To support the WMO/UNEP Scientific Assessment of Ozone Depletion Report 2022, we are asking modelling groups to perform a number of simulations to allow for both an assessment of the models, the first comprehensive assessment of model performance since CCMVal-2 in 2010, and longer scenario simulations to project ozone recovery.
All the details of the setup and forcings for the first model experiment, the REF-D1 historical hindcast simulation covering 1960 – 2018 and using forcing as closely following the observed historical evolution, can be found here.
In addition to the historical simulation, two scenarios have been defined and assigned a high priority to provide updated projections of ozone recovery (REF-D2) and information on the effects of geoengineering through stratospheric aerosol injection (SEN-D2-geo). Additional, lower priority, scenarios are included to The detailed description of the experimental setup for the scenario simulations can be found here.
Data request, netCDF conversion and data submission
An excel spreadsheet version of the data request for all model experiments can be found here. Model data is to be provided in version 4 netCDF, with all metadata following the CF (Climate and Forecast) conventions. As such, we strongly recommend groups use the CMOR utility for the generation of the netCDF files.
The MIP tables (or CMOR tables) in json format for use with CMOR3 can be found on github at https://github.com/cedadev/ccmi-2022/. Note that the institution_id and source_id for each model submission must be registered in the reference version of the controlled vocabulary json file (CCMI2022_CV.json) used for quality control at CEDA. Before converting data, please e-mail us with the information for your group and model so that it can be included.
Instructions on how to submit data to CEDA can be found in this pdf document.
We do strongly encourage groups to use CMOR for netCDF file conversion as CMOR should check all of the required metadata is correctly formatted and matches the specifications in the controlled vocabulary. If you are not using CMOR, please pay special attention to the case (lower and upper case) of the letters in variables like experiment_id. For example, ‘REFD1’ and ‘RefD1’ are not considered a match for the specified experiment_id ‘refD1’ and quality control will fail.
Ancillary Information on REF-D1 Forcings
The required forcings, largely following datasets produced for the historical period of CMIP6 up to 2014 and then SSP2-4.5 for 2015 – 2018, are detailed in the REF-D1 experimental description. Here, a bit of additional information is provided on particular datasets that deviate from the standard CMIP6 recommendations.
Near-surface methane concentration
A slightly modified time series for the near-surface methane mixing ratio has been recommended for REF-D1 over the period 2015 – 2018. This data has been created by scaling the original SSP2-4.5 methane data by a globally constant scaling factor to bring the year-to-year changes in methane in line with more recent observations given in the annual WMO GHG bulletin and the NOAA/ESRL Global Monitoring Laboratory Atmospheric Greenhouse Gas Index. A comparison of the global average CO2, CH4 and N2O near-surface concentrations from the different Tier 1 and 2 CMIP6 SSPs against these observations for 2012 – 2019, can be found here. The python script that was used to modify the original v1.2.1 SSP 2-4.5 methane data files can be found here.
Ozone Depleting Substances
The WMO-2018 scenario of global-average near-surface mixing ratios was based on observations up to the beginning of 2017, with projected values used for subsequent years. To more accurately reflect the recent behaviour of different ODSs, flask network observations from NOAA ESRL were used to extend the WMO-2018 time series for 2017 – 2018 for CFC-11, CFC-12, CCl4, HCFC-22 and CH3Cl. For convenience, the original WMO-2018 time series was also extrapolated backwards from 1955 to 1949. A comparison of the original WMO-2018 baseline scenario and the revised time series for the REF-D1 simulation can be found here, and the python script that was used to revise the WMO-2018 time series can be found here.
The time series of equatorial zonal winds to be used for introducing, or constraining, a QBO for the historical hindcast simulations has been extended from the dataset produced for CCMI-1. The python script that was used to extend the dataset using the Singapore winds from FUB can be found here, and plots of the CCMI-1 and the extended time series can be found here.
Ancillary information for the scenario forcings
For modelling groups that do not internally generate a QBO, we ask that zonal winds in the QBO domain are relaxed towards a prescribed timeseries of monthly average zonal winds for the entire 1960 – 2100 period. A figure detailing the construction of the extended timeseries from observations and the full timeseries can be found here. A compressed tar file (.tgz) of the input files and programs to produce the extended QBO timeseries can be found here.