running dflowfm_blackbox examples

[cortesel]

Hi,

I am familiarizing with OpenDA in d3dfm and I am testing the provided examples. For instance, I am running the "simple_waal_calibration_roughness" example. The first simulation runs without issues but soon after I get the following error:

Exit value: 0
Instance computation done
Error running algorithm step.
Error message: BBModelInstance.getExchangeItem(Obs01.waterlevel):
Exchange item 'Obs01.waterlevel' not found
Existing exchange items are:

Error type :RuntimeException
Stacktrace :java.lang.RuntimeException: BBModelInstance.getExchangeItem(Obs01.waterlevel):
Exchange item 'Obs01.waterlevel' not found
Existing exchange items are:

This is likely a basic mistake on my part but I am not able to find a solution. I don't understand if this is related to how the history file stores the variables. I checked and station ids are correct and the variable 'waterlevel' is there in the history file. I would be very greatful to anyone giving me any advise. I get the exact same error with other examples like "simple_waal_kalman". FYI: I am using the 2025.01 version of Deflt3D-FM and openda 3.3.0.

Best,
Luca

[robot144]

Dear Luca,

OpenDA is trying to match each of your measured time series to a model result. The attempt to list all available items from the model returns an empty list. This can happen if there is no history output from the model yet, eg because the Delft3D-FM run crashed. In the directory where the run was supposed to run you can check the status manually, looking at the files in that folder. One option is that the Openda configuration for this experiment does not point to the proper location of your Delft3D-FM installation.

Best regards, Martin

[cortesel]

Dear Martin,

Thank you very much for your reply. Delft3D-FM runs without issues. OpenDA generates a new folder in which it copies all model files and creates the output folder where I can see the outputs (diagnostic file, restart files, map and history output). The .dia file shows that the simulation went trough and all the outputs are there. To run Delft3D-FM OpenDA uses a .bat that points to "run_dimr.bat" in the Delft3D-FM installation directory. The path is correct.

However, I think I found out the issue. OpenDA is looking for "station" within the variable "waterlevel" in the history output file. After examing the netCDF file, I saw that the IDs of the observation points are actually named "station_name". This creates a mismatch that likely causes the empty list. I added the lines: stationDimensionVarName=station</arg and stationIdVarName=station_name in of the dflowfmWrapper.xml. I am not sure this is the best way, but now OpenDA runs.