Radon and CO2 as natural tracers to investigate the recharge dynamics of karst aquifers

• Published in: Journal of hydrology (Amsterdam) Vol. 406; no. 3-4; pp. 148 - 157
• Main Authors: SAVOY, Ludovic, SURBECK, Heinz, HUNKELER, Daniel
• Format: Journal Article
• Language: English
• Published: Kidlington Elsevier 06-09-2011
• Subjects: Aquifers; Carbon dioxide; Discharge; Dynamic tests; Dynamics; Earth sciences; Earth, ocean, space; Exact sciences and technology; Floods; Freshwater; Geochemistry; Hydrogeology; Hydrology. Hydrogeology; Isotope geochemistry; Isotope geochemistry. Geochronology; Karst; Mineralogy; Silicates; Soil (material); Water geochemistry; Jura Switzerland; Central Europe; Switzerland; Europe; radon; floods; carbonate rocks; mixing; permeability; ground water; ground-water recharge; Rn-222; aquifers; hydrochemistry; limestone; Epikarst; discharge; soils; base flow; carbon dioxide; models; underground streams; Karst hydrogeology; tracers; concentration; karst; unsaturated zone; contamination; sedimentary rocks; Soil; radioactive isotopes; travel time
• ISSN: 0022-1694; 1879-2707
• DOI: 10.1016/j.jhydrol.2011.05.031

This study investigated the use of radon (222Rn), a radioactive isotope with a half-life of 3.8days, and CO2 as natural tracers to evaluate the recharge dynamics of karst aquifer under varying hydrological conditions. Dissolved 222Rn and carbon dioxide (CO2) were measured continuously in an underground stream of the Milandre test site, Switzerland. Estimated soil water 222Rn activities were higher than baseflow 222Rn activities, indicating elevated 222Rn production in the soil zone compared to limestone, consistent with a 226Ra enrichment in the soil zone compared to limestone. During small flood events, 222Rn activities did not vary while an immediate increase of the CO2 concentration was observed. During medium and large flood events, an immediate CO2 increase and a delayed 222Rn activity increase to up to 4.9Bq/L and 11Bq/L, respectively occurred. The detection of elevated 222Rn activities during medium and large flood events indicate that soil water participates to the flood event. A soil origin of the 222Rn is consistent with its delayed increase compared to discharge reflecting the travel time of 222Rn from the soil to the saturated zone of the system via the epikarst. A three-component mixing model suggested that soil water may contribute 4-6% of the discharge during medium flood events and 25-43% during large flood events. For small flood events, the water must have resided at least 25days below the soil zone to explain the background 222Rn activities, taking into account the half-life of 222Rn (3.8days). In contrast to 222Rn, the CO2 increase occurred simultaneously with the discharge increase. This observation as well as the CO2 increase during small flood events, suggests that the elevated CO2 level is not due to the arrival of soil water as for 222Rn. A possible explanation for the CO2 trend is that baseflow water in the stream has lower CO2 levels due to gas loss compared to water stored in low permeability zones. During flood event, the stored water is more rapidly mobilised than during baseflow with less time for gas loss. The study demonstrates that 222Rn and CO2 provides value information on the dynamics of groundwater recharge of karst aquifer, which can be of high interest when evaluating the vulnerability of such systems to contamination.