Missing keywords for (3D) coastal models

[SCLaan]

Describe the bug
For 3D models like DCSM-FM we deviate from some default settings. To do so we make use of several keywords that are not available in FMModel of hydrolib-core yet:

Update 5 April 2023:
In Stendert's list below I (@arthurvd) have listed if and how we will support this (awaiting approval by some colleagues):
✔️ : in User Manual, support it
🎓 : Research keyword in 6th gen models, support it
❔ : undecided
❌ : no support in hydrolib-core

in [numerics]:

• Keepstbndonoutflow 🎓
• Baroczlaybed 🎓 Extra added by developer
• Barocponbnd 🎓
• MinTimestepBreak ✔️

in [physics]:

• ITcap 🎓
• Tempmax 🎓
• Tempmin 🎓
• Salimax 🎓
• Salimin 🎓
• Heat_eachstep 🎓
• RhoairRhowater 🎓
• Nudgetimeuni 🎓 Extra added by developer
• IniWithNudge 🎓

in [time]:

• AutoTimestep 🎓
• Startdatetime 🎓
• Stopdatetime 🎓

in [trachytopes]:

• TrtMxR ✔️

in [output]:

• FouFile ✔️
• fouUpdateStep 🎓 Extra added by developer
• WaqInterval ✔️ was in hydrolib-core already
• StatsInterval ✔️ was in hydrolib-core already
• Wrihis_turbulence ✔️
• Wrihis_wind ✔️
• Wrihis_rain ✔️
• Wrihis_infiltration ✔️ Extra added by developer
• Wrihis_temperature ✔️ Extra added by developer
• Wrihis_waves ✔️
• Wrihis_heat_fluxes ✔️
• Wrihis_salinity ✔️ Extra added by developer
• Wrihis_density ✔️
• Wrihis_waterlevel_s1 ✔️ Extra added by developer
• Wrihis_bedlevel ✔️ Extra added by developer
• Wrihis_waterdepth ✔️ Extra added by developer
• Wrihis_velocity_vector ✔️ Extra added by developer
• Wrihis_upward_velocity_component ✔️ Extra added by developer
• Wrihis_sediment ✔️
• Wrihis_constituents ✔️
• Wrihis_zcor ✔️ Extra added by developer
• Wrihis_taucurrent ✔️
• Wrimap_evaporation ✔️
• Wrimap_waterdepth ✔️
• Wrimap_velocity_magnitude ✔️
• Writek_CdWind ❌ Tekplot no official support/not maintained
• Wrimap_CdWind ❌ does not exist!
• Wrimap_tidal_potential ✔️
• Wrimap_sal_potential ✔️ Extra added by developer
• Wrimap_internal_tides_dissipation ✔️ Extra added by developer
• Richardsononoutput ✔️ was in hydrolib-core already

To Reproduce
Steps to reproduce the behavior:

1. Run the lines:

from pathlib import Path
import hydrolib.core.dflowfm as hcdfm

oldPath = r'p:\11208054-004-dcsm-fm\models\3D_DCSM-FM\2013-2017\B05\DCSM-FM_0_5nm.mdu'
mduOld  = hcdfm.FMModel(filepath=oldPath,path_style='unix',recurse=False)

newPath = r'p:\11208054-004-dcsm-fm\models\3D_DCSM-FM\2013-2017\B05_hydrolib\DCSM-FM_0_5nm.mdu'
mduOld.filepath=newPath
mduOld.save(path_style='unix',recurse=False)

2. Start a short simulation (done): p:\11208054-004-dcsm-fm\models\3D_DCSM-FM\2013-2017\B05_hydrolib\
3. Compare the diagnostics output to see which keywords in the used mdu changed:

• p:\11208054-004-dcsm-fm\models\3D_DCSM-FM\2013-2017\B05\DFM_OUTPUT_DCSM-FM_0_5nm\DCSM-FM_0_5nm_0001.dia
• p:\11208054-004-dcsm-fm\models\3D_DCSM-FM\2013-2017\B05_hydrolib\DFM_OUTPUT_DCSM-FM_0_5nm\DCSM-FM_0_5nm_0001.dia

Expected behavior
After reading and writing an mdu there should be no differences in settings (and in the resulting simulation).

Version info (please complete the following information):

• OS: Windows 10

[veenstrajelmer]

I notice at least some of these keywords are not in the D-Flow FM User Manual. >> https://issuetracker.deltares.nl/browse/UNST-6822

The behaviour of silently dropping keywords is not desireable, at least a warning should be given. Given this usecase, I think it is not constructive to raise errors upon not-implemented keywords.