Scheduled Downtime
On Tuesday 24 October 2023 @ 5pm MT the forums will be in read only mode in preparation for the downtime. On Wednesday 25 October 2023 @ 5am MT, this website will be down for maintenance and expected to return online later in the morning.
Normal Operations
The forums are back online with normal operations. If you notice any issues or errors related to the forums, please reach out to help@ucar.edu

Why model starts at different carbon value with same land restart file ?

R

rambhari01

Ram
New Member
Dear CESM Community,

What version of the code are you using?
- using CESSM2.2.2


Describe your problem or question:
I am starting two different runs using the composets given below.

1850_CAM60_CLM50%BGC_CICE%PRES_DOCN%DOM_MOSART_SGLC_SWAV.

&

FHIST_CAM60_CLM50%BGC_CICE%PRES_DOCN%DOM_MOSART_SGLC_SWAV


The difference between these 2 runs sought to be as first would be 1850 control run with everything to be of 1850 and 2nd is supposed to use the FHIST (as time) so that it can use the historical forcings and we fiexd the CO2 to 1850 only.

Both of these runs use the same initial land restart files as I want to branch these runs form an existing run (same strart point). for this I used the namelist "

finidat = "F1850_CAM6CLM5BGC_NOcrop_CTRL_I3.clm2.r.1910-01-01-00000.nc" " in use_nml_clm for both of these runs.

After running both of these runs for several years, I plotted the carbon stocks for these 2 runs. Everything seems okay but I it can be seen that both of these run start at different total carbon values for these 2. Which is unexpected. Since I am using the above mentioned same restart files file both of these run I expected to start them from the same carbon value.

However, when i was using this restart file for the "FHIST.." compset. initially, model thrown an error and it suggested to use "use_init_interp = .true." in name list variable. When i used this model runs smoothly but I noticed a difference in starting carbon values.

Please help me understanding is it the different time values in compset (1850 and FHIST) has made this difference, how?

Please see the attached plot and see the different start C value s for both of these run (blue and red) instead of using the same land restart files.



Thank you.

-Ram
 

Attachments

  • TOTCOLC_TOTECOSYSC_TOTLITC_TOTSOMC_TOTVEGC_TOTFIRE_ANNUAL_time_series.pdf
    29 KB · Views: 2
O

oleson

Keith Oleson
CSEG and Liaisons
Staff member
The FHIST comes out of the box with a year 2000 surface dataset, while the other configuration uses an 1850 dataset, so maybe the differences you are seeing are just due to the different surface distributions...
 
R

rambhari01

Ram
New Member
Thanks Keith,
Yes this explains the difference.
Could you please suggest how I can make/change those settings such that my “FHIST” should use the 1850 equilibrium surface data.
I am using FHIST because my experiments is planned to start from 1850 and proceed with everything historical but only CO2 is fixed to 1850 for the land component (CLM). So I decided to use FHIST where I can fix the CO2 for land surface model by setting to “constant”.

Thank you very much.

-Ram
 
O

oleson

Keith Oleson
CSEG and Liaisons
Staff member
Specify the 1850 fsurdat in your user_nl_clm:

fsurdat = 'the 1850 file'

Do want the landcover in the FHIST to be transient or stay at 1850?
 
O

oleson

Keith Oleson
CSEG and Liaisons
Staff member
The FHIST is setup with some of the streams fixed to cycle over 2000. So you may want to change those as well (to cycle over 1850?).

&ndepdyn_nml
ndep_taxmode = 'cycle'
ndep_varlist = 'NDEP_month'
ndepmapalgo = 'bilinear'
stream_fldfilename_ndep = '/glade/campaign/cesm/cesmdata/inputdata/lnd/clm2/ndepdata/fndep_clm_hist_b.e21.BWHIST.f09_g17.CMIP6-historical-WACCM.ensmean_1849-2015_monthly_0.9x1.25_c180926.nc'
stream_year_first_ndep = 2000
stream_year_last_ndep = 2000
/
&popd_streams
popdensmapalgo = 'bilinear'
stream_fldfilename_popdens = '/glade/campaign/cesm/cesmdata/inputdata/lnd/clm2/firedata/clmforc.Li_2017_HYDEv3.2_CMIP6_hdm_0.5x0.5_AVHRR_simyr1850-2016_c180202.nc'
stream_year_first_popdens = 2000
stream_year_last_popdens = 2000
/
&urbantv_streams
stream_fldfilename_urbantv = '/glade/campaign/cesm/cesmdata/inputdata/lnd/clm2/urbandata/CLM50_tbuildmax_Oleson_2016_0.9x1.25_simyr1849-2106_c160923.nc'
stream_year_first_urbantv = 2000
stream_year_last_urbantv = 2000
urbantvmapalgo = 'nn'
/
 
R

rambhari01

Ram
New Member
Hi Keith,

Thank you very much for your detailed response. I am planning to use the rest of all transient except the CO2. which is supposed to fixed to 1850.

So, I think I need to set these above stream years ("_years_first_") to 1850 and "_years_last_" to 2016, since these files are timeseries from 1850 to 2016.?
Also what should be the replacement for "cycle" as pointed above?

Also could you please suggest if there are anything else what I need to change in order to smoothly branch out an historical run with fixed CO2 from the '1850' control run?

Thank you very much.

-Ram
 
slevis

slevis

Moderator
Staff member
So you changed the fsurdat file to 1850 as suggested. Your simulation's lnd_in file probably already includes a flanduse_timeseries file, so that should be fine.

Yes, change the year_first and year_last variables to the start and end years of your simulation. If you plan to stop the simulation in 2015, I would not worry about "cycle"; however, running beyond 2015 your ndep will cycle back to the beginning of the timeseries. You could change "cycle" to "extend" but even better may be to switch the ndep file to one that continues beyond 2015 (see other files in the same directory).

Similarly there are popdens files that go beyond 2016.
 
R

rambhari01

Ram
New Member
Hi Slevis, Keith,

Thank you very much for the confirmation on these. I tested that changing the fsurdat and it help for resolving the difference.

However, I dont find any of namelist variable as "flanduse_timeseries" file used in both of these runs (1850 or FHIST) used in lan_in file. Could you please point me the exact namelist variable for this?

OR "flanduse_timeseries" is the namelist that define the landuse/land cover file, that i should look for?

Also, i would greatly appreciate if you can suggest how should I make this change for landcover use for FHIST (historical timeseries and fixed to 1850) ?

Thank you very much.

-Ram
 
slevis

slevis

Moderator
Staff member
If you see no flanduse_timeseries file in the FHIST case's lnd_in, then I suggest creating an IHist case similar to the FHIST case. You can see options for IHist cases by running "./query_config --compsets" in the same directory where you run ./create_newcase. Pick an IHist case similar to your FHIST case and pick the same grid. Then see what flanduse_timeseries file the IHist case prescribes and add the same one to your FHIST case's user_nl_clm.
 
R

rambhari01

Ram
New Member
Thank you very much.

I will do this and will reach if more queries comes up.

Meanwhile, A connected query on the above discussion, I am running those 2 runs as mentioned above and below

1. 1850_CAM60_CLM50%BGC_CICE%PRES_DOCN%DOM_MOSART_SGLC_SWAV.
&
2. FHIST_CAM60_CLM50%BGC_CICE%PRES_DOCN%DOM_MOSART_SGLC_SWAV

These both compsets don't show any flanduse_timeseries, I am wondering what land cover these runs are using by default?

Also, when i tested changing the "fsurdat" file (from 1850 compset) and other suggested "ndep", "popdens" and "urbantv" files with changing the start and end years to 1850. I founds that both runs are now start from the point exactly as we expected. So now I think that both of these runs are now using the same land cover and other related settings for 1850 years.

Please suggest if above understanding for this implementation is correct or not?

Thank you.

-Ram
 
You must log in or register to reply here.
Top