Hello everyone,
I've been working on ACI by running a pair of simulations: one with PD emissions from 2010 and the other with PI emissions from 1850. By examining the differences in the TOA radiation between these cases, I can derive the aerosol-cloud interaction (ACI) values from the simulation.
I crafted my own run script for this purpose. However, the results I obtained differ significantly from those presented by Gettelman et al. (2019, see ref below). Specifically, in their paper, the ACI for CESMv2 CAM6 is approximately -1.8 (as seen in Fig. 3D). In contrast, my simulations indicate a global mean ACI of -2.6. This discrepancy concerns me, leading me to suspect potential errors in my scripts. I've attached the scripts for your reference.
Basic settings:
TOA_PD=(FSNT_d1-FSNTC_d1) - (FLNT_d1-FLNTC_d1) using PD case outputs.
TOA_PI = (FSNT_d1-FSNTC_d1) - (FLNT_d1-FLNTC_d1) using PI case outputs.
ACI=TOA_PD-TOA_PI. Then calculate the global mean.
I have some guessing that might be influencing the results:
Ref:
Gettelman, A., Hannay, C., Bacmeister, J. T., Neale, R. B., Pendergrass, A. G., Danabasoglu, G., ... & Mills, M. J. (2019). High climate sensitivity in the Community Earth System Model version 2 (CESM2). Geophysical Research Letters, 46(14), 8329-8337.
I've been working on ACI by running a pair of simulations: one with PD emissions from 2010 and the other with PI emissions from 1850. By examining the differences in the TOA radiation between these cases, I can derive the aerosol-cloud interaction (ACI) values from the simulation.
I crafted my own run script for this purpose. However, the results I obtained differ significantly from those presented by Gettelman et al. (2019, see ref below). Specifically, in their paper, the ACI for CESMv2 CAM6 is approximately -1.8 (as seen in Fig. 3D). In contrast, my simulations indicate a global mean ACI of -2.6. This discrepancy concerns me, leading me to suspect potential errors in my scripts. I've attached the scripts for your reference.
Basic settings:
- 2-yr simulation starting from 2010-01-01
- Compset: F2010climo
- resolution: f09_f09_mg17
- CAM_CONFIG_OPTS val '-phys cam6 -chem trop_mam4 -offline_dyn -nlev 56' (need -chem trop ? )
- CAM_CONFIG_OPTS --append --val='-cosp'
- met_data_file = '2010/MERRA2_0.9x1.25_20100101.nc'
- met_rlx_time = 6.
- met_qflx_factor = 1.
- ncdata ='/glade/p/cesmdata/cseg/inputdata/atm/cam/inic/fv/f.e20.FCSD.f09_f09_mg17.cesm2.1-exp002.001.cam.i.2005-01-01-00000_c180801.nc'
- rad_diag_1 = 'A:Q:H2O', 'N:O2:O2', 'N:CO2:CO2', 'N:ozone:O3', 'N:N2O:N2O', 'N:CH4:CH4', 'N:CFC11:CFC11', 'N:CFC12:CFC12',
- ext_frc_specifier/ srf_frc_specifier= tons of files under ‘/glade/p/cesmdata/cseg/inputdata/atm/cam/chem/emis/CMIP6_emissions_1750_2015/’
- ext_frc_type /srf_emis_type= 'CYCLICAL'
- ext_frc_cycle_yr /srf_frc_cycle_yr = 2010 (1850 in PI script)
TOA_PD=(FSNT_d1-FSNTC_d1) - (FLNT_d1-FLNTC_d1) using PD case outputs.
TOA_PI = (FSNT_d1-FSNTC_d1) - (FLNT_d1-FLNTC_d1) using PI case outputs.
ACI=TOA_PD-TOA_PI. Then calculate the global mean.
I have some guessing that might be influencing the results:
- I've linked nearly all the emission files I could find, from ‘CMIP6_emissions_1750_2015’ folder on Cheyenne, to ext/srf_frc_specifier. Could I have overdone it, or possibly overlooked something crucial?
- I use F2010climo and utilized an ncdata from 2005. How can I ensure I'm selecting the most appropriate ncdata for 2010? Like can I use 2005 ncdata for a run starting at 2000 as there might be no data for 2000.
- I just submitted a new pair of runs using F2000climo and start time =2000, not sure if this is the case.
- Is it possible that I skipped essential steps before submission?
Ref:
Gettelman, A., Hannay, C., Bacmeister, J. T., Neale, R. B., Pendergrass, A. G., Danabasoglu, G., ... & Mills, M. J. (2019). High climate sensitivity in the Community Earth System Model version 2 (CESM2). Geophysical Research Letters, 46(14), 8329-8337.