Modelling of water flow through typical Bavarian soils: 2. Environmental deuterium transport

Abstract Seven experimental lysimeters were filled with different soil materials and used to study water flow and tracer transport through the unsaturated zone under natural atmospheric conditions during an eight year period (1984–1991). The study of soil characteristics and water flow was presented in Part 1. The present paper focuses on environmental deuterium transport. The concentrations of environmental deuterium were measured in both precipitation and the water outflowing from the lysimeters. Two model approaches were used to describe tracer transport. The numerical solution of water flow and two-phase mass transport equations, referred to here as the variable flow dispersion model (VFDM), can exactly simulate spatial and temporal distribution of flow and transport parameters. The fitting parameters of VFDM are the soil characteristics, the longitudinal dispersivity, the fraction of immobile water and the transfer constant between mobile and immobile water. However, the application of the VFDM for detailed transport simulations requires an adequate database, which is typically not available. That was the reason to introduce a second model, the lumped-parameter steady-state dispersion model (DM). This model assumes dispersive distribution of the transit time of a tracer particle between input and output from the system, has two fitting parameters: the apparent dispersion constant and the mean transit time of the tracer, and requires only input and output concentrations of tracer as a function of time. Both models were calibrated to the experimental data with similar accuracy. Using the VFDM it was possible to fit output concentrations with the same accuracy assuming that the entire volume of water takes part in motion. The results obtained with the DM demonstrated that this simple steady-state approach can be applied successfully to describe tracer transport under variable flow conditions. Both models have shown that, under applied experimental conditions, it is practically impossible to discover the existence of immobile water in the system.