Posted Tue, 16 Dec 2014 12:35:20 GMT by Sven Fuchs Dr.
I run a 14 zone, steady-state thermal transport model were thermal conductivity (TC of solid) and radiogenic heat production (source, RHP of solid) of all 14 zones are the parameters to be optimized. In the FePest problem settings these are two groups each with 14 zones.

Using the standard options, the FePest run results in the optimization (zonally constant) of one parameter group only (TC for instance). RHP is not affected. If I active the second group (RHP) no optimization is done at all. Neither for TC nor for RHP.

This can be only circumnavigated by setting log transformation and factor change for both groups.
Does anyone know why?[/color]

[color=blue]Does anyone know the reason that pest does not optimize both parameter groups with the standard settings of FePest?[/color]

This problem is also present when optimizing lateral varying values (pilot points). In this case, group 1 are the pilot points from zone 1 for one parameter (TC, for instance), group2 those from zone 2 aso. With standard setting and by consideration of one parameter only, FePest optimize only zone 1 and ignores zone 2 too 14.
Has someone observed similar actions of FePest and has found a solution or can provide an explanation?[/color]

I am not sure if this can be solved similar to the approach described above, but I will give you a notice as soon as FePest has performed this (time-consuming) job.

I appreciate any help you can give.

Best regards,

Posted Mon, 12 Jan 2015 14:37:29 GMT by Carlos Andres Rivera Villarreyes Global Product Specialist - FEFLOW
Dear Sven,

Did you check if the parameter definitions are set as either [color=red][b]none [/b][/color] or [color=red][b]log[/b][/color] (in Default section)? The definition "none" refers that parameter will be adjusted by using its default units (not transformation at all). If a parameter group is set as [color=red][b]"fixed"[/b][/color], this will not be adjusted during the optimization.

There are parameters (e.g. conductivity, transmissivity, etc.) which are easily adjusted by using a log-transformation. [color=blue][i]"Often a proper choice of whether and what parameters should be logarithmically transformed has a pronounced effect on optimisation efficiency; the transformation of some parameters may turn a highly nonlinear problem into a reasonably linear one"[/i] [/color](PEST Manual, 1-8). If we are talking about a coupled process (flow-heat), a high degree of non-linearity is inherent in the problem.

In general, the "standard" (or default) settings need to be reviewed by the modeller. Several of these settings have been taken from the PEST manual, however there may be cases where they are not applicable for all the models.

Best regards,
Posted Wed, 18 Mar 2015 19:08:23 GMT by Lucien Blandenier

I have a similar problem. I simply tested to calibrate the In/outflow on the entire surface of a 1 layer 3D model but no change is obtained. I get some results when a consider only on a part of the model or when I reduce the number of observation point... but I need to calibrate the Inflow on the entire model using all the observation points.

Any idea?


Posted Tue, 24 Mar 2015 20:17:15 GMT by Carlos Andres Rivera Villarreyes Global Product Specialist - FEFLOW
Dear Lucien,

Maybe the recharge parameter is not sensitive to the observation data provided. You can easily demonstrate this hypothesis by means of the Sensitivity tool provided in FePEST.


You must be signed in to post in this forum.