Measurement of the 36Cl deposition flux in central Japan: natural background levels and seasonal variability.

Essential parameters for the applications of (36)Cl as a tracer in groundwater studies include the initial (36)Cl/Cl ratio, at the time of recharge, and/or the natural background deposition flux of (36)Cl in the recharge area. To facilitate the hydrological use of (36)Cl in central Japan, this study aimed to obtain a precise estimate of the long-term average local (36)Cl flux and to characterize its seasonal variability. The (36)Cl in precipitation was continuously monitored in Tsukuba, central Japan over a period of >5 years. The (36)Cl flux showed a clear seasonal variation with an annual peak during the spring, which was attributed to the seasonal variability of tropopause height. The long-term average (36)Cl flux (32±2atoms m(-2)s(-1)), estimated from the measured data, was consistent with the prediction from the (36)Cl latitudinal fallout model scaled using the global mean production rate of 20atoms m(-2)s(-1). The initial (36)Cl/Cl ratio was estimated to be (41±6)×10(-15), which is similar to that of pre-bomb groundwater in the Tsukuba Upland. An observation period covering an 11-year solar cycle would yield more accurate estimates of the values, given the increased (36)Cl flux during the solar minimum.

Estimation of groundwater residence time using the 36Cl bomb pulse.

We propose a methodology for estimating the residence time of groundwater based on bomb-produced (36)Cl. Water samples were collected from 28 springs and 2 flowing wells located around Mt. Fuji, Central Japan. (36)Cl/Cl ratios in the water samples, determined by accelerator mass spectrometry (AMS), were between 43 × 10(-15) and 412 × 10(-15). A reference time series of the above-background (i.e., bomb-derived) (36)Cl concentration was constructed by linearly scaling the background-corrected Dye-3 data according to the estimated total bomb-produced (36)Cl fallout in the Mt. Fuji area. Assuming piston flow transport, estimates of residence time were obtained by comparing the measured bomb-derived (36)Cl concentrations in spring water with the reference curve. The distribution of (36)Cl-based residence times is basically consistent with that of tritium-based estimates calculated from data presented in previous studies, although the estimated residence times differ between the two tracers. This discrepancy may reflect chlorine recycling via vegetation or the relatively small change in fallout rate, approximately since 1975, which would give rise to large uncertainties in (36)Cl-based estimates of recharge for the period, approximately since 1975. Given the estimated ages for groundwater from flowing wells, dating based on a (36)Cl bomb pulse may be more reliable and sensitive for groundwater recharged before 1975, back as far as the mid-1950s.