CO2 emissions gridding

[znichollscr]

Not exactly harmonisation, but related enough.

@lchini @jkikstra @ssmithClimate @gidden good if you have at least seen this. @benmsanderson I am also assuming you will be interested and have thoughts. Please tag any others who should be in here.

@ssmithClimate kicked off the discussion with the following:

We had one additional dimension in the CO2 gridded emissions variable that corresponded to negative emissions (CO2 update) (see below). Is that fine for this time too?

e.g.: double sector(sector=9); :ids = "0: Agriculture; 1: Energy; 2: Industrial; 3: Transportation; 4: Residential, Commercial, Other; 5: Solvents production and application; 6: Waste; 7: International Shipping; 8: Negative CO2 Emissions";

Last time we simply took the negative CO2 emissions and distributed them over space with one static map of global cropland? Is that sufficient this time?

Do we need supplementary files that provide more information on fossil CO2 sequestration?

My two cents. I think the extra sector is fine.

To confirm, are we only talking about scenario gridding here or are we also discussing gridding of historical data? (I'm assuming the former. If it's the latter, my answers below won't make much sense.)

Re distribution, I guess it really depends what we think these emissions are. I think it would be nice to do better than a single static map, but obviously we're dealing with time pressure so there's limits to how ambitious we can be. I guess it may also be dependent on what information we can get from IAMs/other sources. My immediate question would be this: I would have thought that all land-based sequestration would come from the land cover changes. Hence, I would have thought that all negative emissions would be mapped to places where things like DACS are being done, which I would have assumed wasn't related to crop land.

Re a supplementary file. I think it would be helpful because there are real differences between different types of sequestration. Depending on how negative emissions end up being dealt with though, all of this information may be promoted to 'main' files so we might not end up with any supplementary files.

Having said all that, in my head, the way this would work is the following. It's pretty clear to me now that this isn't how it worked in CMIP6, so maybe I'm just thinking of something that is practically impossible, I look to everyone else for guidance on this. Anyway, to how I thought it would work:

• we get as much information as we can out of the IAMs. Ideally, that tells us not only how much negative CO2 is being assumed but also where it is
• we use that information to split the negative CO2 into different sectors. The reason being that how you grid DACS will be different from BECCS which will be different from ocean alkalisation (for example).
  • for 'engineering' solutions (which ESMs don't model), we create files that prescribe a CO2 flux to the ESMs
  • for 'land-based' solutions, I was assuming that we'd be creating land-use change files, and ultimately the ESMs would be the model that quantifies how much negative CO2 this change does or doesn't make
• so we'd end up with the negative CO2 emissions from IAMs being converted into a combination of gridded negative CO2 fluxes and gridded land use changes, then it's over to the ESMs from there
• (then there's the question of how we extend these gridded emissions to the period where there is no IAM modelling, but maybe one thing at a time)

[ssmithClimate]

Yes, not historical yet, since removals are insignificantly small. But we might add that additional dimension at some point if that becomes larger. (Note fossil Co2 re-injection, which is not small, is not net negative - so we don't worry about that historically other than that we assume that's taken account of in the Annex I inventories we calibrate to for fugitive CO2 emissions.)

More specifically, the negative emissions from BECCS, at large scale, would formally be from when the crops were grown that are subsequently transformed. (Thus the zero order approach of distributing as cropland.) It would also be potentially a global market, so hard to pin down where the BECCS occurred.

DACC in the future would be a point source removal, but the IAMs won't have data on where that point source was located.

[znichollscr]

More specifically, the negative emissions from BECCS, at large scale, would formally be from when the crops were grown that are subsequently transformed. (Thus the zero order approach of distributing as cropland.) It would also be potentially a global market, so hard to pin down where the BECCS occurred.

Ok. I guess my question is, does it make more sense to have these as negative emissions or just as land-use change? My instinct was as land-use change, leaving it up to the ESMs to decide how much this BECCS can actually sequester, but maybe I misunderstand.

DACC in the future would be a point source removal, but the IAMs won't have data on where that point source was located.

Interesting, I would have assumed that IAMs would give some sense of which region this occurs in, but maybe that's ambitious.

[ssmithClimate]

No, the BECCS isn't just land use change. Normally you grow the crop (takes up CO2) and that CO2 is respired or otherwise released back into the atmosphere within a year or so. So it's no net change. But with BECCS the CO2 gets taken up by crops and never comes back, so it's a net negative. But the uptake occurs at the field level.

They would know which region DACC occurs but not necessarily anything more specific than that.

[znichollscr]

Thanks @ssmithClimate, very clear and thanks for being patient with me who knows nothing :)

[ssmithClimate]

There are an infinite number of things of which each of us knows nothing....