Restart fails

[jedwards4b]

Restart test ERS.ne30_g16.B1850C5.gust_intel fails with error
NetCDF: Index exceeds dimension bound
when Reading history variable CFAD_DBZE94_CS
from the history restart file.

double CFAD_DBZE94_CS(cosp_ht, cosp_dbze, ncol) ;
CFAD_DBZE94_CS:mdims = 9, 6 ;
CFAD_DBZE94_CS:_FillValue = -1.e+30 ;
CFAD_DBZE94_CS:missing_value = -1.e+30 ;
CFAD_DBZE94_CS:units = "fraction" ;
CFAD_DBZE94_CS:long_name = "Radar Reflectivity Factor CFAD (94 GHz)" ;
CFAD_DBZE94_CS:cell_methods = "time: mean" ;

[nusbaume]

Hmm, this variable is coming from the COSP satellite simulator. To be honest I am not sure a user would ever run COSP with this model version, especially because it is basically unusable for any paleoclimate simulation (as there are no satellites to compare against). Thus maybe we can fix this issue by just turning COSP off entirely?

@jiang-zhu What are your thoughts?

[jedwards4b]

Changes related to cosp that @jiang-zhu wanted in the test case:

CAM_CONFIG_OPTS=-cosp
cat user_nl_cam
! Users should add all user specific namelist changes below in the form of
! namelist_var = new_namelist_value
cosp_lisccp_sim = .true.
cosp_llidar_sim = .true.

I'll try again without cosp.

[jiang-zhu]

@nusbaume @jedwards4b
I agree. We can turn off cosp for paleoclimate runs.
I left it on mainly because it does not impact the model throughput and cost (<5%).

[jiang-zhu]

If a similar porting effort is planned for the CESM1.3_iHESP (no isotopes), then people may need the cosp module.

[jedwards4b]

As I understand it this model as written can only be used with isotopes. @jiang-zhu can you confirm?

[jiang-zhu]

@jedwards4b
Yes, you are right!
The physical climate of this code can reproduce the CESM1.3_iHESP bit-for-bit, although I only tested f19_g16.