Temporal variability of transit time distributions in response to climatic variability: do stable water isotopes and tritium tell the same tale?

Transit time distributions (TTDs) of water moving through a catchment can be estimated using water isotope data. However, different isotopes are characterized by different information contents, which may affect the estimation of TTDs. Stable isotopes, such as ²H and ¹⁸O can provide insights into the part of TTDs that describes water ages of up to a few years. However, they are “blind” to ages older than that. Radioactive isotopes, such as tritium (³H), on the other hand, have been shown to describe old water, and thus the tails of TTDs much better. Direct comparisons of the different information contents and the resulting differences in catchment TTDs estimated from stable and radioactive isotopes are rare, mostly due to very limited data availability. The objectives of this study are therefore to quantify the differences in TTDs together with their temporal variability and sensitivity to climatic variability in multiple components of the hydrological system estimated from both tritium and stable isotopes using a distributed wise-process based model in the Neckar river basin in Germany. More specifically, we test the hypotheses that (1) stable isotope- and tritium-based estimates of TTDs exhibit significant differences for both young and old water ages, (2) they are characterized by distinct sensitivities to climatic variability, and that (3) combined use of stable isotopes and tritium results in more robust estimates of TTDs. The analysis is carried out based on long term hydrological (1958-2016) and isotope data (1990-2016), using a distributed hydrological model coupled with StorAge Selection (SAS) functions, which is simultaneously calibrated (and evaluated) with respect to multiple variables and hydrological signatures including, amongst others, streamflow, tritium, and stable isotope data.