diagnosing coupled wrf/wrf-hydro crash

[willrudisill]

Hello WRF Hydro community,

I'm running WRF coupled WRF-Hydro and I'm encountering an issue that I think is on the WRF-Hydro end of things. I'm getting a fatal error after for about ~ 48 model hours. When I do cat rsl.out* | grep FATAL I get the following: " FATAL ERROR: SMCRT fully depleted upon disaggregation...          94". 

At the bottom of this post I've attached my namelist.input and hydro.namelist files for a simulation thats currently running. I'm running WRF v4.1.2 compiled with intel compilers on my university's cluster. I'm using a total of 112 cores across 4 nodes. 

The WRF domain is 1km and apporoximately ~300x300 grid cells with 50 vertical levels. The LBCS are produced using ndown.exe and wrfout files from a previous wrf run. As a test, I ran WRF without being coupled to WRF hydro for the same domain configuration and it ran fine. I've also tested my compile of WRF Hydro successfully with the test cases. 

The hydro domain (again, copied below) has a 250 m resolution (1/4 of the WRF domain) and a 10 second terrain routing model timestep. Right now I'm testing the run with CHANRTSWCRT turned OFF, but SUBRTSWCRT and OVRTSWCRT turned on. 

Does anyone have any suggestions for things that I might check? Adjust the terrain routing timestep? Reduce the resolution of the routing grid? Thanks in advance. 

[Sumana Sarkar]

Dear Willrudisill,

I did faced similar problem too....To resolve  I have run the model with SPATIAL_SOIL = 0. Also I have rechecked the specified parameters in my simulation.The problem got solved...

Did you run test cases successfully before opting for your customized run??

Best regards,

Susana

--
You received this message because you are subscribed to the Google Groups "wrf-hydro_users" group.
To unsubscribe from this group and stop receiving emails from it, send an email to wrf-hydro_use...@ucar.edu.
To view this discussion on the web visit https://groups.google.com/a/ucar.edu/d/msgid/wrf-hydro_users/0d9f79ab-1673-477d-a827-d4ffa339c557o%40ucar.edu.

[willrudisill]

Thanks for the quick reply Susana. I might give that a shot. Ideally I would like to have the spatial soil option enabled. 

Just now I increased the terrain routing timestep from 10 to 30... no crash yet, but we will see. 

[willrudisill]

Actually, I misspoke in my last response. I am running WRF/WRF-Hydro compiled without the spatial soil option.  SPATIAL_SOIL=0 in the setEnvar.sh script... I forgot that this was a compile time option and not a namelist option.