Abstract
An inverse method to estimate the unsaturated hydraulic conductivity in seepage flow from field observations is presented. Considering the water movement in soil significantly affected by the soil temperature, the soil column of interest is assumed to be non-isothermal, and therefore the problem is based on coupled 1D water movement and thermal conduction equations. Since the saturated hydraulic conductivity could be definitely known, the inverse problem associated with the unsaturated hydraulic conductivity is reduced to that of identifying the relative hydraulic conductivity (RHC) from the hydro-geological information available. For functional representation of RHC, the free-form parameterized function is employed in lieu of the conventional fixed-form function. Values of the parameters included in the functions are optimally determined according to a simulation-optimization algorithm. For easy application of the method, a utilitarian observation system with simple instrumentation is specially contrived which implements collection of the hydro-geological data relatively easily in-situ available. Validity of the method developed is examined through its practical application to a real soil column in an upland crop field. The results show that the water movement model provides the forward solutions of high reproducibility, when coupled with thermal conduction model and calibrated through identifying the RHC by use of a free-form function.
