Potential Improvements for HEC-HMS Automated Parameter Estimation
Abstract
Model independent enhancements and adaptations to the Gauss-Marquardt-Levenberg (GML) method of computer-based parameter estimation are described and demonstrated as potential improvements to existing HEC-HMS automatic calibration capabilities. In contrast to existing HEC-HMS automated parameter estimation capabilities, these methods support global optimization, the ability to simultaneously calibrate multiple subwatershed systems represented within an HEC-HMS model, and they also provide information about individual parameter sensitivities and parameter correlation during and at the end of the calibration process. Moreover, their model independent nature allows one to include into the calibration process (1) state information other than simply stream discharge data, (2) multiple periods, rather than a single time window, of the calibration dataset(s), and (3) the ability to weight data in order to accommodate a prediction specific calibration effort or to accommodate suspect and/or missing observations. The methods are demonstrated by calibrating HEC-HMS models to subwatershed systems in the Goodwin Creek Experimental Watershed.
Document Details
- Document Type
- Technical Report
- Publication Date
- Aug 01, 2006
- Accession Number
- ADA455236
Entities
People
- Brian E. Skahill
Organizations
- Engineer Research and Development Center