The data described below is output from the Bern3D intermediate complexity model and idealized CO2 increase-decrease simulations used in Jeltsch-Thömmes et al., Environ. Res. Lett. 15 (2020) 124026, https://doi.org/10.1088/1748-9326/abc4af


The data are provided as .csv and .nc files
There are different types of data


1) TIMESERIES DATA (Fig. 1 and 2)
=================================
The name of the files indicates the variable:
	co2_ts.csv			change in atm. co2 [ppm]
	cumulativeEmissions_ts.csv	cumulative emissions [GtC]
	cumulativeAOflux_ts.csv	cumulative atm-ocean C flux [GtC]
	cumulativeABflux_ts.csv	cumulative atm-land C flux [GtC]
	sat_ts.csv			change in surface air temperature [degC]
	ohc_ts.csv			change in ocean heat content [10^24 J]
	amoc_ts.csv			change in Atlantic meridional overturning circulation strength [Sv]
	seaice_ts.csv			fraction of pre-industrial sea-ice area remaining [fraction of PI]
	
The first row in the .csv files contains the header, which indicates the experiment. The naming convention is as follows:
c4k#_###

c4 indicates the maximum co2 as times pre-industrial (4 times)
k# indicates the equilibrium climate sensitivity of the respective simulation in degrees C (k2 to k5)
### indicates the rate of CDR:
	010:	0.1% yr^-1
	010:	0.3% yr^-1
	010:	0.5% yr^-1
	010:	0.7% yr^-1
	100:	1% yr^-1
	200:	2% yr^-1
	400:	4% yr^-1
	600:	6% yr^-1


2) HYSTERESIS DATA (Fig. 3)
===========================
The name of the files indicates the variables:
	cumulativeEmissions_sat.csv		cumulative emissions and change in surface air temperature [degC]
	cumulativeEmissions_OHCsurf.csv	cumulative emissions and change in upper ocean heat content (0-700 m) [10^24 J]
	cumulativeEmissions_o2thermo.csv	cumulative emissions and change in thermocline (200-600 m) o2 [mmol m^-3]
	cumulativeEmissions_OM_arag.csv	cumulative emissions and fraction of water in the uppermost 175 m with omegar_aragonite saturation state >3 [fraction]

each file contains the time (simulation year) as well as cumulative emissions (cumuEmis) and the respective variable (same naming as in filename) for all the experiments (see timeseries data for naming convention)


3) SPATIAL DATA (Fig. 4 and 5)
==============================
All data for Fig. 4 and 5 are contained in one single .nc file (fig4_5_data.nc) with a varibale for each map shown in Fig. 4 and 5:
	c4k2_100_sat		hysteresis (down-path minus up-path) in surface air temperature at cumulative emissions of 1000 GtC, ECS=2 degC, in [degC]
	c4k3_100_sat		hysteresis (down-path minus up-path) in surface air temperature at cumulative emissions of 1000 GtC, ECS=3 degC, in [degC]
	c4k5_100_sat		hysteresis (down-path minus up-path) in surface air temperature at cumulative emissions of 1000 GtC, ECS=5 degC, in [degC]
	
	c4k2_100_o2thermo	hysteresis (down-path minus up-path) in thermocline (200-600 m) o2 at cumulative emissions of 1000 GtC, ECS=2 degC, in [mmol m^-3]
	c4k3_100_o2thermo	hysteresis (down-path minus up-path) in thermocline (200-600 m) o2 at cumulative emissions of 1000 GtC, ECS=3 degC, in [mmol m^-3]
	c4k5_100_o2thermo	hysteresis (down-path minus up-path) in thermocline (200-600 m) o2 at cumulative emissions of 1000 GtC, ECS=5 degC, in [mmol m^-3]
	
	c4k3_100_Om_arag_up	mean aragonite saturation state of the uppermost 175 m at cumulative emissions of 1000 GtC on the up-path, ECS=3 degC, [unitless]
	c4k3_100_Om_arag_do	mean aragonite saturation state of the uppermost 175 m at cumulative emissions of 1000 GtC on the down-path, ECS=3 degC, [unitless]






The files can be readily importet in python, for example, by:
    import pandas as pd
    import xarray as xr
    
    # for the .csv files
    df = pd.read_csv('path+filename', sep=',', header=0, index_col=None)
    
    # for the .nc files
    ds = xr.open_dataset('path+filename')


For additional information or in case of questions please contact:
Aurich Jeltsch-Thömmes
aurich.jeltsch-thoemmes@unibe.ch