Parameterization guidelines and considerations for hydrologic models
Imparting knowledge of the physical processes of a system to a model and determining a set of parameter values for a hydrologic or water quality model application (i.e., parameterization) are important and difficult tasks. An exponential increase in the literature has been devoted to the use and development of these models over the years. Few articles, however, have been devoted to developing general parameterization guidelines to assist in hydrologic model application, which is the main objective of this article along with discussing a few important parameters and extracting several case studies from the literature. The following guidelines were extracted from reviewing a special collection of 22 articles along with other relevant literature: (1) use site-specific measured or estimated parameter values where possible, (2) focus on the most uncertain and sensitive parameters, (3) minimize the number of optimized parameters, (4) constrain parameter values to within justified ranges, (5) use multiple criteria to help optimize parameter values, (6) use "soft" data to optimize parameters, and (7) use a warm-up period to reduce model dependence on initial condition state variables. A few soil and hydrology related parameters common to many models are briefly described along with a discussion of measurement and estimation methods and parameter sensitivity (curve number, Manning‘s "n", soil bulk density and porosity, soil hydraulic conductivity, soil field capacity and wilting point, and leaf area index). Weather and management inputs are also discussed, as they are critical hydrologic system information that must be imparted to the model. Several case studies from previously reported research illustrate implementation of the parameterization guidelines. This research will help improve model parameterization, resulting in more consistency, better representation of the field or watershed, and a reduced range of parameter value sets resulting in acceptable model simulations.