The folders contain netcdf files from the simulations conducted for the paper. The name of folders are the same than the names used in the paper:

arctic_2004  
arctic_2050  
arctic_2050_EM2004  
arctic_2050_shipping

The netcdf files contain monthly data for 20 years each and values between 60° and 90°N.

The most important variables discussed in the paper are listed below (the files contain more variables). In most cases, the units are written to the netcdf files, otherwise there are mentioned below. 
If you want to use other variables than the ones listed below, please contact Anina Gilgen (anina.gilgen@env.ethz.ch) or David Neubauer (david.neubauer@env.ethz.ch).

activm-stream:
	REFFL_ACC_CW: in-cloud effective cloud droplet radius
	REFFI_ACC_CW: in-cloud effective ice crystal radius
	LWC_ACC_CW: in-cloud liquid water content
	IWC_ACC_CW: in-cloud ice water content
	CDNC_ACC_CW: in-cloud cloud droplet number concentration
	ICNC_ACC_CW: in-cloud ice crystal number concentration
	CDNC_BURDEN_ACC_CW: in-cloud cloud droplet number concentration burden

	the same variables without the "_CW" at the end -> all-sky values

	Note: For the in-cloud variables, I set the missing values (occurring in gridboxes where no clouds are present during the whole month in the original data) to zero in the data. You can set them back to NAN if wished. 
	When averaging (e.g. calculating temporal and zonal means), it makes a difference whether 0 or NAN is used. 
	It makes a difference for regions where only a few gridboxes have values, e.g. at 250 hPa for CDNC (difference 2050 and 2004). Without setting the values to zero,
	the differences would be considerably higher, but would be only representative for a few gridboxes.   
burden-stream:
	burden_SS: sea salt burden
	burden_DMS: DMS burden
	burden_BC: black carbon burden
	burden_OC: organic carbon burden
	burden_SO4: sulphuric acid burden

debugsm-stream:
	ddf01: number of ice crystals freezing per timestep due to contact freezing (dust)
	ddf02: number of ice crystals freezing per timestep due to immersion freezing (BC)
	ddf03: number of ice crystals freezing per timestep due to immersion freezing (dust)

echamm-stream:
	friac: sea ice cover
	precip: total precipitation [kg/(m2*s)]
	aprc: convective precipitation
	tsurf: surface temperature
	aclcov: cloud cover
	CF_LW: long-wave cloud radiative effect [W/m2]
	CF_SW: short-wave cloud radiative effect [W/m2]
	srad0: net short-wave radiation at TOA
	trad0: net long-wave radiation at TOA
	wind10: wind speed at 10m altitude

emim-stream:
	emi_SS: sea salt emissions
	emi_BC_ships: black carbon ship emissions
	emi_OC_ships: organic carbon ship emissions
	emi_SO2_ships: sulphur dioxid ship emissions

forcing-stream:
	FORCING_TOTAL_TOP: radiative forcing of (all) aerosols at the TOA (as defined in the paper) [W/m2]
	FORCING_CLEAR_TOP: radiative forcing of (all) aerosols at the TOA (as defined in the paper), clear-sky conditions [W/m2]

hamm-stream:
	rwet_NS: wet radius of soluble nucleation mode
	rwet_KS: wet radius of soluble Aitken mode
	rwet_AS: wet radius of soluble accumulation mode
	rwet_CS: wet radius of soluble coarse mode
	rwet_KI: wet radius of insoluble Aitken mode
	rwet AI: wet radius of insoluble accumulation mode
	rwet_CI: wet radius of insoluble coarse mode
	D_COND_SO4: condensation of sulphate on aerosol
	D_NUC_SO4: nucleation of sulphate

radm:
	ABS_2D_550nm: absorption optical thickness (at 550nm)
	TAU_2D_550nm: optical thickness (at 550nm)

sfcbc:
	FSW_TOTAL_TOP: radiative forcing of BC deposited on snow

tracerm:
	NUM_NS: number concentration of soluble nucleation mode
	NUM_KS: number concentration of soluble Aitken mode
	NUM_AS: number concentration of soluble accumulation mode
	NUM_CS: number concentration of soluble coarse mode
	NUM_KI: number concentration of insoluble Aitken mode
	NUM_AI: number concentration of insoluble accumulation mode
	NUM_CI: number concentration of insoluble coarse mode
	BC_TOT: total BC mass concentration [kg/kg]
	DU_TOT: total dust concentration [kg/kg]

The results from the radiative kernel method (all in W/m2) can be found in the directory "RK". The subdirectory 2050-2004 contains the differences in cloud radiative effects between 
the simulation arctic_2050_EM2004 and arctic_2004. The subdirectory 2050ship-2050 contains the differences in cloud radiative effects between the simulation
arctic_2050_shipping and the simulation arctic_2050. The subsubdirectories mean, stda, and stdb refer to the average and the two standard deviations, respectively, as described in the Appendix of the paper.
The file names are structured the following way:
	JA/SO/SO_2: refers to the time period over which the results are averaged. JA=July/August, SO=September/October. SO_2 also refers to September/October, but contains missing values (see paper for more details).
	fdbk/fdbkd: fdbk contains the standard short-wave (SW_cld_fdbk) and long-wave (LW_cld_fdbk) changes in cloud radiative effect using the radiative kernel method. In fdbkd, these 
		changes are split into contributions from changes in cloud top altitude (SW_cld_fdbkp and LW_cld_fdbkp), changes in cloud optical thickness (SW_cld_fdbkt and LW_cld_fdbkt), 
		changes in cloud cover (SW_cld_fdbk0 and LW_cld_fdbk0), and the residuals (SW_cld_fdbkr and LW_cld_fdbkr). 
	low/non-low/: changes in the cloud radiative effects refer either to all altitudes (neither low nor non-low) or are split into changes in low clouds (low) and free-tropospheric clouds (non-low), 
		as defined in the paper.