TRITIUM/HELIUM-3 DATING OF BASEFLOW
Long-term time series of 3H in baseflow provide an indicator of the mean subsurface residence time (MRT) and hence a useful index of vulnerability/sustainability of surface water. Unfortunately, long-term times series are uncommon and short-term time series are relatively insensitive to the MRT. We are investigating the use of 3H/3He dating as a substitute for long-term time series. A gravel aquifer that is discharging into surface water in the Southern Vienna Basin, Austria was chosen as a test site. The aquifer receives the majority of recharge in the form of infiltration from losing sections of rivers. Long-term 3H records for both precipitation and baseflow are available. The preliminary results indicate that the apparent age of the water depends on sampling location and depth below the top of the stream. Apparent 3H/3He ages range from about 1 to 7 years with the oldest ages occurring near or below the water sediment interface at upstream locations near springs and seeps. The results from CFC sampling indicate that groundwater is contaminated with respect to atmospherically-derived CFCs and concentrations in baseflow decrease with distance downstream as a result of re-equilibration with the atmosphere. The observed decrease in CFCs with distance downstream was evaluated using a gas exchange model that suggests an exchange velocity of 2 cm/hr for CFC-12. This exchange velocity was then used to predict the exchange velocity for 3He. The model was then used to predict the observed change in 3He (and hence apparent 3H/3He age) with an excellent agreement with observed values. This result suggests that while gas exchange is an active mechanism for noble gases in streams, it is possible to predict this exchange and hence "correct" 3H/3He ages for this process. Samples collected near the sediment water interface (i.e. away from gas exchange) have apparent 3H/3He ages that range from 4 to 7 years. The long-term 3H record suggests that the MRT in the aquifer is slightly older ranging from about 8 to 15 years, depending on the physical conceptualization of the system. The relatively small difference between the 3H time series and the exchange-corrected 3H/3He age is being further investigated by directly sampling the aquifer along with additional sampling of baseflow.