NOAH-MP V5 run as a single cell model (1D driver)

[YiannisS]

For a previous version of Noah-MP LSM there was a specific model configuration to run the model as a single cell model (forcing for one point location and output for only this point location)

here is the link to this model:

https://ral.ucar.edu/sites/default/files/docs//simple-driver-noahmp-v11tar.gz

Is there such an option for the latest Noah-MP v5 or a workaround to archive this type of model configuration?

 

[Ming Chen]

I am not sure of the update. I will forward your question to our expert.

 

[cenlinhe]

Yes, here is the model code to run Noah-MP offline: GitHub - NCAR/hrldas: HRLDAS (High Resolution Land Data Assimilation System)
The instruction of compiling the model code is here: hrldas/tutorial/Note0_Download_Compile.ipynb at master · NCAR/hrldas
The instruction of running it in a single cell mode is here: hrldas/tutorial/Note1_Single_Point.ipynb at master · NCAR/hrldas

 

[Ming Chen]

@cenlinhe
Thank you so much !

 

[Tomasio]

Hi @cenlinhe
How can I solve this problem (when I run single cell mode):

lsminit: out of range ISLTYP 1 1 -2147483648
Note: The following floating-point exceptions are signalling: IEEE_INVALID_FLAG

 

[cenlinhe]

Did you follow the instruction of single-point run in the above link: hrldas/tutorial/Note1_Single_Point.ipynb at master · NCAR/hrldas
It looks like you did not give a correct soil type (ISLTYP) initial value. Did you successfully generate HRLDAS_setup file? If yes, please double check your soil type (and other variables) in HRLDAS_setup file to ensure they have a reasonable (not -9999 or NaN) value.

 

[thomas1983]

Thanks @cenlinhe for the clarification. I confirmed that the single-point setup itself is correct. Using the same HRLDAS_setup_single_point.nc, forcing files, and simulation period, the model runs successfully to completion in a controlled Docker environment and writes the expected restart file. However, on the HPC system the run fails at initialization with lsminit: out of range ISLTYP, where ISLTYP is read as a fill value. This indicates an environment/compilation issue on the HPC side (likely NetCDF/HDF5 incompatibility or NetCDF-4 support), rather than a problem with the input data or workflow.