SOC per Depth

[Chr]

Dear CLM modellers,

I have a question regarding the SOC output. I noticed that there are several variables available: TOTSOMC, TOTSOMC_1m, and SOILC_CHANGE (among others). I have two questions about it:

1. Is there a possibility to receive TOTSOMC per depth of interest (other than 1m), e.g 0-10, 10 -30 cm etc.? I saw the output variable bulkdensity is in xyzt, with z = 25 layers (default)
2. I figured that SOILC_CHANGE is the difference in SOC between two timesteps, when extrapolated from s to the time period of interest. Does this refer to TOTSOMC, and is it possible to receive it per depth as well?

Thank you for your help.

 

[oleson]

1. I think SOILC_vr (vertically-resolved soil carbon) would be what you are looking for. But I would double-check to make sure the sum of SOILC_vr is equal to TOTSOMC.

2. I don't think SOILC_CHANGE is necessarily directly related to TOTSOMC. It seems to be a FUN-related (Fixation and Uptake of Nitrogen) variable:

soilc_change(p) = (npp_active_no3_acc_total(p) + npp_active_nh4_acc_total(p) &
+ npp_nonmyc_no3_acc_total(p) &
+ npp_nonmyc_nh4_acc_total(p) + npp_fix_acc_total(p))/dt &
+ npp_Nretrans(p)
soilc_change(p) = soilc_change(p) + burned_off_carbon / dt

 

[Chr]

Thanks a lot, this worked when multiplying SOILC_vr with the delta z as described in Table 2.3, p.18, CLM5 Tech Note.

A related question:

It says in the CLM5 Tech Note, that one can adjust the soil layer depths. I would like to alter the number and magnitude of levsoi, but I am not sure how to do this.

 

[oleson]

You can specify the "depth to bedrock" for each gridcell. This is the "zbedrock" field on the surface dataset as specified by the namelist item "fsurdat".

 

[Chr]

Good morning Keith,

I have a follow-up question regarding the SOILC_vr and SOILN_vr variables. I do currently the following:


hist_mfilt = 365,365
hist_nhtfrq = -24,-24

hist_dov2xy = .true.,.false.
hist_type1d_pertape = ' ','PFTS'

hist_fincl1 = 'GRAINC_TO_FOOD','SOILC_vr','SOILN_vr'
hist_fincl2 = 'GRAINC_TO_FOOD'


This case works fine. But my case times out as soon as I change to:

hist_fincl2 = 'GRAINC_TO_FOOD','TOTSOMC','TOTSOMN'

I found the following error message in the cesm.log:

pio_support::pio_die:: myrank= -1 : ERROR:
pionfwrite_mod::write_nfdarray_double: 250 :
NetCDF: Numeric conversion not representable


and when I change to:

hist_fincl1 = 'GRAINC_TO_FOOD','SOILC_vr','SOILN_vr'

I get the following error message in the cesm.log:

pio_support::pio_die:: myrank= -1 : ERROR: box_rearrange.F90.in:
233 : box_rearrange_comp2io: size(compbuf)= 7920
not equal to size(compdof)= 10160


Do you know why this happens?

Thanks for your help.

 

[oleson]

I think it is probably because TOTSOMC, TOTSOMN, SOILC_vr, SOILN_vr are column-level variables and you are requesting pft-level output (PFTS).
Try COLS.