. We further discuss the various advantages of the implemented procedure with respect to more traditional glacier monitoring techniques.

Near-Surface Geophysical Characterization of a Marble Deposit to Promote a Sustainable Small-Scale Mining

Small-scale mining (SSM) is responsible for almost all the production of non-metallic minerals in the world and represents around 80% of the mining in Brazil. The lack of direct geological information increases the level of uncertainty associated with the exploratory process, compromises mine planning, limits mineral extraction, and contributes to maximizing environmental issues. In this research, near-surface geophysical methods, including Electrical Resistivity, Capacitive Resistivity, Ground Penetrating Radar (GPR), and Transient Electromagnetic (TEM), were applied to characterize a marble deposit in an SSM located in the Campos do Jordão region, São Paulo state, southeast Brazil. The geophysical methods used provide indirect information about the subsurface geology based on the contrast in electrical and electromagnetic properties. Resistivity results show the efficiency of locating marble deposits, as well as fracture zones. GPR profiles allowed for the investigation of the structural heterogeneities in the subsurface. Geophysical data and lithological information from drill holes were integrated into Micromine software and guided the development of a geological model and a conceptual pit model. The information inferred from the pit modeling allowed us to analyze the potential of the deposit and should be used to assist in developing sustainable mining planning. The results of this work demonstrate that the investment in geophysical research can support the modernization of an SSM and contribute to more sustainable and productive mining.

Internal structures of the Last Glacial Period multi-phase cold-climate aeolian dunes, Moravian Sahara, Czechia

Aeolian dunes in the Moravian Sahara developed during at least two distinct cold periods of the Last Glacial Period (LGP). Their two-phase development was previously deduced solely from the different orientations of dune crests without knowing the dunes' internal structure. Therefore, the aim of this study is to reveal the internal structure of the multi-generational dune system, developing a relationship between these dunes and bedrock. A combination of data was obtained from boreholes and ground-penetrating radar (GPR). GPR profiles were measured in places where the boreholes were drilled to obtain high-resolution combination data. Furthermore, models of aeolian sediment thickness and ground-water level were made using a spatial interpolation method (ordinary kriging with spherical semivariogram model) to interpret the GPR results and to investigate a relationship between the bedrock and dune patterns. Internal dune structures confirm the hypothesis of multi-generational development of this dune field. In addition, the aeolian sediment distribution was influenced by the bedrock topography. Finally, this study confirms that a change in LGP atmospheric circulation over Central Europe is the main controlling factor in forming the Moravian Sahara dune field.

Geophysical and chemical characteristics of peatland in coastal wetland, southern Thailand

Peatlands, as an important carbon sink, store a large amount of atmospheric carbon that is vulnerable to becoming a carbon emission source due to the disturbance of the peat layer by natural and anthropogenic processes. The Kuan Kreng peat swamp forest is the second largest peatland in Thailand and is in coastal wetland serving for carbon stock in the subsurface. To assess carbon stock in peatland, the thickness and distribution of the peat layer are necessary. This study determines physical and chemical characteristics of the peatlands in the coastal wetland, southern Thailand by conventional core study and geophysical surveys, including ground penetrating radar (GPR) and electrical resistivity imaging (ERI). There is a significant relationship between chemical and physical properties from the peat layer, which is useful to approximately predict the physical properties and peat layers in the geophysical profiles. The resistivity profiles exhibit high resistivity response, peat between 21.9 and 145.0 ohm-m, interpreted as the peat layers in the shallow subsurface. The high resolution GPR profiles provide an average GPR velocity in peat of 0.040 m/ns and a relative dielectric constant of 54.9 for the peat layers. The peat layer is characterized by a high amplitude of reflection in the GPR profiles. The thickness of the peat layers was estimated from the geophysical surveys and the drilling cores having an average thickness of 18 cm. Average values of bulk density (0.19 g/cm3) and TOC (31.18 wt. %) from the drilling core samples are used to calculate the carbon density in the peat layers giving a result of 59.24 Kg C/m3. Hence, the carbon stock at the Kuan Kreng peat swamp forest is estimated at least 7.53 megaton carbon, which is much higher than aboveground biomass carbon in this area. The innumerably high amount of carbon stock could raise awareness of peatland disturbances.

GPR detection localization of underground structures based on deep learning and reverse time migration

Ground penetrating radar (GPR) is widely used in detection localization of underground structure anomalous bodies. However, it is impossible to achieve accurate imaging localization of underground structures targets only by the waveform of GPR detection profiles due to the interference of clutter signals in engineering application. In this paper, a RefineNet deep learning network based on hybrid loss function is proposed to remove clutter signals in GPR profiles, and a reverse time migration (RTM) imaging algorithm based on total variation(TV) regularization is developed to perform migration imaging on GPR profiles processed by RefineNet network to realize accurate imaging localization of underground structures targets. Numerical simulations demonstrate that the developed RefineNet deep learning network can significantly remove clutter signals and improve GPR profiles quality, and the proposed RTM imaging algorithm based on TV regularization can accurately obtain location and shape information of underground structures targets. The full-scale experiment and field detection show that the proposed algorithm can accurately obtain the location of underground pipelines, cavities and other anomaly bodies. It is concluded that the developed algorithm provides a strong technical support for the GPR detection localization of underground structures targets.

Using ground-penetrating radar to investigate deposits from the Storegga slide tsunami and other sand sheets in the Shetland Islands, UK

We use ground-penetrating radar (GPR) to investigate the geometry, inland extent and continuity of sand layers interpreted as tsunami deposits in the Shetland Islands, UK. Four sites where sand layers within peat deposits have been recorded in previous studies are used to provide ground truth. In addition, we describe survey results from one site where deposits are not exposed to test the potential of GPR to identify candidate tsunami deposits in areas that are not well documented. Sand layers can be clearly imaged at all five locations because they are interbedded with peat and the contrast in lithology gives a good reflection on GPR profiles, even very thin sand layers, <1 cm thick, that are beneath the theoretical resolution of the GPR. 2D and 3D surveys show that tsunami deposits appear to drape a buried topography. Most sand layers form continuous reflections, although some gaps are attributed to later erosion, most likely by streams. Sand layers have been traced up to 150 m inland and 10 m above the present shoreline, which is consistent with data from boreholes. If a similar sized event occurred today, it would have a devastating impact on the Shetland Islands.

Archaeological geophysical investigation of Uzun Rama Steppe kurgans, Goranboy

Geophysical methods can provide valuable information when imaging man-made subsurface structures prior to archaeological excavations. An archaeological geophysical survey was conducted on funeral chambers (also known as kurgans) from the Early Bronze Age, in the Shadili-Uzun Rama Steppe of the Goranboy region, near the Kurakchay river gorge in Azerbaijan. This multimethod survey is based on a ground-penetrating radar profile, heat map of the total magnetic intensity, and contiguous profiles of electrical resistivity and seismic refraction. Brief processing and default inversion methods enabled us to obtain geophysical images in accordance with the lithology (a subsoil comprised of alluvial-proluvial conglomerate terrasse). The shape of the kurgans was recovered mainly through electrical resistivity tomography, enabling future targeted excavations. Overall, this study further testifies that geophysics provides valuable information for archaeological investigations, although higher resolution could be achieved by using more advanced field methodology and processing.

Non-invasive (georadar) investigation of groundhog (Marmota monax) burrows, Pennsylvania, USA

Zoogenic impact plays a critical role in stream processes, especially bank stability and resulting channel dynamics. This study focuses on bioturbation by groundhogs (Marmota monax) along the riparian zone of Mill Creek (Bucks County, Pennsylvania, USA). Several complexes comprising at least 32 active burrows (average diameter: 25.9 cm) were geolocated, with morphometric measurements obtained at selected sites. Two networks were imaged using high-frequency 800 MHz ground-penetrating radar (GPR) and included: 1) a grid of parallel 3-m-long transects on the south bank, and 2) an 11-m-long profile on the north bank. Post-processed electromagnetic signal traces (A-scans) comprising 2D radargrams (B-scans) revealed voids as reverse-polarity anomalies (hollow inclined shafts and tunnels), allowing for a general assessment of burrow depth and orientation. At the southern cutbank site, a large burrow had an entrance diameter of 0.3 m and a westerly dip. A sloping tunnel section was detected at ~0.5 m depth, based on the geometry of point-source (transverse) hyperbolic diffractions corresponding to the roof and a floor ‘pull-up’. The second locality traversed three open burrow entrances adjacent to large tree roots. This survey along a tributary channel shows multiple hyperbolics below adjacent openings, with the latter showing the characteristic signal ‘breakout’. GPR data show hyperbolic signatures ~0.3–0.4 m below the ground surface. Along this transect, burrowing activity appears to increase with proximity to the northern bank of Mill Creek. An example of a depth slice (bedding-plane view) from a nearby riverbank demonstrates the potential for 3D visualization (C-scans) of burrow networks using a grid of closely spaced GPR profiles. Groundhog burrows constrain maximum long-term level of the groundwater table and serve as important zoogeomorphic structures in diverse ecotones, including developed landscapes. Abundant evidence of bank slumping, incision, and treefall suggests that burrowing activity likely weakens root systems and enhances groundwater flow, thereby initiating or accelerating geomorphic cascades leading to slope failure.

Ground penetrating radar (GPR) analysis of the deformation in the submerged drainage route at Ashour Bridge, El Bahira governorate, Egypt

ABSTRACT An underground structure called a siphon allows irrigation of water canals and channels to pass underneath water drains. Three groups of loads that may have an impact on the siphon’s Geotechnical stability are: (1) loads brought about by traffic and truck duties on bridges; (2) loads brought about by drain water on the siphon’s body; and (3) loads brought about by irrigation channels and the dynamic movement of sandals and bots. To address the problem, surveying was done to see if the body of the submerged Syphon had any deformations. A ground penetrating radar study of the research area will be used to pinpoint the exact location of the damaged parts and the areas of deformation in the body of the submerged Syphon. On the southwesterly side (the exit), three GPR profiles were made, and on the northeastern side (the mouth), five. To display the covered drainage body’s infrastructure, the measured GPR data were processed differently. Deformities and distortion areas in the body of the submerged siphon were discovered at various places and depths on the two sides of the siphon after the processed data were analyzed.

Set of Engineering and Geophysical Methods for Determining the Volume of Homogeneous Industrial Waste at Sites of Accumulated Environmental Harm

The paper presents the application results of geophysical methods (micro-near-field time-domain electromagnetic sounding and ground penetrating radar profiling) performed in 2022 to determine the volume of homogeneous industrial waste in collection ponds using the example of the phenol lake within the city limits of Ulan-Ude , The Republic of Buryatia. The boundaries and volumes of industrial waste disposal were determined, and the position of the overlying and bottom beds of the phenol pound were recorded. The migration channel of hydrophilic chemical pollutants from the bed of the studied object was identified and the reservoir was installed to collect, over the years of infiltration processes, fluid saturated with hydrophilic chemical pollutants from the phenol pound. 3D models of the industrial waste deposits in the phenol lake and 3D models of the identified collector were built. It was concluded that the used set of geophysical methods was effective for solving similar engineering, environmental and geoecological problems.

The current state of Lake Dålk (Larsemann Hills, East Antarctica)

The study focuses on the current state of englacial Lake Dålk (Larsemann Hills Oasis), which was completely devastated in January 2017 by a catastrophic outburst and formed again in February 2020. A set of works including ground-penetrating radar (GPR) profiling, geodetic survey, core drilling, thermometric measurements and isotopic analysis was performed. As a result, new data were obtained about the modern boundaries of the reservoir, ice thickness and depths, as well as about the peculiarities of freezing of the lake after re-formation. Based on the field data obtained, one-dimensional mathematical modeling of Lake Dålk freezing was carried out, the results of which showed further scenarios of its evolution.

Integration of in‐mine seismic and GPR surveys to gain advanced knowledge of Bushveld Complex orebodies

AbstractImproving the exploration of deep‐seated mineral deposits and assessing the stability of the mine pillars require that geophysical techniques are deployed in a fast and cost‐effective manner with minimal environmental impact. This research presents results from in‐mine reflection seismic experiments and a ground penetrating radar (GPR) survey conducted at the Maseve platinum mine, South Africa. The research aims to develop and implement methods to image platinum group metal (PGM) deposits and geological structures near mine tunnels and assess the stability of pillars. The seismic experiments were conducted using a sledgehammer source (10 lb), conventional cabled geophones (14 Hz), and a landstreamer with 4.5 Hz vertical component geophones. The GPR survey was conducted using a Noggin 500 GPR system with 500 MHz centre frequency. An image of the underlying orebody and geological structures down to 100 m from the mine tunnel floor (∼500 m below ground surface) was produced. We correlated the coherent reflections beneath the tunnel floor with a known Upper Group (UG2) PGM orebody. The final seismic section shows that the UG2 mineralisation is dissected by near‐vertical dykes, faults and fractures. These structures, faults in particular, are interpreted to have been active post‐mineralisation, implying that they may have contributed to the current complex geometry of the deposit. Four GPR profiles were collected around a stability pillar adjacent to the seismic lines. The radargram sections were processed to improve the signal‐to‐noise ratio (S/N). The results show different patterns of fracturing and stress‐induced structures. These fractures were shown to be sub‐vertical and, possibly, constitute complex micro‐structures within the pillar, which could compromise the pillar stability and integrity. The study demonstrates that in‐mine seismic and GPR surveys can be cost‐effective and valuable for mineral exploration.

Integrated Geophysical and Geochemical Analyses for Assessment of Potential Coal Prospects in Tirah Area, Khyber Pakhtunkhwa, Pakistan

This study presents a comprehensive investigation of the Paleocene coal in the Tirah area, District Khyber, Khyber Pakhtunkhwa, Pakistan, utilizing onsite Electrical Resistivity Tomography (ERT) and Ground Penetrating Radar (GPR) experiments combined with geochemical analyses of obtained coal-rock samples. ERT and GPR profiles, along with geochemical data, were collected and meticulously analyzed to assess the potential and quality of coal seams based on coal ranking, aiming to delineate their spatial distribution and future exploitation prospects. The study reveals a wide range of resistivity values, spanning from 8.93 Ωm to 2472 Ωm. The uppermost layer, comprising silt and clay with water saturation, exhibits resistivity values ranging from 8.93 Ωm to 50 Ωm. The subsequent stratum, characterized by wet sandstone with minor shale and clay, has resistivity values between 50 Ωm and 95.3 Ωm. Significantly, the zone with resistivity values between 95.3 Ωm and 800 Ωm is identified as the probable host of the coal seams. Based on the results, the coal seams are estimated to exist at depths ranging from 14 to 23 m, with resistivity values between 95.3 Ωm and 800 Ωm. GPR results further corroborate these findings, revealing coal-bearing strata at various depths, thus confirming the ERT results. Additionally, proximate and ultimate analysis provided insights into the quality of the coal, with average concentration values of moisture content (6.95%), ash content (22.12%), volatile matter (28.55%), fixed carbon (42.40%), carbon content (55.35%), hydrogen (4.7%), nitrogen (0.95%), sulfur (4.82%), and oxygen (10.06%). Comparative analyses of the obtained results with local coals from the Salt Range in Punjab, Pakistan, and coal from Ogboyaga, Nigeria, confirm that the studied coal ranks as sub-bituminous. This comprehensive assessment offers valuable insights into the potential and quality of Paleocene coal in the studied region and is applicable to other areas with similar geological settings.

Application of Ground-Penetrating Radar with the Logging Data Constraint in the Detection of Fractured Rock Mass in Dazu Rock Carvings, Chongqing, China

Geologic interpretation results from conventional ground-penetrating radar (GPR) detection methods tend to have a certain degree of uncertainty. In order to improve the reliability of ground-penetrating radars in the detection of rock mass fissures in grottoes, this study proposes a ground-penetrating radar detection method with the logging data constraint, which is applied to detect the fractured rock mass in the Baodingshan Scenic Area of Dazu Rock Carvings, Chongqing, China. First, conventional logging and borehole televiewer data were compared and verified, yielding detailed lithological and wellbore fissure information. Next, electromagnetic wave velocity was calibrated using GPR profile and the depth of the stratigraphic interface determined by borehole data. Utilizing this calibrated velocity, we are able to accurately calculate the depth values of anomalies in GPR interpretation profiles. Subsequently, we compared the preliminary GPR interpretation profile with the borehole televiewer images. After eliminating false anomalies caused by interference, we obtained more reliable location information for detection targets such as fissures, fracture zones, and weak interlayers. The results of fissure detection in the Dazu Rock Carvings indicate that the detection results of ground-penetrating radar are verified and supplemented under the constraints of stratigraphic and well-wall fissure information obtained by logging. This effectively mitigates the influence of multiplicity and false anomalies of GPR detection on interpretation results. GPR with the logging data constraint enhances the accuracy of the fissure detection results, providing novel technical means for the protection and restoration of grotto relics.

Integration of Ground Penetrating Radar and Electrical Resistivity Methods to Investigate Subsurface Features at Kish Archaeological site, Babylon, Iraq

Several geophysical techniques, including magnetic, Electrical Resistivity Imaging (ERI), and Ground Penetrating Radar (GPR), have been used to investigate archaeological features because they are non-destructive and inexpensive compared to traditional methods such as excavation. GPR has been used in the Governorate of Babylon, Kish archeological site to investigate the subsurface ancient features.The GPR survey carried out on three parallel profiles. The length of each profile is 29 m and the spacing between each adjacent profile is 15 meters. The antennas 450 MHz and 750 MHz were used. The data was processed using GPR SliceV7 software. The GPR profiles indicate the presence of buried features, which may be an ancient wall. The wall is buried at a depth range between 0.5 and 3 m. Outside the study area, some part of the wall is visible on the surface. Three Electrical resistivity image (ERI) profiles (1, 2 and 3) were carried out by using wenner array to check the main features detected by GPR method. The ERI results are confirmed the GPR profiles (1, 2, and 3). Both methods indicate the existence of an ancient wall trending NW-SE. The thickness of the wall is approximately 3m and is located at distance 7.5-10.5 meters at profile 1 in the south and at distance 10-13 m in profiles 2 and 3. The depth of the wall ranges from 3–3.5 m. Another deep wall was also detected in the GPR profile 1 and in the three profiles of ERI method. The current work confirmed the importance of using ERI method with GPR method to investigate shallow subsurface features.

GPR and Magnetic Techniques to Locate Ancient Mining Galleries (Linares, Southeast Spain)

Old mining districts have created numerous subsurface cavities, often at shallow depths. The resulting subsidence risk is a major territorial planning problem, especially when these holes are in urban expansion areas. Ground-penetrating radar (GPR) and magnetic techniques can help to detect and to characterise these shallow mining structures based on the strong contrast of electromagnetic and magnetic properties (dielectric constant and magnetic susceptibility) between the rock and the backfill of cavities. In the present study, these techniques were used to locate old mining cavities near the city of Linares, located south of Spain and connected to the area’s old mining district. GPR and magnetometry (total magnetic field and vertical magnetic gradient) were performed on a grid in one of the most important veins in the sector. By comparing both working methods, the vein structure within the granite can be detected. On the one hand, the magnetic prospecting technique (magnetic anomalies) has allowed us to detect when the vein is covered by metallic elements of natural or anthropogenic origin. On the other hand, strong reflections and hyperbolic events associated in GPR profiles confirm the presence of cavities related to old mining operations. Shallow magnetic anomalies not associated to GPR variations are related to the slag present in the study area (detected in the outcrop) or to unexploited vein mineralizations.

GPR survey on underwater archaeological site: A case study at Jenipapo stilt village in the eastern Amazon region, Brazil

In this work we present and discuss the results of a pioneering Ground Penetrating Radar (GPR) survey performed at the underwater archaeological stilt village called Jenipapo in the eastern Amazon region. Jenipapo stilt village is located on the Turiaçu River, Maranhão State, Northeast of Brazil. The region is known as the Maranhense wetland where settlements were built by indigenous populations that inhabited this region during the pre-colonial period (∼AD 1 – 1100). These settlements were built on the margin of rivers and lakes, being supported by wooden pillars, corresponding to prehistoric stilt houses. The objectives of this research are to refine the wooden pillars location at the Turiaçu River bottom and guide the search for underwater archaeological artifacts. A total of twelve GPR profiles were acquired with 270 MHz antennas using a rubber boat. The GPR results indicated clear signal reflection from the river bottom and several diffraction hyperbolas suspended over the river bottom related to the top of the wooden pillars. To support the interpretation of the results, a GPR numerical modeling was performed to simulate the diffraction hyperbolas suspended in the water related to the wooden pillars. The synthetic result presented a good agreement with real data. Finally, the location of several diffraction hyperbolas can guide the divers to find and collect archaeological artifacts around the underwater wooden pillars that are the best evidence of the stilt houses of indigenous settlements found in the eastern Amazon region. Several archaeological artifacts were found during previous work, including ceramic materials. The results of this research show the capability of the GPR method to map an underwater wooden pillar and increase the probability of finding archaeological artifacts on the river bottom, which contributes to improve the knowledge of the Amazonian archaeology in Brazil.