search

ufs-community / ccpp-physics

UFS fork for CCPP
Other
3 stars 32 forks source link

Garbage values for dynamics tendencies of the x and y (u and v) winds when doing a restart run #175

Open ericaligo-NOAA opened 4 months ago

ericaligo-NOAA commented 4 months ago

Description

After following the instructions at the link below for writing out dynamics and physics tendencies, all of the tendencies had reasonable values until a restart run was made. The first output time from the restart run and the subsequent history files contained garbage values for the x and y (u and v) wind tendencies from the dynamics. The u and v tendencies from the physics had reasonable values.

Additional Context

Machine = Hera Compiler = Intel SDF = FV3_HRRR

Output

Attached are two plots of the u and v dynamics tendencies at the first output time at level 33 during the restart run. The run was restarted at f06 and output was every 3 minutes. Note, a test was made where the restart interval was set to 1 hour (instead of 3 minutes), and the garbage values were also seen in the first history file of the restart run.

udyn_bug vdyn_bug

dustinswales commented 4 months ago

@ericaligo-NOAA Can you provide a link to the directions you followed? I will see if I can replicate this and find a solution.

ericaligo-NOAA commented 4 months ago

@ericaligo-NOAA Can you provide a link to the directions you followed? I will see if I can replicate this and find a solution.

https://ccpp-techdoc.readthedocs.io/en/latest/ParamSpecificOutput.html

ericaligo-NOAA commented 4 months ago

@ericaligo-NOAA Can you provide a link to the directions you followed? I will see if I can replicate this and find a solution.

https://ccpp-techdoc.readthedocs.io/en/latest/ParamSpecificOutput.html

Note that the u and v dynamics tendencies are okay/make sense before I do the restart. The values are very large at the first restart time and for the remainder of the forecast. The other physics and dynamics tendencies are fine for the restart.

grantfirl commented 4 months ago

I'm pretty sure that this isn't a CCPP issue and would be better discussed in https://github.com/ufs-community/ufs-weather-model/discussions/2141