README.txt file accompanying NCAR CESM2 simulation output for 279-year Warming run with CO2 increasing at 1%/year and 514-year Cooling run with CO2 decreasing at 1%/year.


Citation for data:

I. Eisenman and K. Armour (2024). The radiative feedback continuum from Snowball Earth to an ice-free hothouse. Nature Communications, https://doi.org/10.1038/s41467-024-50406-w


Description of data:

The Warming and Cooling simulations are both branched from the end of year 500 of a preexisting CESM2 pre-industrial (PI) simulation ("b.e21.B1850.f09_g17.CMIP6-piControl.001" by Danabasoglu et al., 2020). The kernel analysis uses the mean annual cycle averaged over years 480-499 of this PI simulation as the reference climate. Output from this PI simulation can be downloaded from
https://www.earthsystemgrid.org/dataset/ucar.cgd.cesm2.b.e21.B1850.f09_g17.CMIP6-piControl.001.atm.proc.monthly_ave.html

For the first 150 years of the Warming run, we use a preexisting CESM2 simulation with CO2 increasing at 1%/year ("b.e21.B1PCT.f09_g17.CMIP6-1pctCO2.001" by Danabasoglu, 2019); data from this simulation is shared elsewhere and is mirrored here for convenience. This is followed by our simulation "WARMING1ppyCESM2cheyenne", which we ran until it failed with an error after year 0279. Hence the data for the Warming run comes from the following simulations:
- Year 0001-0150: b.e21.B1PCT.f09_g17.CMIP6-1pctCO2.001
- Year 0151-0279: WARMING1ppyCESM2cheyenne

For the Cooling run, we decreased CO2 at 1%/year in the simulation "COOLING1ppyCESM2cheyenne". After year 140, we adjusted the processor (PE) layout for a more economical (but slower) load balancing, which required us to branch to a new case, "COOLING1ppyCESM2cheyenne2". This simulation failed with an error after year 278, and we commented out the lines calling the error in the model code and re-ran starting from 12 years earlier using the case name "COOLING1ppyCESM2cheyenne3". This simulation failed with an error after year 332, and we commented out the lines calling this error in the model code and re-ran starting in year 333 using the case name "COOLING1ppyCESM2cheyenne5" (there is no simulation "COOLING1ppyCESM2cheyenne4" in this collection). This simulation failed with an error after year 514. Hence the data for the Cooling run comes from the following simulations:
- Year 0001-0140: COOLING1ppyCESM2cheyenne
- Year 0141-0266: COOLING1ppyCESM2cheyenne2 [ignoring the additional years 0267-0278 in this simulation)
- Year 0267-0332: COOLING1ppyCESM2cheyenne3
- Year 0333-0514: COOLING1ppyCESM2cheyenne5

Further details regarding these simulations are given in the "Methods" section of the paper (Eisenman & Armour, 2024).


Downloading the data:

Model output from these runs, including all fields used in the kernel analysis, is split into several downloadable records on Zenodo. Links to the records can be found at https://eisenman-group.github.io

Please contact Ian Eisenman (eisenman@ucsd.edu) for model output fields not included in this collection.