No manning sensitivity in Toolbox

[Arend van Harten]

Dear all, 

I have an issue that when performing a sensitivity analysis in the toolbox it gives 0 sensitivity for the various manning coefficients (chn, fpn, ovn). However, when I manually change them in the editor I do see a difference in the daily flow results. So they are sensitive but the toolbox doesn't seem to be able to change them. Does anyone know the cause of this and a potential solution? 

Arend

[Celray James]

what are you testing sensitivity analysis against?

[Arend van Harten]

Daily river flow

Op maandag 24 maart 2025 om 17:45:18 UTC+2 schreef Celray James:

[Celray James]

How many other parameters are there? also number of sample size and parameter ranges could play a role. If you wish, I could check it, export to sptp (using the export section in SPT) and send me

[Arend van Harten]

First i ran it together with 10 other parameters in scenario "calibration_only_Lugoj". But i also ran it in a seperate "manningtest" scenario, gettin the same results of 0.00000 sensitivity. When running a calibration I get the same performance indicatior for several inputs in a very wide calibration range, so I don't know what else to test. If you could have a look that would be great. See the project in this wetransfer: https://we.tl/t-MeKMms50FB

thanks in advance!

Op maandag 24 maart 2025 om 17:58:31 UTC+2 schreef Celray James:

[Celray James]

The manning's parameters are not sensitive for daily flow for your model set-up. I checked using a larger sample size using calibration.cal. How were you changing them in SPE? did you use Hard Calibration Section or you edited Channels database?

C. James

[Arend van Harten]

I edited the chn and fpn the database (it did require a significant change of parameters tho). And changed the ovn in the land management section, this requierd a much smaller change so they sould come out as sensitive. 

Op dinsdag 25 maart 2025 om 19:22:56 UTC+2 schreef Celray James:

[Celray James]

That explains things a little. there seems to be a bug with adoption of parameters prescribed in calibration.cal for the Manning's parameters. I will do further tests and raise this with other developers.

[Arend van Harten]

Okay thank you! Is there anything you recommend me to do? Should I calibrate the manning values manually from the SPE? or is there a way of adding them to the calibration.cal so I can calibrate in the toolbox?

Op dinsdag 25 maart 2025 om 20:46:43 UTC+2 schreef Celray James: