Radon signature of CO2 flux constrains the depth of degassing: Furnas volcano (Azores, Portugal) versus Syabru-Bensi (Nepal Himalayas).

Substantial terrestrial gas emissions, such as carbon dioxide (CO2), are associated with active volcanoes and hydrothermal systems. However, while fundamental for the prediction of future activity, it remains difficult so far to determine the depth of the gas sources. Here we show how the combined measurement of CO2 and radon-222 fluxes at the surface constrains the depth of degassing at two hydrothermal systems in geodynamically active contexts: Furnas Lake Fumarolic Field (FLFF, Azores, Portugal) with mantellic and volcano-magmatic CO2, and Syabru-Bensi Hydrothermal System (SBHS, Central Nepal) with metamorphic CO2. At both sites, radon fluxes reach exceptionally high values (> 10 Bq m-2 s-1) systematically associated with large CO2 fluxes (> 10 kg m-2 day-1). The significant radon‒CO2 fluxes correlation is well reproduced by an advective-diffusive model of radon transport, constrained by a thorough characterisation of radon sources. Estimates of degassing depth, 2580 ± 180 m at FLFF and 380 ± 20 m at SBHS, are compatible with known structures of both systems. Our approach demonstrates that radon‒CO2 coupling is a powerful tool to ascertain gas sources and monitor active sites. The exceptionally high radon discharge from FLFF during quiescence (≈ 9 GBq day-1) suggests significant radon output from volcanoes worldwide, potentially affecting atmosphere ionisation and climate.

Substratum influences uptake of radium-226 by plants.

Radium-226, an alpha emitter with half-life 1600 years, is ubiquitous in natural environments. Present in rocks and soils, it is also absorbed by vegetation. The efficiency of 226Ra uptake by plants from the soil is important to assess for the study of heavy metals uptake by plants, monitoring of radioactive pollution, and the biogeochemical cycle of radium in the Critical Zone. Using a thoroughly validated measurement method of effective 226Ra concentration (ECRa) in the laboratory, we compare ECRa values of the plant to that of the closest soil, and we infer the 226Ra soil-to-plant transfer ratio, RSP, for a total of 108 plant samples collected in various locations in France. ECRa values of plants range over five orders of magnitude with mean (min-max) of 1.66 ± 0.03 (0.020-113) Bq kg-1. Inferred RSP values range over four orders of magnitude with mean (min-max) of 0.0188 ± 0.0004 (0.00069-0.37). The mean RSP value of plants in granitic and metamorphic context (0.073 ± 0.002; n = 50) is significantly higher (12 ± 1 times) than that of plants in calcareous and sedimentary context (0.0058 ± 0.0002; n = 58). This difference, which cannot be attributed to a systematic difference in emanation coefficient, is likely due to the competition between calcium and radium. In a given substratum context, the compartments of a given plant species show coherent and decreasing RSP values in the following order (acropetal gradient): roots > bark > branches and stems ≈ leaves. Oak trees (Quercus genus) concentrate 226Ra more than other trees and plants in this set. While this study clearly demonstrates the influence of substratum on the 226Ra uptake by plants in non-contaminated areas, our measurement method appears as a promising practical tool to use for (phyto)remediation and its monitoring in uranium- and radium-contaminated areas.

Hydrogeological control on carbon dioxide input into the atmosphere of the Chauvet-Pont d'Arc cave.

Carbon dioxide (CO2) concentration (CDC) is an essential parameter of underground atmospheres for safety and cave heritage preservation. In the Chauvet cave (South France), a world heritage site hosting unique paintings dated 36,000 years BP, a high-sensitivity monitoring, ongoing since 1997, revealed: 1) two compartments with a spatially uniform CDC, a large volume (A) (40,000 to 80,000 m3) with a mean value of 2.20 ± 0.01% vol. in 2016, and a smaller remote room (B) (2000 m3), with a higher mean value of 3.42 ± 0.01%; 2) large CDC annual variations with peak-to-peak amplitude of 2% and 1.6% in A and B, respectively; 3) long-term changes, with an increase of CDC and of its annual amplitude since 1997, then faster since 2013, reaching a maximum of 4.4% in B in 2017, decreasing afterwards. While a large effect of seasonal ventilation is ruled out, monitoring of seepage at two dripping points indicated that the main control of CDC seasonal reduction was transient infiltration. During periods of water deficit, calculated from surface temperature and rainfall, CDC systematically increased. The carbon isotopic composition of CO2, correlated with water excess, is consistent with a time-varying component of CO2 seeping from above. The CO2 flux, which is the primary driver of CDC in A and B, inferred using box modelling, was found to confirm the relationship between water excess and reduced CO2 flux into A, compatible with a more constant flux into B. A buoyancy-driven horizontal CO2 flow model in the vadose zone, hindered by water infiltration, is proposed. Similarly, pluri-annual and long-term CDC changes can likely be attributed to variations of water excess, but also to increasing vegetation density above the cave. As CDC controls the carbonate geochemistry, an increased variability of CDC raises concern for the preservation of the Chauvet cave paintings.

Radon emanation from human hair.

Bio-indicator of long time exposure to pollutants, human hair is studied in contaminated areas. The number of studies on background environments remains small, and factors impacting human hair radioactivity in contaminated and background areas remain poorly known. Radon-222, a radioactive noble gas of half-life 3.8 days, is the alpha decay daughter of radium-226 in the uranium-238 chain. Radon emission depends on radium concentration (CRa) and probability of decaying radium to liberate radon (i.e., the emanation coefficient E). The radon-222 emanating power (i.e., radon emanation or effective radium-226 concentration, ECRa) is measured in the laboratory from human hair of a cohort of 93 individuals living in uranium non-contaminated areas using a high-sensitivity method based on 371 long accumulation sessions. E of human hair is also determined. ECRa values from human hair are heterogeneous, ranging from 0.059 ±â€¯0.008 to 3.7 ±â€¯0.1 Bq kg-1 (mean: 0.484 ±â€¯0.006 Bq kg-1). We find 2.6 ±â€¯0.1 and 2.5 ±â€¯0.1 times larger values for females than males and for color-treated than natural hair, respectively. By contrast, E is homogeneous (mean: 0.33 ±â€¯0.11; n = 9). Our data suggest a different behavior of accumulation/elimination processes of heavy elements in females and non-negligible radium concentration in hair dye products. Our results demonstrate 226Ra-238U disequilibrium in human hair, indicating secondary radium intake, and that ECRa mainly depends on CRa. Other factors such as age and sampling time are also studied. The impact of factors on ECRa from human hair in uranium non-contaminated areas is ordered as follows: (body site?) > sex > hair dyeing > dietary/drinking habits > natural color > time period > geographical location > age. Any human hair-based study should take into consideration these factors. Our method, cost-effective and easy to implement, may be applied to large numbers of samples for large-scale epidemiological studies, and may also be useful for criminal investigations.

Persistent CO2 emissions and hydrothermal unrest following the 2015 earthquake in Nepal.

Fluid-earthquake interplay, as evidenced by aftershock distributions or earthquake-induced effects on near-surface aquifers, has suggested that earthquakes dynamically affect permeability of the Earth's crust. The connection between the mid-crust and the surface was further supported by instances of carbon dioxide (CO2) emissions associated with seismic activity, so far only observed in magmatic context. Here we report spectacular non-volcanic CO2 emissions and hydrothermal disturbances at the front of the Nepal Himalayas following the deadly 25 April 2015 Gorkha earthquake (moment magnitude Mw = 7.8). The data show unambiguously the appearance, after the earthquake, sometimes with a delay of several months, of CO2 emissions at several sites separated by > 10 kilometres, associated with persistent changes in hydrothermal discharges, including a complete cessation. These observations reveal that Himalayan hydrothermal systems are sensitive to co- and post- seismic deformation, leading to non-stationary release of metamorphic CO2 from active orogens. Possible pre-seismic effects need further confirmation.

Effective radium concentration in topsoils contaminated by lead and zinc smelters.

Trace elements (TE) are indicative of industrial pollution in soils, but geochemical methods are difficult to implement in contaminated sites with large numbers of samples. Therefore, measurement of soil magnetic susceptibility (MS) has been used to map TE pollutions, albeit with contrasted results in some cases. Effective radium concentration (ECRa), product of radium concentration by the emanation factor, can be measured in a cost-effective manner in the laboratory, and could then provide a useful addition. We evaluate this possibility using 186 topsoils sampled over about 783km(2) around two former lead and zinc smelters in Northern France. The ECRa values, obtained from 319 measurements, range from 0.70±0.06 to 12.53±0.49Bq·kg(-1), and are remarkably organized spatially, away from the smelters, in domains corresponding to geographical units. Lead-contaminated soils, with lead concentrations above 100mg·kg(-1) <3km from the smelters, are characterized on average by larger peak ECRa values and larger dispersion. At large scales, away from the smelters, spatial variations of ECRa correlate well with spatial variations of MS, thus suggesting that, at distance larger than 5km, variability of MS contains a significant natural component. Larger ECRa values are correlated with larger fine fraction and, possibly, mercury concentration. While MS is enhanced in the vicinity of the smelters and is associated with the presence of soft ferrimagnetic minerals such as magnetite, it does not correlate systematically with metal concentrations. When multiple industrial and urban sources are present, ECRa mapping, thus, can help in identifying at least part of the natural spatial variability of MS. More generally, this study shows that ECRa mapping provides an independent and reliable assessment of the background spatial structure which underlies the structure of a given contamination. Furthermore, ECRa may provide a novel index to identify soils potentially able to fix leached components.

Effective radium concentration in agricultural versus forest topsoils.

Effective radium-226 activity concentration (ECRa), the radon-222 source term, was measured in the laboratory with 724 topsoil samples collected over a ∼110 km(2) area located ∼20 km south of Paris, France. More than 2100 radon accumulation experiments were performed, with radon concentration measured using scintillation flasks, leading to relative uncertainties on ECRa varying from 10% for ECRa = 2 Bq⋅kg(-1) to less than 6% for ECRa > 5 Bq⋅kg(-1). Small-scale dispersion, studied at one location with 12 samples, and systematically at 100 locations with three topsoils separated by 1 m, was of the order of 7%, demonstrating that a single soil sample is reasonably representative. Agricultural topsoils (n = 540) had an average (arithmetic) ECRa of 8.09 ± 0.11 Bq⋅kg(-1), and a range from 2.80 ± 0.22 to 19.5 ± 1.1 Bq⋅kg(-1), while forest topsoils (n = 184), with an average of 3.21 ± 0.14 Bq⋅kg(-1) and a range from 0.45 ± 0.12 to 9.09 ± 0.55 Bq⋅kg(-1), showed a clear systematic reduction of ECRa when compared with the closest agricultural soil sample. Large-scale organization of ECRa was impressive for agricultural topsoils, with homogeneous domains of several kilometers size, characterized by smooth variations smaller than 10%. These patches emerged despite heavy human remodeling; they are controlled by the main geographical units, but do not necessarily coincide with them. Valleys were characterized by larger dispersion and less organization. This study illustrates how biosphere and anthroposphere modify the soil distribution inherited from geological processes, an important baseline needed for the study of contaminated sites. Furthermore, the observed depletion of forest topsoils suggests an atmospheric radon signature of deforestation.

Optimized measurement of radium-226 concentration in liquid samples with radon-222 emanation.

Measuring radium-226 concentration in liquid samples using radon-222 emanation remains competitive with techniques such as liquid scintillation, alpha or mass spectrometry. Indeed, we show that high-precision can be obtained without air circulation, using an optimal air to liquid volume ratio and moderate heating. Cost-effective and efficient measurement of radon concentration is achieved by scintillation flasks and sufficiently long counting times for signal and background. More than 400 such measurements were performed, including 39 dilution experiments, a successful blind measurement of six reference test solutions, and more than 110 repeated measurements. Under optimal conditions, uncertainties reach 5% for an activity concentration of 100 mBq L(-1) and 10% for 10 mBq L(-1). While the theoretical detection limit predicted by Monte Carlo simulation is around 3 mBq L(-1), a conservative experimental estimate is rather 5 mBq L(-1), corresponding to 0.14 fg g(-1). The method was applied to 47 natural waters, 51 commercial waters, and 17 wine samples, illustrating that it could be an option for liquids that cannot be easily measured by other methods. Counting of scintillation flasks can be done in remote locations in absence of electricity supply, using a solar panel. Thus, this portable method, which has demonstrated sufficient accuracy for numerous natural liquids, could be useful in geological and environmental problems, with the additional benefit that it can be applied in isolated locations and in circumstances when samples cannot be transported.

Stationary and transient thermal states of barometric pumping in the access pit of an underground quarry.

The transition zone between free and underground atmospheres hosts spectacular phenomena, as demonstrated by temperature measurements performed in the 4.6m diameter and 20m deep vertical access pit of an abandoned underground quarry located in Vincennes, near Paris. In summer, a stable stratification of the atmosphere is maintained, with coherent temperature variations associated with atmospheric pressure changes, with a barometric tide S2 larger than 0.1°C peak to peak. When the winter regime of turbulent cold air avalanches is initiated, stratification with pressure induced signals can be restored transiently in the upper part of the pit, while the lower part remains fully mixed and insensitive to pressure variations. The amplitude of the pressure to temperature transfer function increases with frequency below 5×10(-4)Hz, with values at 3×10(-5)Hz varying from 0.1°C·hPa(-1) at the bottom up to 2°C·hPa(-1) towards the top of the pit. These temperature variations are accounted for by cave breathing, which is pressure induced motion of air amplified by the large volume of the quarry. This understanding is supported by a numerical model including advective heat transport, heat diffusion, and heat exchange with the pit walls. Mean lifetime in the pit is of the order of 9 to 13h, and barometric pumping results in an effective ventilation rate of the quarry of the order of 10(-7)s(-1). This study illustrates the important role of barometric pumping in heat and matter transport between atmosphere and lithosphere. The resulting stationary and transient states, revealed in this pit, are probably a general feature of functioning interface systems, and therefore are an important aspect to consider in problems of contaminant transport, or the preservation of precious heritage such as rare ecosystems or painted caves.

Measuring effective radium concentration with large numbers of samples. Part I--experimental method and uncertainties.

Effective radium concentration EC(Ra), product of radium concentration and radon emanation, is the source term for radon release into the pore space of rocks and the environment. To measure EC(Ra), we have conducted, over a period of three years, more than 5500 radon-222 accumulation experiments in the laboratory with scintillation flasks, and about 700 with integrating solid state nuclear track detectors, leading to experimental values of EC(Ra) for more than 1570 rock and soil samples. Through detailed systematic checks and intercomparison between various repeated experiments, the experimental uncertainty has been assessed, and ranges from 30% (1 σ) for EC(Ra) values smaller than 0.2 Bq kg(-1) to about 8-10% for EC(Ra) values larger than 50 Bq kg(-1). The detection limit, defined as the 90% probability for obtaining a non-zero experimental EC(Ra) value at 68% confidence level, depends on the mass of the sample with respect to the volume of the accumulation volume, and typically varies between 0.04 and 0.09 Bq kg(-1). To measure EC(Ra) from large numbers of samples with sufficient accuracy and uncertainty for our purpose, i.e. for the most natural objects encountered in the environment, the accumulation method with scintillation flask emerged as particularly useful and robust. Properties of EC(Ra) and interpretations inferred from this large data set are presented in the companion paper.

Measuring effective radium concentration with large numbers of samples. Part II--general properties and representativity.

Effective radium concentration EC(Ra), product of radium concentration and radon emanation, is the source term for radon release into the pore space of rocks and the environment. Over a period of three years, we performed more than 6000 radon-222 accumulation experiments in the laboratory with scintillation flasks and SSNTDs and we obtained experimental EC(Ra) values from more than 1570 rock and soil samples. With this method, which allowed the measurement of EC(Ra) from large numbers of samples with sufficient accuracy and uncertainty, as detailed in the companion paper, the dependence of the emanation factor on temperature and moisture content is revisited. In addition, with such a large EC(Ra) dataset, dispersion of EC(Ra) can be studied at sample-scale (cm to dm) and at scarp-scale (m to tens of m). Furthermore, we are able to discuss the representativity of obtained EC(Ra) values at field-scale, and to investigate the spatial variations of EC(Ra) over kilometric scales, within geological formations and across formations and faults. This experimental study opens new perspectives in the understanding of radium geochemistry and illustrates the importance of studying the radon source term with large numbers of samples for the modelling of geological and environmental processes, and also for the assessment of the radon health hazard.

Estimating the importance of factors influencing the radon-222 flux from building walls.

Radiation hazard in dwellings is dominated by the contribution of radon-222 released from soil and bedrock, but the contribution of building materials can also be important. Using a simple air mixing model in a 2-story house with an attic and a basement, it is estimated that a significant risk arises when the Wall Radon exhalation Flux (WRF) exceeds 10×10(-3) Bq·m(-2)·s(-1). WRF is studied using a multiphase advection-diffusion 3-layer analytical model with advective flow, possibly induced by a pressure deficit inside the house compared with the outside atmosphere. To first order, in most circumstances, the WRF is proportional to the wall thickness and to the radon source term, the effective radium concentration EC(Ra), which is the product of the radium-226 concentration by the emanation coefficient E. The WRF decreases with increasing material porosity and exhibits a maximum for water saturation of about 50%. For EC(Ra)=10 Bq·kg(-1), in many instances, WRF is larger than 10×10(-3) Bq·m(-2)·s(-1) and, therefore, EC(Ra)=10 Bq·kg(-1) can be considered as the typical limit not to be exceeded by building materials. An upper limit of the WRF is obtained in the purely advective regime, independent of porosity or moisture content, which can thus be used as a robust safety guideline. The sensitivity of WRF to temperature, due to the temperature sensitivity of EC(Ra) or the temperature sensitivity of radon Henry constant can be larger than 5% for the seasonal variation in the presence of slight pressure deficit. The temperature sensitivity of EC(Ra) is the dominant effect, except for moist walls. Temperature and moisture variation effects on the WRF potentially can account for most observed seasonal variations of radon concentration in houses, in addition to seasonal changes of air exchange, suggesting that the contribution of walls should be considered when designing remediation strategies and studied with dedicated experiments.

Measuring effective radium concentration with less than 5 g of rock or soil.

Radon generation in natural systems and building materials is controlled by the effective radium concentration EC(Ra), product of the radium concentration C(Ra) and the emanation factor E. An experimental method is proposed to measure EC(Ra) in the laboratory by radon accumulation experiments using less than 5 g of sample inserted in 125 mL scintillation flasks. Accumulation curves with fine temporal resolution can be obtained, allowing the simultaneous determination of the effective leakage rate. The detection limit, defined as the EC(Ra) value giving a probability larger than 90% for a determination with a one-sigma uncertainty better than 50%, is moderate, varying from 2 to 5 Bq kg(-1) depending on the conditions. Obtained punctual uncertainties on EC(Ra) vary from about 10 to 20% at 10 Bq kg(-1) to less than 3% for EC(Ra) larger than 500 Bq kg(-1). The representativity of small samples to estimate meaningful values at site or system level is, however, a definite limitation of the method, and the sample dispersion needs to be considered carefully in every case. Nevertheless, the value obtained with 5 g or less differs on average by 9 ± 13% from the value given by standard methods using 100 g or more, thus is sufficiently reliable for most applications. When EC(Ra) is sufficiently large, the temperature sensitivity of EC(Ra) can be measured reliably with this method, with obtained mean values ranging from 0.39 ± 0.05% °C(-1) for Compreignac granite, to 2.8 ± 0.2% °C(-1) for La Crouzille pitchblende, both from the centre of France. This method is useful to study dedicated problems, such as the small scale variability of EC(Ra), and in circumstances when only a small amount of sample is available, for example from remote areas or from precious materials such as historical building stones.

Temporal signatures of advective versus diffusive radon transport at a geothermal zone in Central Nepal.

Temporal variation of radon-222 concentration was studied at the Syabru-Bensi hot springs, located on the Main Central Thrust zone in Central Nepal. This site is characterized by several carbon dioxide discharges having maximum fluxes larger than 10 kg m(-2) d(-1). Radon concentration was monitored with autonomous Barasol™ probes between January 2008 and November 2009 in two small natural cavities with high CO(2) concentration and at six locations in the soil: four points having a high flux, and two background reference points. At the reference points, dominated by radon diffusion, radon concentration was stable from January to May, with mean values of 22 ± 6.9 and 37 ± 5.5 kBq m(-3), but was affected by a large increase, of about a factor of 2 and 1.6, respectively, during the monsoon season from June to September. At the points dominated by CO(2) advection, by contrast, radon concentration showed higher mean values 39.0 ± 2.6 to 78 ± 1.4 kBq m(-3), remarkably stable throughout the year with small long-term variation, including a possible modulation of period around 6 months. A significant difference between the diffusion dominated reference points and the advection-dominated points also emerged when studying the diurnal S(1) and semi-diurnal S(2) periodic components. At the advection-dominated points, radon concentration did not exhibit S(1) or S(2) components. At the reference points, however, the S(2) component, associated with barometric tide, could be identified during the dry season, but only when the probe was installed at shallow depth. The S(1) component, associated with thermal and possibly barometric diurnal forcing, was systematically observed, especially during monsoon season. The remarkable short-term and long-term temporal stability of the radon concentration at the advection-dominated points, which suggests a strong pressure source at depth, may be an important asset to detect possible temporal variations associated with the seismic cycle.

Spatiotemporal variation of radon and carbon dioxide concentrations in an underground quarry: coupled processes of natural ventilation, barometric pumping and internal mixing.

Radon-222 and carbon dioxide concentrations have been measured during several years at several points in the atmosphere of an underground limestone quarry located at a depth of 18 m in Vincennes, near Paris, France. Both concentrations showed a seasonal cycle. Radon concentration varied from 1200 to 2000 Bq m(-3) in summer to about 800-1400 Bq m(-3) in winter, indicating winter ventilation rates varying from 0.6 to 2.5 x 10(-6) s(-1). Carbon dioxide concentration varied from 0.9 to 1.0% in summer, to about 0.1-0.3% in winter. Radon concentration can be corrected for natural ventilation using temperature measurements. The obtained model also accounts for the measured seasonal variation of carbon dioxide. After correction, radon concentrations still exhibit significant temporal variation, mostly associated with the variation of atmospheric pressure, with coupling coefficients varying from -7 to -26 Bq m(-3) hPa(-1). This variation can be accounted for using a barometric pumping model, coupled with natural ventilation in winter, and including internal mixing as well. After correction, radon concentrations exhibit residual temporal variation, poorly correlated between different points, with standard deviations varying from 3 to 6%. This study shows that temporal variation of radon concentrations in underground cavities can be understood to a satisfactory level of detail using non-linear and time-dependent modelling. It is important to understand the temporal variation of radon concentrations and the limitations in their modelling to monitor the properties of natural or artificial underground settings, and to be able to assess the existence of new processes, for example associated with the preparatory phases of volcanic eruptions or earthquakes.

Persistence of radon-222 flux during monsoon at a geothermal zone in Nepal.

The Syabru-Bensi hydrothermal zone, Langtang region (Nepal), is characterized by high radon-222 and CO(2) discharge. Seasonal variations of gas fluxes were studied on a reference transect in a newly discovered gas discharge zone. Radon-222 and CO(2) fluxes were measured with the accumulation chamber technique, coupled with the scintillation flask method for radon. In the reference transect, fluxes reach exceptional mean values, as high as 8700+/-1500 gm(-2)d(-1) for CO(2) and 3400+/-100 x 10(-3) Bq m(-2)s(-1) for radon. Gases fluxes were measured in September 2007 during the monsoon and during the dry winter season, in December 2007 to January 2008 and in December 2008 to January 2009. Contrary to expectations, radon and its carrier gas fluxes were similar during both seasons. The integrated flux along this transect was approximately the same for radon, with a small increase of 11+/-4% during the wet season, whereas it was reduced by 38+/-5% during the monsoon for CO(2). In order to account for the persistence of the high gas emissions during monsoon, watering experiments have been performed at selected radon measurement points. After watering, radon flux decreased within 5 min by a factor of 2-7 depending on the point. Subsequently, it returned to its original value, firstly, by an initial partial recovery within 3-4h, followed by a slow relaxation, lasting around 10h and possibly superimposed by diurnal variations. Monsoon, in this part of the Himalayas, proceeds generally by brutal rainfall events separated by two- or three-day lapses. Thus, the recovery ability shown in the watering experiments accounts for the observed long-term persistence of gas discharge. This persistence is an important asset for long-term monitoring, for example to study possible temporal variations associated with stress accumulation and release.

Temporal variations of radon concentration in the saturated soil of Alpine grassland: the role of groundwater flow.

Radon concentration has been monitored from 1995 to 1999 in the soil of the Sur-Frêtes ridge (French Alps), covered with snow from November to April. Measurements were performed at 70 cm depth, with a sampling time of 1 h, at two points: the summit of the ridge, at an altitude of 1792 m, and the bottom of the ridge, at an altitude of 1590 m. On the summit, radon concentration shows a moderate seasonal variation, with a high value from October to April (winter), and a low value from May to September (summer). At the bottom of the ridge, a large and opposite seasonal variation is observed, with a low value in winter and a high value in summer. Fluctuations of the radon concentration seem to be associated with temperature variations, an effect which is largely delusory. Indeed, these variations are actually due to water infiltration. A simplified mixing model is used to show that, at the summit of the ridge, two effects compete in the radon response: a slow infiltration response, rich in radon, with a typical time scale of days, and a fast infiltration of radon-poor rainwater. At the bottom of the ridge, similarly, two groundwater contributions compete: one slow infiltration response, similar to the response seen at the summit, and an additional slower response, with a typical time scale of about a month. This second slower response can be interpreted as the aquifer discharge in response to snow melt. This study shows that, while caution is necessary to properly interpret the various effects, the temporal variations of the radon concentration in soil can be understood reasonably well, and appear to be a sensitive tool to study the subtle interplay of near surface transfer processes of groundwater with different transit times.

Seasonal variations of natural ventilation and radon-222 exhalation in a slightly rising dead-end tunnel.

The concentration activity of radon-222 has been monitored, with some interruptions, from 1997 to 2005 in the end section of a slightly rising, dead-end, 38-m long tunnel located in the Phulchoki hill, near Kathmandu, Nepal. While a high concentration varying from 6 x 10(3) Bq m(-3) to 10 x 10(3) Bq m(-3) is observed from May to September (rainy summer season), the concentration remains at a low level of about 200 Bq m(-3) from October to March (dry winter season). This reduction of radon concentration is associated with natural ventilation of the tunnel, which, contrary to expectations for a rising tunnel, takes place mainly from October to March when the outside air temperature drops below the average tunnel temperature. This interpretation is supported by temperature measurements in the atmosphere of the tunnel, a few meters away from the entrance. The temporal variations of the diurnal amplitude of this temperature indeed follow the ventilation rate deduced from the radon measurements. In the absence of significant ventilation (summer season), the radon exhalation flux at the rock surface into the tunnel atmosphere can be inferred; it exhibits a yearly variation with additional transient reductions associated with heavy rainfall, likely to be due to water infiltration. No effect of atmospheric pressure variations on the radon concentration is observed in this tunnel. This experiment illustrates how small differences in the location and geometry of a tunnel can lead to vastly different behaviours of the radon concentration versus time. This observation has consequences for the estimation of the dose rate and the practicability of radon monitoring for tectonic purposes in underground environments.

Modelling the effect of air exchange on 222Rn and its progeny concentration in a tunnel atmosphere.

The effect of air exchange on the concentration of 222Rn and its progeny in the atmosphere of the Roselend tunnel, in the French Alps, is estimated using a box modelling scheme. In this scheme, the atmosphere is divided into a small number of well mixed zones, separated by flow restricted interfaces, characterized by their exchange rate. A four-box model, representing the three sections of the tunnel present until 2001 and an adjacent inner room, accounts for the spatial variations of the background 222Rn concentration, and for the time structure of transient bursts observed regularly in this tunnel since 1995. A delay of the order of one day, observed during some transient bursts in the inner room with respect to the end of the tunnel, is accounted for if the bursts are assumed to be mainly generated in the end section of the tunnel, and stored temporarily in the inner room via air exchange. The measured radon concentration is reproduced by this model for an air exchange rate of 1.6x10(-6) s-1 between the room and the tunnel, in a context of a global ventilation rate of 10(-5) s-1 in the tunnel. Gradual onset and decay phases, varying from burst to burst, are also suggested. The equilibrium factor of 222Rn with its progeny, measured in 2002 with values varying from 0.60+/-0.05 to 0.78+/-0.06, is interpreted with a five-box model representing the five sections of the tunnel present after 2001. This model indicates that the equilibrium factor does not provide additional constraints on the air exchange rates, but the value of the deposition rate of the unattached short-lived radon progeny can be inferred, with results varying from 0.2 to 6 h-1 in the various sections. This study illustrates the benefits of a simple modelling tool to evaluate the effect of natural ventilation on 222Rn and its progeny concentration in underground cavities, which is important for radioprotection and for a reliable characterization of signatures of hydrogeological or geodynamical processes. Conversely, this study shows that 222Rn and progeny measurements provide a non-invasive method for characterizing natural ventilation conditions in delicate underground cavities, such as painted caves.

Spatial and time variations of radon-222 concentration in the atmosphere of a dead-end horizontal tunnel.

The concentration of radon-222 has been monitored since 1995 in the atmosphere of a 2 m transverse dimension, 128 m long, dead-end horizontal tunnel located in the French Alps, at an altitude of 1600 m. Most of the time, the radon concentration is stable, with an average value ranging from 200 Bq m(-3) near the entrance to about 1000 Bq m(-3) in the most confined section, with an equilibrium factor between radon and its short-lived decay products varying from 0.61 to 0.78. However, radon bursts are repeatedly observed, with amplitudes reaching up to 36 x 10(3) Bq m(-3) and durations varying from one to several weeks, with similar spatial variations along the tunnel as the background concentration. These spatial variations are qualitatively interpreted in terms of natural ventilation. Comparing the radon background concentration with the measured radon exhalation flux at the wall yields an estimate of 8+/-2 x 10(-6) s(-1) (0.03+/-0.007 h(-1)) for the ventilation rate. The hypothesis that the bursts could be due to transient changes in ventilation can be ruled out. Thus, the bursts are the results of transient increased radon exhalation at the walls, that could be due to meteorological effects or possibly combined hydrological and mechanical forcing associated with the water level variations of the nearby Roselend reservoir lake. Such studies are of interest for radiation protection in poorly ventilated underground settings, and, ultimately, for a better understanding of radon exhalation associated with tectonic or volcanic processes.