Water and solute flux from the subsurface drains of macroporous agricultural fields are simulated using two-dimensional single-porosity and dual-porosity models. Field-averaged (i.e. effective) parameters are calibrated from drainage outflow and validated for water flow and the transport of non-reactive solutes applied at discrete locations on the field. Both the single-porosity and the dual-porosity simulations capture the observed trends in the drainage hydrographs, with the dual-porosity model performing slightly better than the single-porosity model. The values of the effective hydrologic parameters, however, were not fully characteristic of macroporous soils. The physical meaning of the effective parameters was further questioned as neither the single-porosity nor the dual-porosity models could simulate the rapid transport of solutes to the subsurface drain. The discrepancies between the simulated and observed solute flux indicate that the actual spatial area contributing to drainage outflow in the field experiment was much greater than the integrated area of the simulated domain over which the effective parameters were calibrated. Supplementary simulations using parameters calibrated from solute flux data (outflow solute concentration multiplied by the outflow water flux) also fail to match both water and solute fluxes. The failure of the simulations is attributable to factors such as non-unique parameters and problems with representing a three-dimensional heterogeneous domain as a two-dimensional homogeneous system, including a misrepresentation of macropore flow paths. This study shows the fallacy of interpreting a hydrograph fit as evidence of the physical meaning of model parameters.
Search
Reviews
[Write a review of this article]
Find related articles from these CiteULike users
