Carbon cycle outputs

[maritsandstad]

We should get carbon cycle output of

• airborne fraction,
• and ability to calculate TCRE from 1pct CO2.

For those I need to have good definitions. Also, a wishlist of other carbon cycle outputs such as for instance contents of carbon pools (ocean, bio) also somewhat clearly defined and preferred units. Even things we can not get from the current model, but which we might want to implement later.

[maritsandstad]

Looked at the old RCMIP data output templates, and I guess these might provide some hints.

There is:

• Net Land to Atmosphere (Mt CO2/yr)
• Net Ocean to Atmosphere fluxes of CO2 (Mt CO2/yr) (also CH4, but we most definitely can't do that easily, for land there is also breakdown to feedbacks, and feedbacks other and permafrost, again, not something we can easily do).
• Cumulative for the same variables. (Mt CO2) Also total CO2 in carbon pools of
• Atmosphere (Mt CO2)
• Soil (Mt CO2)
• Detritus (Mt CO2)
• Plant (Mt CO2) (And I think in principle the pulse response function for biostuff that we have is vaguely aware of these in a very indirect way...)
• Net primary producticity (Mt CO2/yr (This is also something I think that our code in principle is vaguely aware of...)
• Airborne Fraction|CO2 (Dimensionless) (fraction of (cumulative) emitted CO2 which is still in the atmosphere - this is the def in the document for reference)
• Instantaneous TCRE (K / MtCO2) (warming per unit cumulative CO2 (this should simply be your 'Surface Air Temperature Change' divided by 'Cumulative Emissions|CO2') And speaking of Cumulative emissions, those for:
• CO2
• CO2|MAGICC AFOLU
• CO2|MAGICC Fossil and Industrial
• CO2|Other

From this, getting cumulative emission is no 1, and should be super easy to do (unless I'm missing something obvious), it should basically just be summing stuff, maybe even knowable from the inputs?

Then the instantaneous TCRE is basically for free from this definition, just need to actually add something that puts it in the inputs...

Then I wonder if airborne fraction is as easy as using a conversion factor between change in ppm and emitted CO2 and dividing that by the cumulative thing?

If it's this easy, then I can start implementing and outputting next week (modulo unexpected other things). But things are usually not always that easy, so I might be completely misunderstanding things...

The other stuff require a bit more thinking about what's in the carbon cycle model, and how to output it, but at least we can start thinking.