With the development of increasingly complex hydrologic models that use a wide range of parameters to represent hydrologic processes both in space and time, many challenges arise with respect to simulation and quantification of uncertainty. The goal of this research is to introduce strategies to effectively and efficiently estimate and quantify hydrologic responses. A robust framework for parameter estimation and uncertainty quantification is proposed. The procedure also considers temporal variations over a time-series. Specifically, two issues of traditional estimation schemes and uncertainty quantification methods were addressed: overparameterization and reduction of parameter uncertainty through quantitative information. Parameters were categorized as distributed, inactive, or lumped by combining traditional concepts from identifiability and overparameterization with approaches from sensitivity analyses. This led to decreased dimensionality and thus less required computational demand. The framework takes into account climatic conditions over large scales. As a result, the modeler can investigate parameter uncertainty subbasin-by-subbasin as well as temporal variations. The result is a novel estimation scheme capable of subjectively investigating likelihood to extract quantitative information, improving communication of hydrologic simulation data, and ultimately improving reliability of hydrologic models. The techniques proposed and demonstrated here were programmed within the MATLAB programming environment using the Linux platform. The hydrologic model used in this study was the Variable Infiltration Capacity (VIC) model. The finalized scripting environment will be made available to the modeling community.
Uncertainty; Quantification; Sensitivity; Parameter identifiability; Large scale hydrologic model; Climatic gradient; VIC; GLUE; Bayesian; Pareto;
Level of Degree
First Committee Member (Chair)
Second Committee Member
Third Committee Member
Jia, Lijuan. "Toward improved evaluation of large scale hydrologic models: estimation and quantification of parameter uncertainty." (2015). https://digitalrepository.unm.edu/ce_etds/18