Assessing an efficient hybrid of Monte Carlo technique (GSA-GLUE) in Uncertainty and Sensitivity Analysis of vanGenuchten Soil Moisture Characteristics Curve

Abstract

Studying model uncertainty and identifying the parameter uncertainty in the modeling of water flow through the soil is useful to improve water and soil management. This research aimed to assess the uncertainty of tshe parameters soil water retention curve (SWRC) models using an efficient hybrid of the Monte Carlo technique e.g. generalized likelihood uncertainty estimation (GLUE). GLUE estimates the parameters of vanGenuchten, vanGenuchten-Mualem, and vanGenuchten-Burdine models for four soil classes. Also, to evaluate the relative importance of the model parameters, generalized sensitivity analysis (GSA) was performed. The results of the uncertainty analysis showed that among the studied models, the vanGenuchten-Mualem model with the indices of S = 0.05, T = 0.4, d-factor = 0.25 and, PCI = 100 was considered as the most accurate model with the least uncertainty. Also, the results of GSA were demonstrated that alpha and n parameters were sensitive parameters in the models. Consequently, identifying the uncertainty of the SWRC model structure and its paramet