CSDMS 2014 annual meeting poster Mary Hill

From CSDMS
Presentation provided during CSDMS annual meeting 2014

Exploring how parameter importance to prediction changes in parameter space

Mary Hill, USGS Boulder Colorado, United States. mchill@usgs.gov
Olda Rakovec, UFZ Liepzig , Germany. oldrich.rakovec@ufz.de
Martyn Clark, NCAR Boulder Colorado, United States. mclark@ucar.edu

Abstract:

This talk presents a novel hybrid local-global method that measures how model parameter importance is distributed as parameter values change. DELSA (Distributed Evaluation of Local Sensitivity Analysis) is demonstrated using rainfall-runoff models constricted using FUSE, and results are compared to the Sobol’ global sensitivity analysis method. Insights from DELSA can be combined with field data to identify the most relevant parts of parameter space to focus data collection and model development.

Much of what will be discussed is described in:

Rakovec, O., M. C. Hill, M. P. Clark, A. H. Weerts, A. J. Teuling, and R. Uijlenhoet (2014), Distributed Evaluation of Local Sensitivity Analysis (DELSA), with application to hydrologic models, Water Resources Research, 50, doi:10.1002/2013WR014063.
Image.png
Comparing Sobol' to the less computationally demanding DELSA method.

* Please acknowledge the original contributors when you are using this material. If there are any copyright issues, please let us know and we will respond as soon as possible.