Coupled Test Case User Guide: "MPI_ABORT " "grid%tsk unreasonable" Error in ./real.exe

[Dehai Liao]

Hi, 

I followed the Coupled WRF | WRF-Hydro Test Case User Guide (WRF v4.6.1, WRF-Hydro v5.4.0). 
I have configured and complied the WPS and WRF-Hydro. When I am running the command "mpirun -np 2 ./real >& real.log", I got the problem as shown in the "real.log":

starting wrf task 0 of 2
starting wrf task 1 of 2
--------------------------------------------------------------------------
MPI_ABORT was invoked on rank 1 in communicator MPI_COMM_WORLD
with errorcode 1.

NOTE: invoking MPI_ABORT causes Open MPI to kill all MPI processes.
You may or may not see output from other processes, depending on
exactly when Open MPI kills them.
--------------------------------------------------------------------------

in the rsl.error.0001:

Assume Noah LSM input
d01 2013-09-09_18:00:00 forcing artificial silty clay loam at 13 points, out of 820
-------------- FATAL CALLED ---------------
grid%tsk unreasonable
-------------------------------------------

I have checked the geog_data_path (WPS_GEOG) and variables. The SKINTEMP variable was not found in the geo_em files, but was found in the met_em files. Other variables like LANDMASK, SOILTEMP, HGT_M and ALBEDO12M were both available. I am using sf_surface_physics = 4, 4.

I have re-run geogrid.exe, ungrib.exe and metgrid.exe but the issue still exists.

Could anyone tell how to examine whether the geographic data was full or not? If not, then how to get the relieable variables and fix it?

Or if there are any other reasons for the issue and the way to solve this problem?

Any help would be appreciated! Thank you!