Estimating groundwater inflow to a shallow, poorly-mixed wetland using environmental tracers

摘要

Four environmental tracers - ~(222)Rn, ~2H, ~(18)O and chloride - were used to assess patterns and magnitudes of surface water - groundwater interactions in a shallow wetland near Mount Gambier, South Australia. Concentrations of these tracers within the wetland are largely determined by the rates of surface water and groundwater inflow and outflow, and by the tracer concentrations in these inflows and outflows. However, unlike ~2H, ~(18)O and chloride, radon activity in precipitation and surface water inflow is zero, and so these inflows do not contribute radon to the lake. Radon is lost by radioactive decay and gas exchange with the atmosphere, as well as surface water and groundwater outflow. The stable isotopes ~2H and ~(18)O are lost through evaporation as well as outflow, while chloride in only lost through lake outflow. This paper describes interpretation of measurements of radon activity within the wetland, which was measured on three occasions between May and October 2006. Measured activities within the surface water display very large spatial variability, the pattern of which is similar on each occasion, suggesting that it is related to the locations of groundwater inflow and mixing processes. The mean groundwater inflow rate has been estimated from the mean radon activity using a mass balance approach. The components of the radon budget are (i) radon contribution from groundwater inflow, (ii) radon diffusive flux from lake bottom sediments (iii) radon loss due to gas exchange, (iv) radon loss due to radioactive decay, (v) radon loss due to groundwater or surface water outflow. Also required to complete the water balance are the surface water inflow rate, direct precipitation on the lake, and evaporation from the lake. The radon diffusive flux has been estimated from measurements of radon production within the sediments and a diffusive transport model, which is calibrated by measurements of radon activity in sealed chambers that can receive radon only from diffusion and lose it only by radioactive decay. Radon loss due to gas exchange is inferred from the loss rate of SF_6, following its injection into isolated areas of the lake, while the rate of radioactive decay is known. The radon activity in groundwater inflow is measured from sampling piezometers surrounding the lake, and the groundwater inflow rate is then determined by mass balance. Estimated groundwater inflow rates vary between 11 and 17 m~3/day, and are most sensitive to the radon activity of groundwater inflow, the gas exchange coefficient, lake area and the accuracy with which the mean radon activity in the wetland can be measured. Importantly, they are relatively insensitive to the surface water inflow rate, which is poorly known.