Dear all, I'm doing uncertainty analysis on rainfall runoff model by GLUE method, in the GLUE method I have to decide parameter ranges and their optimal values , on what basis that is decided please help.
Hi Praveen,
One option is to go for physical ranges depending on the physical properties of the parameters. The other option, assuming you have a parameter sampling procedure implemented, is go numerical and do a preliminary assessment of the likelihood response surface (dotty-plots) and define the ranges from it. In both cases just make sure parameter ranges make sense to you.
Hope it helps
Rodrigo
I agree with Dr Rojas, check the physical ranges of each parameter and start the numerical simulation with this initial set, if you find a wide range in one parameter then you can reduce its range and run the GLUE again.
Hi Praveen,
In addition to the previous answers, it is important that you make the parameter ranges wide enough in order to sample the complete parameter space (and thus carry out sufficient model runs). In GLUE the threshold value for the likelihood function will largely determine the uncertainty bounds for the parameters and as such is the most important choice when setting up a GLUE analysis.
Good luck,
Martijn
- Badges
- Science topic
- Similar topics
- Geoscience
- Hydrology