Pre-/Post-Processing scripts

[MuellerSeb]

We want to clean up the repository and eventually create a new toolbox for mHM. For now we focus on pre- and post-processing scripts shipped with mHM. Some of these are outdated and/or not compatible with the current version of mHM. So...

If you have:

• 🙏 wishes
• 💡 ideas
• 🔢 working examples

for pre- / post-processing scirpts, just leave a comment here.

[StephanThober]

Wish:

I would like to have a script that automatically calculates the water balance for a given mHM setup and output files within information on average P, simulated and observation-based runoff ratio, mean soil moisture, average Q, KGE, NSE etc. so that it is possible for users to get a quick overview about the setup at hand and how well mHM is representing it.

[rohinikumar82]

And to add on that - average temp, and avg. PET, avg. ET, avg. GW recharge, avg. soil percolation flux..

[silentin]

I think my evaluation script already does most of this. If you can give me a clear description of what you need, I can probably write the script for it without too much work.
I don't know when I'll get around to it, though.

[StephanThober]

Fantastic, is this already part in the mhm-tools repo in git.ufz.de/mhm/mhm-tools ? I think it would be great to have. It does not need to fulfill all variables, this is something that could be added later on.

[pallavhydro]

Preprocessing

1. basin extractor

Postprocessing

1. Point time series extractor for any point in the domain from mHM/mRM_Fluxes_States.nc.

[pallavhydro]

More Postprocessing wishes

There is quite some analysis that scripts can help after optimisation run in mHM (dds_results.out)

2. A simple Objective function evolution plot showing value of the objective function (y axis) with iteration number (x-axis)

3. In the unfortunate situation when the optimisation run doesn't complete (for some reason), mHM doesn't produce FinalParam.nml namelist of optimised parameters. Would be nice to have a script that can generate mhm_parameters.nml based on user input iteration number from dds_results.out, regardless whether the optimisation was complete or not. ( @rohinikumar82 's idea )

4. Sensitivity plot showing objective function vs parameter. Not sure if its useful to have all parameters as a multiplot or just leave the selection to user. Note 1: Might be nice to check whether the parameter optimisation flag was on or off so that it can be included/ excluded. Note 2: Currently dds_results.out stores the best iteration with optimisation and not the actual iteration and corresponding parameters. Would it make more sense if Sensitivity plot could plot the actual iteration objective function value and the corresponding parameters?

5. Currently mHM only saves dds_results.out and the output corresponding to the last iteration after the optimisation run. If mHM were to store the output of all iterations into discharge_iterations.nc or daily_discharge_iterations.out, then we can have a script for Spaghetti plot of all hydrographs with the optimised one highlighted. Perhaps uncertainty bands can also be added. This would be really cool.