|Van Genuchten, Martinus|
Submitted to: Vadose Zone Hydrol
Publication Type: Proceedings
Publication Acceptance Date: 7/21/2007
Publication Date: 11/1/2007
Citation: Pachepsky, Y.A., Guber, A.K., Van Genuchten, M.T., Simunek, J., Jacques, D., Nemes, A., Nicholson, T.J., Cady, R.E. 2007. The multiplicity of flow and transport models in unsaturated zone – curse or blessing. Vadose Zone Hydrol. Vol. VIII, ZNS'07:pp 9-17. Interpretive Summary:
Technical Abstract: Many conceptually different models have been developed to simulate flow and transport in vadose zone. For practical purposes, parameters in these models are often estimated from readily available data using pedotransfer functions. Many pedotransfer functions have been developed, and it is not known which one is the most suitable for a specific site. This adds to the model multiplicity. Selection of the best model is always uncertain at best. The objective of this work is to make an argument for using several models rather of looking for the best model. One method of using several conceptually different models has been recently termed ‘model abstraction’. Model abstraction is the methodology for reducing the complexity of a simulation model while maintaining the validity of the simulation results with respect to the question that the simulation is being used to address. Simpler models can be used along with the more complex to improve the reliability and reduce uncertainty of simulations, to make the modeling and its results more explicable and transparent to the end users, and to enable more efficient use of available resources in data collection and computations. An example of the systematic model abstraction will be presented that has been developed for the case of the infiltration in variably saturated soil under natural rainfall conditions when the complex original model gave inexplicable simulation results. Abstracting base model led to more transparent process description with a simpler model and more robust parameter set without the loss in accuracy of simulating soil water fluxes as the key output. Another way to take advantage of the model multiplicity is to combine results obtained from several models using weights reflecting models’ performance. This approach called multimodeling is appropriate for the case of using several pedotransfer functions to simulate flow and transport in unsaturated zone. The similarity in results from different models may require the dimensionality reductions using, for example, the singular value decomposition. An example of multimodeling for the case of the infiltration in variably saturated soil under natural rainfall conditions will be presented to show that that monitoring of the soil water regime in combination with multimodel simulations can be a viable approach to simulating field water flow in the vadose zone. Overall, the model multiplicity presents an opportunity to combine expert and factual knowledge on unsaturated flow and transport in different natural settings. A systematic use of several models can advance understanding monitoring data and improve the use of monitoring resources.