Im trying to run MCIP for a WRF run with 61 land use categories (WUDAPT categories). But I get the following error:
**********************************************************************
*** SUBROUTINE: GETLUSE
*** UNKNOWN LAND USE SOURCE AND MAX CATEGORIES
*** LAND USE SOURCE = MODIFIED_IGBP_MODIS
*** NUMBER OF CATEGORIES = 61
**********************************************************************
*** ERROR ABORT in subroutine GETLUSE
ABNORMAL TERMINATION IN GETLUSE
Date and time 1:00:00 Nov 10, 2012 (2012315:010000)
Error running mcip
Im guessing this is related with the WUDAPT land use categories as I had no problem running MCIP with the default MODIS ones.
Is there an specific flag or something I should take into account when running with WUDAPT land use categories?
Thank you very much in advance
Kind regards
Solange
Thank you for pointing out this error in MCIP. The changes for the WUDAPT categories in WRF have been a little rocky over the past couple of years. The MODIS land use was extended with the WUDAPT categories beginning in WRFv4.3 (released 10 May 2021), then updated in WRFv4.4.1 (released 25 Aug 2021) to resolve incompatibilities with NLCD40 and PX LSM, then updated and renumbered in WRFv4.4.2 (released 19 Dec 2022) to correct compatibility issues with NLCD40 and NOAH LSM.
Over this period, MCIP was not updated for the various physics and land use changes that were introduced in WRF, and that’s on the “to do” list. Unfortunately, the accommodating WUDAPT with 61 categories will require some intervention in both MCIP and CMAQ’s CTM to ensure that the correct information is persisted from WRF through MCIP to CMAQ and that we dynamically adapt to the various changes that were version-specific in WRF.
At this time, I do not have an easy solution to offer you on getting WUDAPT categories through MCIP. I’ll put it in my queue for MCIP modifications. I will post here when I have a workable solution (in consultation with others on the team to ensure that we address the issue comprehensively). I may reach out to you (via the Forum) to request a copy of your dataset for testing purposes.
So if I understand correctly, is not possible to use WUDAPT classification with MCIP even for WRF 4.3.2 or 4.3.3 outputs (which i believe are 41 categories if Im not mistaken). Is this correct?
I think that is correct, but I would need to check it. There are some characteristics that are associated with land use categories by classification scheme, and those characteristics may not be set up correctly in MCIP to account for the additional WUDAPT categories.
As far as I could tell, we did not update MCIP for any changes to the WRF physics or land use options beyond WRFv.4.1. The overlapping land use indices could be problematic, even if MCIP finished correctly. The initial implementation of WUDAPT used categories 31-41 (I think), but indices 31-40 were already assigned for the NLCD classification. With WRFv4.4.2 (I think), NCAR shifted WUDAPT to instead use indices 51-61 to avoid clashing with NLCD.
Testing of MCIP with WUDAPT will require adaptations for both PX and Noah LSMs, and it may require additional testing within CMAQ’s CTM to ensure that all of the values are passed through and used correctly with both M3Dry and STAGE.
Sorry to say, but this one may take a bit of time, and I’ll need to find how to inject it with my other priorities. I wish I had a quick and dirty fix for you, but I want to make sure that the programs don’t just run to completion and, instead, they give you the intended uses of the data.
@SolL I had similar issues. To solve this issue, I tried to convert LCZ landuse dataset to traditional 3 urban categories used in the WRF model which is also compatible with CMAQ.
Thank you for your suggestion. Yes, I believe that the only way to use the LCZ landuse is to modify the wrfout files by mapping those categories to the MODIS or UGS ones.
I thought about it, but I ended up not doing it. Did you tried it and succeeded in runing mcip?
I didn’t make any changes to wrfout files. I applied this methodology after I ctreated my geo_em files. However, i validated the model with both LCZ and 3UB land use and they had similar performance.
If you want to modify your wrfout files, remember to adjust the global attribute as well.