Search

The goal of data classification is to organize them into relevant groups using algorithms. In this study, two clustering algorithms are applied to classify a set of geophysical measurements performed around the Lascaux Cave (Dordogne, France). Based on a non-destructive geophysical method, electrical resistivity tomography (ERT), the data (resistivity values) are supposed to characterize the rock mass around the cave. The rock mass must be divided into an optimal number of homogeneous domains with specific thermal properties to integrate them in the future thermo-aeraulic simulations of the cave. Since the data are georeferenced, a given resistivity cluster corresponds to a specific spatial domain of the rock mass. This study aims to compare two different clustering methods, the Hierarchical Agglomerative Clustering (HAC) and the K-means methods, on the resistivity data set. Thus, the objective of this study is to determine which of the two methods leads to a partition of the massif with an optimal number of classes, allowing us to find the geological structures partially known thanks to previous studies. The results of these methods are analyzed in light of two different indices, the Silhouette Index (SI) and the Coefficient of Variation (CV). Based on these indices alone, K-means might seem to be the best algorithm, but a fine analysis of each cluster shows that the HAC method gives better results. Indeed, only the HAC method highlights areas already known to be either waterlogged or clayey. Our partial geological knowledge, which is of paramount importance in such a study, also supports the choice of the HAC method. [ABSTRACT FROM AUTHOR]

The Lascaux cave, located in southwest France, is one of the most important prehistoric cave in the world that shows Paleolithic paintings. This study aims to characterize the structure of the weathered epikarst setting located above the cave using Time-Lapse Electrical Resistivity Tomography (ERT) combined with local hydrogeological and climatic environmental data. Twenty ERT profiles were carried out for two years and helped us to record the seasonal and spatial variations of the electrical resistivity of the hydraulic upstream area of the Lascaux cave. The 20 interpreted resistivity models were merged into a single synthetic model using a multidimensional statistical method (Hierarchical Agglomerative Clustering). The individual blocks from the synthetic model associated with a similar resistivity variability were gathered into 7 clusters. We combined the resistivity temporal variations with climatic and hydrogeological data to propose a geo-electrical model that relates to a conceptual geological model. We provide a geological interpretation for each cluster regarding epikarst features. The superficial clusters (no 1 & 2) are linked to effective rainfall and trees, probably a fractured limestone. Another two clusters (no 6 & 7) are linked to detrital formations (sand and clay respectively). The cluster 3 may correspond to a marly limestone that forms a non-permeable horizon. Finally, the electrical behavior of the last two clusters (no 4 & 5) is correlated with the variation of flow rate; they may be a privileged feed zone of the flow in the cave. [ABSTRACT FROM AUTHOR]

This article combining geophysics and archaeology aims to provide a more comprehensive characterization of the La Ruchelle valley located in the south of the Celtic site of La Peyrouse (Saint‐Félix‐de‐Villadeix, Dordogne, France) (occupied between the 3rd century BC and the 2nd century AD) through geological prospection (core sampling, geotechnical prospection and mechanical prospection) and near‐surface geophysics (electrical resistivity tomography). On this site, other studies have been carried out since the discovery of the site, such as magnetic prospecting and a hydrogeological study. Geophysical data coupled with geological prospection were used to understand the geology of the valley: (1) The bottom of the valley is filled with colluvium with a thickness of 6 m maximum; (2) part of the colluvium filling characterize by a very low resistivity is indeed archaeological remains; (3) three very distinct geological horizons have been detected—two of these horizons are characteristic of Campanian limestone (C6d and C6e), and the last consists of colluvium with a filling between 2 and 6 m in the centre of the valley, which is particularly thick. This last very thick formation is in fact filled with archaeological and prehistoric remains. [ABSTRACT FROM AUTHOR]

The Lascaux cave, located in southwest France, is one of the most important prehistoric caves in the world that show Paleolithic paintings. This study aims to simulate the 3D geological environment of the cave by geostatistical modeling based on 2D Electrical Resistivity Tomography (ERT) data. Seventeen ERT profiles were carried out resulting in 64,195 geolocated resistivity values, covering about 14,500 m 2 of the Lascaux site. From this dataset, 3D resistivity geo-model has been built by geostatistical method. We used geotechnical data from 36 boreholes and cone penetrometer tests to validate the geo-model. By modeling from 2D data, we can establish a good estimation of the 3D geological environment of the Lascaux site. Our results show that the site can be generally separated in two parts: the limestone promontory in the center and the detrital formations located in both east and west sides. There is a high contrast between resistivity of the two formations: 45 Ω·m (detrital formation) to 556 Ω·m (limestone) with a rough boundary. In the center, the limestone could also be separate into two parts: a superficial layer of weathered limestone (150 Ω·m) and a deeper layer with compact limestone (556 Ω·m). [ABSTRACT FROM AUTHOR]

Inland aquifers are connected to the ocean through permeable coastal sediments. The mixing zone between sea water and fresh-groundwater forms a so-called subterranean estuary. The dynamics and the shape of this interface influence fluxes of fresh water to the coastal zone. On tidal sandy beaches, tide and waves actions lead to the formation of an intertidal saline circulation cell that complements the structure of the subterranean estuary between sea water and fresh water. The upper beach corresponds to the recharge zone, where large volumes sea water penetrate the sediment at high tide, whereas the lower beach corresponds to the discharge zone of older pore water that remains in the sand for several tidal cycles. The objective of our study was to characterize the shape and the evolution of this saline cell in a high-energy macro-tidal beach, the Truc Vert beach (SW France). In order to assess the distribution of saline, brackish and fresh water in this high-energy beach aquifer, we conducted electrical resistivity tomography (ERT) measurements from the lower beach face to the sand dune in winter 2012–2013. We also measured water table elevation and salinity in three piezometers situated at the base of the dune and in pore waters at low tide on tidal cross-shore transects. Results show that the intertidal saline plume can be identified and localized with the ERT method. The fresh-saline water interface is almost vertical in the upper beach. The saline circulation cell occupies the upper part of the tidal beach over a thickness of 5 m deep. Below this saline cell, a brackish water flows to the ocean: about 30% of fresh water was found in pore water in the lower part of the beach. Shifting of the vertical front between saline and fresh waters of about 2–5 m occurred during the spring–neap tidal cycle. This migration is small relative to the beach size, suggesting that the transition zone is relatively stable. The variations in salinity of the circulation cell suggest that the position of the saline plume is controlled by the spring tide level. Our results differ slightly from studies based on modelling and piezometric measurement made in less exposed micro-tidal sandy beaches. Our data would allow to better constrain numerical models that would enable to quantify water residence time and solute fluxes in subterranean estuaries of high-energy beaches. [ABSTRACT FROM AUTHOR]

The Lascaux Cave, located in the south-eastern part of the department of the Dordogne (24, France), is considered to be one of the most important prehistoric caves in the world. The scope of this study is the protection and conservation of the Lascaux paintings from a hydrogeological and climatic environmental standpoint. Geophysical methods enable us to monitor the environment of the decorated cave in a non-invasive way. Electrical resistivity tomography (ERT) using a pole--dipole array, together with analysis of the local effective rainfall (groundwater recharge) and the flow in the cave, helps us to identify an area where upstream underground water is probably stored e.g. a recharge zone. There is a relationship between resistivity change in this zone and the underground flow measured in the cave, but with a time lag. Thus systematic electrical surveys following rainy periods could be used to predict the beginning of the underground flow and in so doing provide valuable advice for the preservation of the Lascaux Cave. [ABSTRACT FROM AUTHOR]

Abstract: In France, the monitoring of landfill cover after closure of the site is a local problem, since its tightness must be ensured over time. Leaks in the cover are a problem, as they allow water to infiltrate the stored waste. In order to locate such leaks, electrical resistivity tomography was used on an experimental site in which defects had been intentionally made in the cover. Repeated measurements taken on this site showed that the weather conditions preceding the measurements need to be taken into account, as they affect the water content in the cover material. They also showed that there are optimal weather conditions for detecting defects in the cover. A statistical analysis carried out on the electrical resistivity results for all surveys and cover material samples showed that the material was heterogeneous; this variability was mainly due to a difference in particle size (fines content) and in compaction. This study has shown the capacity of electrical resistivity tomography to detect defects and heterogeneity in the cover material, indicating that it is a good means of monitoring the quality of landfill cover both when it is put in place and subsequently. [Copyright &y& Elsevier]

Non-destructive techniques (NDT) are often seen as a practical and efficient way to assess the material and structural condition of existing reinforced concrete structures. However, assessment cannot be reduced to measurement and interpretation, and asset managers and structural engineers often need a quantitative assessment. A combination of several techniques can offer precious help. This article intends to show what kind of improvement can be expected from such a combination. Examples are taken from a series of on-site case studies and laboratory experiments. The focus is on the assessment of water content and concrete porosity as these material properties are key factors regarding their present condition as well as their durability in an aggressive marine environment. Material variability, uncertainties on measurements and model error condition the accuracy of the parameter estimation. Their relative contribution to the quality of material assessment is analysed. [ABSTRACT FROM PUBLISHER]

The study of karst features under a detrital cover is difficult to obtain using punctual traditional tools. This paper presents a 3D geostatistical modeling of ERT (Electrical Resistivity Tomography) data to describe undercover karst features. A case study was carried out on the karstic site of the prehistoric decorated Lascaux cave (France). Geophysical measurements were used to define the limit between outcropping Coniacian or Santonian limestones (southwest) and clayey sands (northeast), with a main orientation of N140°. A geometrical description of the scarp was also constructed; pinnacles and notches were found under the clayey sand detrital formation. By combining 3D ERT with geomorphological and geological observations, the geometry of the stratigraphic limit between the Coniacian and Santonian limestone could be determined. This stratigraphic limit separates two domains, one is a potential aquifer, and acts as a feeder for the intermittent spring at the cave entrance, while the other is less permeable, resulting in a permeability contrast with the later rock type. All these observations help define the geological cave environment and ensure better protection for the paintings. • Undercover karst features are found with 3D ERT geostatistical model. • Geomorphological features found comprises shelf, pinnacles and scarps. • The geological limits are redrawn around the ornamented Lascaux cave. • The top of the Lascaux cave develops within Santonian limestone. • The rest of the cave develops in Coniacian, e. g. the Hall of the Bulls. [ABSTRACT FROM AUTHOR]

Water migration within the rock mass during a fire has not yet been studied. Here, an innovative high-resolution micro-ERT method is developed at a centimetric scale to measure this phenomenon. The method is two-step based: first, a correction of the electrode effect, and second, the correction of resistivity for rapid variations in temperature with high values. Water migration is successfully described within the wall of an underground limestone quarry with three phases: from fire ignition to 36 min, the surface dries to a 3-cm depth and a higher saturated layer forms deeper than 0.10 m; from 36 min to 46 min (fire supply stops), the rock dries out and water migrates from deeper parts of the rock wall, deeper than 20 cm, towards the surface; finally, after fire extinction, the water migrates through a readjustment phase until it returns to its initial state. [ABSTRACT FROM AUTHOR]

Because of the great heterogeneity of the limestone fissures, conduits, and reservoirs in karst areas, the understanding of the infiltration in caves and karst terrains is complicated. For example, the time delays observed between rainfall and the various flows into caves, such as shaft flow and seepage flow, exhibit significant variability. To better understand the infiltration above a typical shallow cave, Villars Cave, South-west France, we used both a very long monitoring (>20 years) isotopic data set of rainfall and dripping rates at two different cave gallery levels (14 m and 36 m below the surface, respectively) and a conceptual model (KarstFor) that mix water flows and isotopes. In order to determine the most appropriate reservoir and flow settings, we used one time-lapse electrical resistivity tomography (ERT) image made above the cave, which were incorporated into the KarstFor model. The modeled variations in storage water and infiltration flows are in a good agreement with the observed dripping discharge rates, indicating seasonal variability and various time delays between water excess ((Rainfall-Evapotranspiration) > 0) and drip rates in the different cave levels. The fact that the model strikingly simulates the small, but constant, isotopic difference of drip water δ18O (δ18O d) values between the upper and the lower galleries of Villars Cave, not only reinforces reliability in the model itself but also reveals the impact of multiple infiltration routes and karst reservoir dynamics on the δ18O d. On an inter-annual scale, the observed δ18O d values follow the variations of the water excess and of the cave temperature, which, in the period analyzed, show a marked trend, possibly induced by the climate variability patterns above the cave. Overall, these results are not only important to better understand the infiltration in karstic zones, but also to better interpret speleothem isotopic records widely used in paleo-climate studies. [ABSTRACT FROM AUTHOR]

ABSTRACT: In this paper, we present the uncertainty analysis of the 2D electrical tomography inverse problem using model reduction and performing the sampling via an explorative member of the Particle Swarm Optimization family, called the Regressive‐Regressive Particle Swarm Optimization. The procedure begins with a local inversion to find a good resistivity model located in the nonlinear equivalence region of the set of plausible solutions. The dimension of this geophysical model is then reduced using spectral decomposition, and the uncertainty space is explored via Particle Swarm Optimization. Using this approach, we show that it is possible to sample the uncertainty space of the electrical tomography inverse problem. We illustrate this methodology with the application to a synthetic and a real dataset coming from a karstic geological set‐up. By computing the uncertainty of the inverse solution, it is possible to perform the segmentation of the resistivity images issued from inversion. This segmentation is based on the set of equivalent models that have been sampled, and makes it possible to answer geophysical questions in a probabilistic way, performing risk analysis. [ABSTRACT FROM AUTHOR]

ABSTRACT In this paper, we present the uncertainty analysis of the 2D electrical tomography inverse problem using model reduction and performing the sampling via an explorative member of the Particle Swarm Optimization family, called the Regressive-Regressive Particle Swarm Optimization. The procedure begins with a local inversion to find a good resistivity model located in the nonlinear equivalence region of the set of plausible solutions. The dimension of this geophysical model is then reduced using spectral decomposition, and the uncertainty space is explored via Particle Swarm Optimization. Using this approach, we show that it is possible to sample the uncertainty space of the electrical tomography inverse problem. We illustrate this methodology with the application to a synthetic and a real dataset coming from a karstic geological set-up. By computing the uncertainty of the inverse solution, it is possible to perform the segmentation of the resistivity images issued from inversion. This segmentation is based on the set of equivalent models that have been sampled, and makes it possible to answer geophysical questions in a probabilistic way, performing risk analysis. [ABSTRACT FROM AUTHOR]

The Chauvet-Pont d'Arc Cave (Ardèche, France), famous for its remarkable rock art, also contains unique thermal-alterations such as rock spalling and color changes on the walls. These alterations resulted from intense fires that have not been observed in the other decorated caves thus far discovered. The functions of these unusual fires challenge archaeologists. To characterize these combustions, we used a numerical tool, previously validated with experimental data, to study the thermo-alterations in the Megaceros Gallery. This unprecedented approach in cave art research enabled us to assess the wood quantities and locations of the hearths responsible for the thermo-alterations. We report here that at least ten fires took place in the Megaceros Gallery while burning more than 170 kg of wood. Both simulation and in situ observations suggest that the branches were arranged in a tepee shape and purposefully positioned, some distance from the walls. This method therefore enables further analysis of the functions of these fires. [ABSTRACT FROM AUTHOR]

• The δ18O p (rainfall δ18O) d-excess, and moisture sources reveal decadal trends in SW-France. • Seasonal and interannual changes of δ18O p not only respond to atmospheric temperature but also significantly to moisture sources. • The impact of moisture sources and temperature on δ18O p were quantitatively evaluated. In the mid-high latitude region, variations of stable isotopic compositions of atmospheric precipitation (δ18O p and δD p) were commonly regarded as reflecting the "temperature effect". However, some studies have indicated that changes in moisture sources are important controlling factors for δ18O p. To clarify whether there are connections between δ18O p and variations of moisture sources in Southwest France (SW-France), whose implications for speleothem δ18O are of great importance, we have used among the longest isotopic time-series from SW-France (Le Mas and Villars stations) and a 5 days' reconstruction of air mass history during the 1997–2016 A.D period based on the HYSPLIT tracking model. We found the percentage of initial moisture sources (PIMS) as important factors controlling the oxygen isotope composition of precipitation in SW-France, whether monthly or inter-annual timescale was considered. Additionally, we observed that the δ18O p preserved the signal of local temperature, supporting the "temperature effect", while no evidence for its "amount effect" has been observed. These quantified links between PIMS/local-temperature and δ18O p appear useful references to understand the link between stable oxygen isotopes and climate parameters. Our long-term monitoring of δ18O p , d-excess, and moisture sources reveals decadal trends, highlighting a tight coupling in hydrologic systems and relatively fast changes on rainfall sources controlled by atmospheric circulations in SW-France. [ABSTRACT FROM AUTHOR]

In 1994, three speleologists discovered the Chauvet–Pont d'Arc Cave, which contains singular thermal marks on walls deep in the cavity. These alterations arose from intense fires, and understanding their characteristics would help archaeologists suggest hypotheses about the function of such activities. In this context, three confined fires were conducted in a former underground quarry to reproduce thermo-alterations similar to those in the Chauvet–Pont d'Arc Cave and extract experimental data. Each fire involved approximately 135 kg of wood, which was continuously supplied by firemen for safety reasons (> 500°C) and burnt in the shape of a tepee 80 cm in diameter for 50 min. This paper presents the validation of a numerical model on this experimentation. The modelling requires coupling between the combustion and wall impact simulations. Thus, a link between the combustion code FireFOAM and the thermo-mechanical code Cast3m was created with Python scripts. The results from the simulation agree with the measurements and the observations. More specifically, the analysis is based on the temperatures, gas and particle concentrations, gas velocities, soot deposition, colour changes at the walls and areas likely to spall. These data were collected from thirty-seven measuring points covering the whole quarry. This validated tool will provide information about the features of the fires that occurred within the Chauvet–Pont d'Arc Cave. [ABSTRACT FROM AUTHOR]

Abstract A numerical simulation of a wood fire in a natural gallery is compared with experimental data. Temperatures, velocities, gases and particles concentrations as well as soot deposits are the measured quantities of the experiment. We expose in this article the comparison with a simulation performed by the open-source solver FireFOAM. In order to improve the accuracy of the modeling, the initial numerical tool has been modified. A boundary condition for temperature based on energy conservation has been added to better estimate thermal losses at the walls. A hypothesis regarding vertical velocity in a confined geometry containing a localized fire is also made. By assuming that the velocity is exclusively horizontal, the error between simulation and experiment is reduced. With the aim of comparing the temperature measurements with the simulation accurately, the thermocouple model commonly used in FireFOAM has been changed. The limitations of the previous models are also discussed. Then, two combustion side issues are presented. First, the model of Beresnev-Chernyak computes soot deposit on walls. Second, the fractional effective doses (FED) for toxicity, radiation and heat hazards have been implemented to evaluate the safety of the environment during the fire. Highlights: • A localized wood fire was made within an instrumented confined gallery. • The experimental data are compared with a FireFOAM simulation. • An accurate thermocouple correction added to the tool allows this comparison. • A thermal boundary condition and a hazard modeling are implemented. • Enforcing the velocity to be only horizontal yields consistent results. [ABSTRACT FROM AUTHOR]

Numerous fire marks occur on the walls of the Chauvet-Pont d’Arc cave. Dating indicated that some of the fires were contemporary to the Aurignacian. Violent thermal shocks were observed in surprisingly narrow areas of the cave. This raises numerous archaeological questions about the function of the fires; the answers depend on the location of the hearths, and the intensity of the fires. Numerical simulation was used here to provide information about the behaviour of fires in such confined spaces. An underground non-archaeological site, in a limestone quarry, was equipped to monitor fires in an environment similar to that of the Megaceros gallery of the Chauvet-Pont d’Arc cave. The fire and the movement of heat and smoke in the quarry were simulated by the open source code “Fire Dynamics Simulator (FDS)”. Results were validated on wall temperatures recorded behind and above the fire. The thermo-mechanical impact of the fire on the rock was simulated with CAST3M software, providing the most probable zones for limestone spalling due to thermal gradients. The validated approach will, in a forthcoming study, be applied to the Chauvet-Pont d’Arc cave, in which coupled simulations in the air and in the rock should indicate the location of the hearths and the intensity of the fires that generated the marks. [ABSTRACT FROM AUTHOR]

A deep sequence of Last Glacial peats and slope deposits at the foot of the Lascaux hill was investigated using a large array of methods (geology, geophysics, 14 C and OSL dating, palynology, macro-remain analysis, entomology). The peaty levels, OSL dated to between ~ 73 and 60 ka, are found within a doline developed in a fluvial terrace of the Vézère River. These gyttja and chalky gyttja deposits covered by fen peat yielded abundant beetle and plant (seeds, pollen) remains. Palaeoecological reconstructions suggest a cool steppe with willow stands compatible with an interstadial of marine isotopic stage (MIS) 4, probably Ognon II (GI-19.2). Aquatic and hygrophilous plants (sedges), which reflect local vegetation, and their dependent insects are equally well represented in the record. The peats are covered by a sequence of slope deposits: a coarse-grained lower unit deposited during late MIS 4, overlain by a heavily decalcified sandy unit with a boreal soil complex correlated with early MIS 3, and finally, a coarse-grained, calcareous unit formed during late MIS 3 and MIS 2 in connection with slope dynamics typical of periglacial environments. The palaeosols yielded a small Middle Palaeolithic lithic assemblage. [ABSTRACT FROM AUTHOR]