Updated aerosol time series

This version includes a much more realistic historical aerosol forcing time series that better matches historical observed temperature and ocean heat uptake constraints (paper submitted).

Tiny change to ssp370-lowNTCF-aerchemmip

Minor release as come of the minor gases were very slightly off in the v2.0 emissions dataset. Impact on ERF is negligible.

ssp370-lowNTCF modifications

A change in the ssp370-lowNTCF emissions to make them consistent with the AerChemMIP protocol has been implemented.

Two versions of ssp370-lowNTCF now exist:
ERF_ssp370-lowNTCF-aerchemmip_1750-2500.csv: the one to use
ERF_ssp370-lowNTCF-gidden_1750-2500.csv: retained for compatibility (based on harmonization in Gidden et al., 2019

A symbolic link has been added so that the ERF_ssp370-lowNTCF_1750-2500.csv file points to the AerChemMIP version. Scripts relying on this should continue to work, but the data has changed (requiring a new version).

RCMIP Phase 1 Final Version

This release is the effective radiative forcing from SSPs to be used in RCMIP Phase 1. It is consistent with the latest AR6 estimates between 1750 and 2014 (and CMIP6 forcing datasets, where possible, between 1850 and 2014), but for the following small differences:

• GHGs in this dataset are based on concentrations from Meinshausen et al. (2017), whereas AR6 ERF time series are based on more recent observations in AR6.
• stratospheric ozone and stratospheric water vapour ERF depend on GHG concentrations, and follow Meinshausen et al. (2017) here rather than AR6 values.

Other updates compared to v0.3.1:

• add "year" column for better integration with pandas (fix #2)
• volcanic ERF fixed to rebase to long-term average of zero (meaning positive offset for quiescent years) and ramp down over 10 years to zero from 2014 (partial fix of #3)
• ozone relationships based on CMIP6 datasets for historical period (fix #6)
• some columns have been given more descriptive names
• "present day" for forcing categories that are assessed in AR6 is defined to be year 2018 in SSP2-4.5. This differs from v0.3.1 which used SSP5-8.5 for year 2017. AR6 now defines 2018 for recent estimates, and my best estimate of transient future aerosol forcing is from SSP2-4.5 simulations which is why the change has been made. Impacts are minimal.

As a further note, ERFs are substantially higher than the nameplate RFs for year 2100. Much of this stems from an increase in ERF compared to stratospherically-adjusted RF (SARF) when taking into account rapid adjustments from greenhouse gases, and new relationships for SARF derived from Etminan et al., (2016). Taking this into account, modellers may wish to use the latest best-estimate doubled-CO2 ERF of 4.01 W/m2 in their setups where this is a free parameter and important (e.g. for calculating climate feedbacks, or ECS).

Zenodo citeable version

No changes since v0.3 except that the repository is now linked to Zenodo and creating a new release triggers the DOI.

Updated post-2100 forcing and land-use rescaling

The post-2100 period has been updated from new emissions files provided by Zeb Nicholls and should now be fully consistent with RCMIP (rcmip.org).

A slight rescaling of the land use ERF is also performed in order to make SSP5-8.5 land use exactly -0.15 W m-2 in 2017.

Rapid adjustments and extension to 2500

Rapid adjustments to CO2 and CH4 have been included from the multi-model study of Smith et al. (2018) : +5% for CO2 and -14% for CH4 (only considering models that include SW absorption of methane).

ERFs have been extended to 2500: for GHGs and Solar they are based on Meinshausen et al. (2017) and Matthes et al. (2017) respectively; emissions of aerosol and ozone pre-cursors are set to their 2100 values for 2100-2500.

First release

First release: contains the 13 forcing classifications from 1750-2100 included in the FaIR 1.3/1.4 series with aerosols split into aerosol-radiation and aerosol-cloud contributions.