Improved Lumped-Parameter and Numerical Modeling of Unsaturated Water Flow and Stable Water Isotopes.

Characterizing unsaturated water flow in the subsurface is a requirement for understanding effects of droughts on agricultural production or impacts of climate change on groundwater recharge. By employing an improved lumped-parameter model (LPM) approach that mimics variable flow we have interpreted stable water isotope data (δ O and δ H), taken over 3 years at a lysimeter site located in Germany. Lysimeter soil cores were characterized by sandy gravel (Ly1) and clayey sandy silt (Ly2), and both lysimeters were vegetated with maize. Results were compared with numerical simulation of unsaturated flow and stable water isotope transport using HYDRUS-1D. In addition, both approaches were extended by the consideration of preferential flow paths. Application of the extended LPM, and thus varying flow and transport parameters, substantially improved the description of stable water isotope observations in lysimeter seepage water. In general, findings obtained from the extended LPM were in good agreement to numerical modeling results. However, observations were more difficult to describe mathematically for Ly2, where the periodicity of seasonal stable water isotope fluctuation in seepage water was not fully met by numerical modeling. Furthermore, an extra isotopic upshift improved simulations for Ly2, probably controlled by stable water isotope exchange processes between mobile soil water and quasi-immobile water within stagnant zones. Finally, although LPM requires less input data compared with numerical models, both approaches achieve comparable decision-support integrity. The extended LPM approach can thus be a powerful tool for soil and groundwater management approaches.