Comparing Two Different Methods of Preferential Flow Simulation, Using Calibration Constrained Monte Carlo Uncertainty analysis

Friday, 19 December 2014
Mario Schirmer1, Mehdi Ghasemizade2,3 and Dirk Radny2, (1)EAWAG Swiss Federal Institute of Aquatic Science and Technology, Duebendorf, Switzerland, (2)EAWAG Swiss Federal Institute of Aquatic Science and Technology, Dübendorf, Switzerland, (3)University of Neuchâtel, Centre for Hydrogeology and Geothermics, Neuchâtel, Switzerland
Many different methods and approaches have been suggested for simulation of preferential flows. However, most of these methods have been tested in lab scales where boundary conditions and material properties are known and under control. The focus of this study is to compare two different approaches for simulating preferential flows in a weighing lysimeter where the scale of simulation is closer to field scales than simulations done in labs. To do so, we applied dual permeability and spatially distributed heterogeneity as two competitive approaches for simulating slow and rapid flow out of a lysimeter. While the dual permeability approach assumes that there is a structure among soil aggregates and that can be captured as a fraction of the porosity, the other method attributes the existence of preferential flows to heterogeneity distributed within the domain. The two aforementioned approaches were used in order to simulate daily recharge values of a lysimeter. The analysis included a calibration phase, which started from March 2012 until March 2013, and a validation phase which lasted a year following the calibration period. The simulations were performed based on the numerical and 3-D physically based model HydroGeoSphere. The nonlinear uncertainty analysis of the results indicate that they are comparable.