Hi,
In your rsl* files, you have many CFL errors; for e.g.,
Code:
rsl_files/rsl.error.0007.txt:d01 2018-03-01_00:06:00 996 points exceeded v_cfl = 2 in domain d01 at time 2018-03-01_00:06:00 hours
rsl_files/rsl.error.0007.txt:d01 2018-03-01_00:06:00 Max W: 249 151 17 W: 13.59 w-cfl: 2.98 dETA: 0.05
rsl_files/rsl.error.0007.txt:d01 2018-03-01_00:06:00 999 points exceeded v_cfl = 2 in domain d01 at time 2018-03-01_00:06:00 hours
rsl_files/rsl.error.0007.txt:d01 2018-03-01_00:06:00 Max W: 249 151 17 W: 13.15 w-cfl: 2.99 dETA: 0.05
rsl_files/rsl.error.0007.txt:d01 2018-03-01_00:06:40 1016 points exceeded v_cfl = 2 in domain d01 at time 2018-03-01_00:06:40 hours
This means the model has become unstable. Take a look at
this FAQ that discusses cfl errors and potential ways to overcome them.
I also have a few other suggestions for your domain and simulation.
1) You probably should increase your number of vertical levels. In newer versions of WRF, the default for the e_vert namelist parameter has been increased to 45, and I would recommend setting yours to that, as well.
2) I see you are only using 10 processors. Although this may be enough, if possible, you likely will want to use more than that to help you run the simulation faster.
This FAQ discusses choosing an appropriate number of processors.
3) Although it may not be a problem, since you are using 10 processors, the processor decomposition becomes 2x5 (the closest prime numbers). If you're going to increase the number of processors, pay attention to the distribution to ensure it's a bit more squared (though it doesn't have to be perfect).
