issue when resubmiting?

[m_valmartin@sheffield_ac_uk]

Hi, I am doing an SSP simulation, in 10 year runs with multiple resubmissions. My compset is SSP245_DATM%GSWP3v1_CLM50%BGC-CROP_SICE_SOCN_MOSART_CISM2%NOEVOLVE_SWAV

PCO2 and TBOT get bump up every time there is a resubmission (see screenshot). Is there an issue with the initial conditions every time the run gets resubmitted? I have "continue_run" as true but perhaps am I missing something else?

Thanks a lot!

Regards,

Maria

 

[oleson]

I assume you are using anomaly forcing. There are oscillations associated with that but they shouldn't be that large.
Can you attach your lnd_in, datm_in, and lnd and atm logs for two consecutive model runs?
Or just point me to your run/case directories.

 

[m_valmartin@sheffield_ac_uk]

Hi Keith, Thanks a lot for getting back to me. I checked /glade/scratch/swensosc/anomaly_forcing/af.tas.CESM.SSP245.2015-2100.nc
and /glade/p/cesmdata/cseg/inputdata/atm/datm7/CO2/fco2_datm_globalSSP2-4.5__simyr_2014-2501_CMIP6_c190506.nc. The anomalies are fine, that is, no jumps in the trends.

Yesterday I did a test run with resubmitting every 25 years and the jump appeared every 10 years too. So it doesnt look like it is an initial condition issue at resubmitting.

My case is /glade/work/mariavm/cases/i.clm5.SSP2.000/

Thanks again. I appreciate your time!

Best,

Maria

 

[oleson]

Thanks. I was able to replicate your problem with my own case. It seems to just be happening over Antarctica. The northern hemisphere looks reasonable for example.
I think the cause is related to the fact that the GSWP3V1 forcing data had originally ended in 2010. When it was extended to 2014, unrealistically warm temperature forcing over Antarctica was used for 2011-2014. So as you loop over 2011-2014 in your anomaly forcing run you'll see a jump in temperature forcing over Antarctica. There is probably something similar happening with pressure which would affect PCO2.