question on mkurbanparMod in mksurfdata_esmf

[Yuan Sun]

Have you made any changes to files in the source tree?
No

Describe your problem or question:
Hi all,

We tried to increase urban parameter variability across 33 global regions to 3999 cities. However, the modified raw data does not fit the mksurfdata_esmf because it is out of memory when allocate(data_i(max_regions, ns_i), stat=ier). I do not have a good idea to address it because the algorithm is to determine a dominant region among all regions.

By the way, in mkurbanparMod.F90, Line 304, why data_i, should be data_o?
allocate(data_o(max_regions, ns_o), stat=ier)
if (ier/=0) call shr_sys_abort('error allocating data_i(max_regions, ns_o)')

Thanks for any comments.

Best,
Yuan

 

[Yuan Sun]

I raised an issue: A typo in the mkurbanparMod.F90 code · Issue #3898 · ESCOMP/CTSM

 

[oleson]

As discussed offline, you could try spreading out the mksurfdata_esmf across more nodes if you have a memory problem.
I agree that the diagnostic statement is incorrect.