Restarts for Geoffroy temperature function and GIR carbon cycle

An old feature given a new lease of life.

Updates to ozone forcing

This micro version introduces an updated treatment to the ozone forcing, which allows for a dependence on N2O concentrations, and a temperature-dependent ozone feedback (this was previously included with the stevenson treatment, but has now been simplified to be a linear function of temperature). The new treatment follows CMIP6 and AerChemMIP experiments:

Thornhill et al., 2021: https://acp.copernicus.org/articles/21/853/2021/acp-21-853-2021.pdf
Thornhill et al., accepted: https://acp.copernicus.org/preprints/acp-2019-1207/
Skeie et al., 2020: https://www.nature.com/articles/s41612-020-00131-0

Default coefficients that link the emissions to the CMIP6 model forcing in Skeie et al. are provided. In fair_scm() use tropO3_forcing=thornhill-skeie to use the new treatment.

Ozone forcing is not split into tropospheric and stratospheric components. For ease of communication with existing versions of FaIR, we maintain the forcing column indices, but all ozone forcing is put into the tropospheric bin (index 4 in 13-species FaIR, or index 31 in 45-species FaIR).

45-species FaIR

This version expands the multispecies version of FaIR to report 45 (previously 41) forcing species. Some fixes for running concentration driven runs with the 45-species FaIR have been implemented as well as including specified forcing and Geoffroy temperature options for inverse FaIR. scmdata interface has been removed.

Fix for 45-species concentration driven runs

Minor version incremented to reflect bug fix that previously used 13-species indices for some species when concentration-driven. All other functionality same as 1.6.0 series.

Compatibility with scmdata-0.7.1

No other changes, just plumbing

Delete rogue print statement

My bad.

v1.6 release candidate

Includes:

• Geoffroy two-layer temperature specification
• Reading of emissions files using the ScmDataFrame
• setup and communication with openscm-runner

Day to day operation unchanged from v1.5 - should be no surprises.

openscm-runner alpha version

This version should contain all the right stuff to run FaIR from openscm-runner.

openSCM testing

Inclusion of airborne fraction

Also testing this version with PyPI upload through openscm-runner

more plumbing for openscm-runner

Also graft the SSP data into the distribution.