To use the data: The data can be downloaded and processed using Python or a programming language that can handle netCDF data. The python (.py) files included were used by the authors for producing figures and analysis for the corresponding research article to this dataset. To use the python files, you need to correctly set up the paths to the correct directories and files in O3_FYSP_paper.py and Analysis_Functions.py. h0 files should be in path + 'Ozone_CH4_paper/h0/', h2 files in path + 'Ozone_CH4_paper/h2/', and solar files in path + 'Ozone_CH4_paper/Solar_files/'. The path can be any directory you have permissions for. For example, on a Linux system, using /localhome/user/ is how the path was set up for the python file when used by the authors. To run the python code, run the O3_FYSP_paper.py file. If you have not used python before, downloading python via Anaconda and installing xarray should allow you to see the code (see here: https://www.anaconda.com/products/individual, and here: http://xarray.pydata.org/en/stable/getting-started-guide/installing.html). Figures produced will be deposited into: path + 'Ozone_CH4_paper/ The h0 and h2 files correspond to the following simulations, which are described in Table 1 of the associated manuscript: PI: b.e21.BWma1850.f19_g17.baseline.cam.h0.0014-0017.nc; b.e21.BWma1850.f19_g17.baseline.cam.h2.0014-0017.zm.nc 150% PAL: b.e21.BWma1850.f19_g17.150pc_o2.002.cam.h0.0034-0037.nc; b.e21.BWma1850.f19_g17.150pc_o2.002.cam.h2.0034-0037.zm.nc 50% PAL: b.e21.BWma1850.f19_g17.50pc_o2.002.cam.h0.0030-0033.nc; b.e21.BWma1850.f19_g17.50pc_o2.002.cam.h2.0030-0033.zm.nc 10% PAL: b.e21.BWma1850.f19_g17.10pc_o2.001.cam.h0.0034-0037.nc; b.e21.BWma1850.f19_g17.10pc_o2.001.cam.h2.0034-0037.zm.nc 5% PAL: b.e21.BWma1850.f19_g17.5pc_o2.002.cam.h0.0032-0035.nc; b.e21.BWma1850.f19_g17.5pc_o2.002.cam.h2.0032-0035.zm.nc 1% PAL: b.e21.BWma1850.f19_g17.1pc_o2.cam.h0.0028-0031.nc; b.e21.BWma1850.f19_g17.1pc_o2.cam.h2.0028-0031.zm.nc CH4 em1: b.e21.BWma1850.f19_g17.1pc_o2.CH4_flux.001.cam.h0.0044-0047.nc CH4 em0.1: b.e21.BWma1850.f19_g17.1pc_o2.0.1PD_CH4_flux.001.cam.h0.0044-0047.nc YS: b.e21.BWma1850.f19_g17.1pc_o2.YoungSun.004.cam.h0.0060-0063.nc YS 4xCO2: b.e21.BWma1850.f19_g17.1pc_o2.YoungSun.4xCO2.005.cam.h0.0089-0092.nc 0.5% PAL: b.e21.BWma1850.f19_g17.0.5pc_o2.001.cam.h0.0046-0049.nc; b.e21.BWma1850.f19_g17.0.5pc_o2.001.cam.h2.0046-0049.zm.nc 0.1% PAL: b.e21.BWma1850.f19_g17.0.1pc_o2.001.cam.h0.0080-0083.nc; b.e21.BWma1850.f19_g17.0.1pc_o2.001.cam.h2.0080-0083.zm.nc In addition, there are four files for calculating the three major loss channels of methane (see Fig. 8 in the associated article at https://doi.org/10.1098/rsos.211165): PI: b.e21.BWma1850.f19_g17.baseline.cam.h2.CH4_loss.0014-0017.zm.nc 10% PAL: b.e21.BWma1850.f19_g17.10pc_o2.001.cam.h2.CH4_loss.0034-0037.zm.nc 1% PAL: b.e21.BWma1850.f19_g17.1pc_o2.cam.h2.CH4_loss.0028-0031.zm.nc 0.1% PAL: b.e21.BWma1850.f19_g17.0.1pc_o2.001.cam.h2.CH4_loss.0080-0083.zm.nc Note that b.e21.BWma1850 corresponds to a WACM6 (CESM2.1) pre-industrial (year 1850) simulation. f19_g17 describes the grid setup. 96 x 144 latitude x longitude, corresponding to 1.875 degrees x 2.5 degrees. The model years of the dataset are given by 0014-0017, for years 14, 15, 16, and 17. 'zm' means a zonal mean was applied to the file. A file is provided which contains gaussian weights for the area of grid cells (gaussian weights vary with latitude): GW_File.nc Solar files are as follows: Present day Sun: SolarForcingCMIP6piControl_c160921.nc Sun's spectrum 2 Gyr ago (2.0 Ga): SolarForcingCMIP6piControl_c160921_2.0Ga.nc

The data was originally produced using the Earth System Model WACCM6. The corresponding paper that describes WACCM can be found at the following DOI: https://doi.org/10.1029/2019JD030943. WACCM6 is a model configuration of the Community Earth System Model version 2 (CESM2). We used CESM2.1.3, which can be downloaded from the following URL: https://www.cesm.ucar.edu/models/cesm2/release_download.html. The atmospheric data from each simulation was output in terms of monthly means (h0 files; h for history), and 5 day averages (h2 files). For both h0 and h2 files, the last 4 years of each simulation was averaged using netCDF operators. ncra for time average (h0 and h2 files), and ncwa -alon for zonal mean (h2 files). The data has been processed using the Python programming language. Two .py files are included named O3_FYSP_paper.py and Analysis_Functions.py. The python files may not be well-commented in all places. To produce the solar file which represent the Sun 2 billion years ago (2 Gyr), we used an existing solar model by Claire et al. 2012 (https://doi.org/10.1088/0004-637X/757/1/95). The model can be downloaded here: http://depts.washington.edu/naivpl/content/models/solarflux The datasets Kasting_Catling_Fig10_data.csv and ROCKE3D_Fig3_O3_col.csv are from the following papers: https://doi.org/10.1146/annurev.astro.41.071601.170049 and https://doi.org/10.3847/1538-4365/aa7a06, respectively. The data was recovered using WebPlotDigitizer (https://automeris.io/WebPlotDigitizer/).

The history of molecular oxygen (O2) in Earth’s atmosphere is still debated; however, geological evidence supports at least two major episodes where O2 increased by an order of magnitude or more: the Great Oxidation Event (GOE) and the Neoproterozoic Oxidation Event. O2 concentrations have likely fluctuated (between 10−3 and 1.5 times the present atmospheric level) since the GOE ∼ 2.4 Gyr ago, resulting in a time-varying ozone (O3) layer. Using a three-dimensional (3D) chemistry climate model, we simulate changes in O3 in Earth’s atmosphere since the GOE and consider the implications for surface habitability, and glaciation during the Mesoproterozoic. We find lower O3 columns (reduced by up to 4.68 times for a given O2 level) compared to previous work; hence, higher fluxes of biologically harmful UV radiation would have reached the surface. Reduced O3 leads to enhanced tropospheric production of the hydroxyl radical (OH) which then substantially reduces the lifetime of methane (CH4). We show that a CH4 supported greenhouse effect during the Mesoproterozoic is highly unlikely. The reduced O3 columns we simulate have important implications for astrobiological and terrestrial habitability, demonstrating the relevance of 3D chemistry-climate simulations when assessing paleoclimates and the habitability of faraway worlds.