[owenversteeg]

> The CO2 impact of flying, and of flying first-class, is understood incredibly well

No, it's not! If you look at NGOs and charities, you find massively different numbers, even for flights with the same characteristics.

I have yet to find a decent model that takes into account everything: cargo, class, seat layouts, personal weight, etc etc. For example, Googling tells me every extra kg of weight burns 0.2-0.7kg of fuel per flight (naturally highly dependent on flight time, aircraft type, current weight etc etc.) If you take 3.15 g CO2/g fuel, and a 2.7x CO2e factor, that's 8.5 kg CO2e/kg fuel. So someone weighing 136 kg (several million Americans weigh more than this [0]) will cause up to 511.7 kg CO2e more emissions than someone weighing 50 kg will! That's more than half a ton of CO2e, and that's just in a simple body weight difference! Throw in a checked bag and a heavier carryon and you're well on your way to a ton of extra CO2e per person for longer flights.

The case of first class also depends strongly on airline-dependent factors: how many economy passengers a first class seat displaces, as well as the weight of the first class seat/bed and other things taken along to improve the journey for first class (blankets? food? a large wine selection? additional attendants, as you mentioned?)

It's an equation with an insane number of variables, as well as economic effects. Example:

There's a weekly (short) flight from <island> to <mainland> that's barely profitable, kept alive by a few passengers flying in first class. In this case, the impact of the first class is huge! Airport fees and taxes ensure that the profits from each additional economy passenger are low, but every first class passenger has a very high impact on the viability of the route.

Furthermore, like you said, there's radiative forcing caused by aviation, of which CO2 only contributes perhaps half. Contrails can contribute far more to global warming than CO2 emissions can, for a flight, depending on the weather. There are NOx emissions which have a cooling effect, which is based on altitude. So a short flight in warm weather in South America will have a much different emissions profile than a long-haul from Canada to Iceland.

I'm going to make a bold claim: the CO2e per passenger kilometer can vary by at least one order of magnitude depending on the flight characteristics (airline, weather, route, direction, location, aircraft, etc) - before you take class into account!

[barunio]

> No, it's not! If you look at NGOs and charities, you find massively different numbers, even for flights with the same characteristics.

Aside from those two sentences, nothing you've written contradicts anything I'm saying, and it's mostly orthogonal to the point.

It is obviously true that a lot of factors can affect the weight of an airplane, and the CO2e impact of a flight will vary based on that weight. And weight is one of many factors, as you mention at the end. But just because something is complicated and may seem overwhelming does not mean that the academic community engaged in researching the topic doesn't understand it.

It isn't possible to know a priori what the impact of a specific flight will be, and it isn't practical to determine the exact ratio of impact of a first-class to economy passenger on one specific flight. But nobody is making that claim, since that isn't how people think about systems in any large-scale domain like this. Instead, we look at large data sets and examine statistical relationships.

Your comment seems to be imply that because the system is high-variance with many variables, it is impossible to understand it in useful ways. That is clearly not the case. If it were, we wouldn't be unable to make decisions about virtually anything in the world outside of extremely narrow domains, and rational policy-making would be impossible. The number of variables in this domain is nothing compared to public health, for example. We can talk about the effectiveness of condoms in preventing infection even though there are a million variables affecting it with numerous outlier cases. In the same way, we can talk about the impact of flying first-class. (In fact, we can do better re: flying because there we have better analytical modeling in addition to the datasets.)

You claim that one would find massively different numbers among researchers. That hasn't been my experience as I've sought ought out detailed information online. Here are just a few references as a starting point:

  * 2008 Union of Concerned Scientists: https://www.ucsusa.org/sites/default/files/legacy/assets/documents/clean_vehicles/greentravel_report.pdf -- Chapter 2 is on air travel and very high-level, but Appendix B provides more detailed numbers across various parameters. It only considers CO2, not CO2e.

  * 2009 academic study, Oxford prof: https://www.eci.ox.ac.uk/research/energy/downloads/jardine09-carboninflights.pdf -- finds that on average, there's a 2x difference in per-passenger emissions between the most dense and least dense seating configurations.

  * 2013 World Bank study: http://documents.worldbank.org/curated/en/141851468168853188/pdf/WPS6471.pdf -- outlines analytical models that include the different variables you're mentioning (and many more), and examines variance in the results as those parameters are adjusted.

The results from all of these studies are consistent across broad parameters (in particular the impact of seating configuration). In fact, it's clear that my back-of-the-envelope calculation of a 69% increase due to first-class is pretty much the lower bound.

[Note, I said in my original answer that the impact of radiative forcing is still being understood, so I'm not talking about that. I intentionally said that CO2 impact is well understood at the start and called out CO2e as a distinct notion for that reason.]

> I'm going to make a bold claim: the CO2e per passenger kilometer can vary by at least one order of magnitude depending on the flight characteristics (airline, weather, route, direction, location, aircraft, etc) - before you take class into account!

That's not a bold claim at all; it's just irrelevant. A cross-country flight on an old plane with a couple layovers in bad weather that has to refuel unexpectedly and is only half-full will have an order of magnitude worse impact on the climate than the best case non-stop scenario. But if one asks, "what is the impact of flying first-class vs economy?", one is not asking "what is the impact of flying first-class in the best possible case vs flying economy in the worst possible case?".

[owenversteeg]

>We can talk about the effectiveness of condoms in preventing infection even though there are a million variables affecting it with numerous outlier cases.

Sure, but you're discussing something completely different from what I care about. Nobody is forced to not wear a condom for work, and you can't buy "condom credits" to offset not wearing a condom. The individual calculus of flying (and whether to fly based on emissions) is completely different from wearing or not wearing a condom.

Ultimately, I can only control my own flying and my own donations, and that's what I'm interested in here. I think I should've been more clear: what I'm looking for is something where I can, as a passenger, see the CO2e of my flight. If I fly DL 259 in economy, what will be the impact of this? If I have the option, does it make sense to replace this with (say) 3 round trip economy flights on KL 1385? (for example, one important conference vs 3 less important ones.) What will it cost to offset my flight?

I took a look at the 3 links you provided and I can't say I'm convinced. The UCS one has pretty pictures, but ignores CO2e which is inexcusable as that is where the bulk of the issue is. I don't see anything resembling an established model from those; it seems to be simply just people gathering (mostly) theoretical data and putting it into Excel. In particular, I am really not convinced that anyone has a good understanding of the factors involved in CO2e and from what I can tell the literature supports me on this.

Quoting one of your own links: different methodologies are responsible for a factor of 2 difference in CO2e and "there is as yet no internationally agreed and adopted methodology for the calculation of aviation emissions" (from Dr. Jardine's report)

>>> I'm going to make a bold claim: the CO2e per passenger kilometer can vary by at least one order of magnitude depending on the flight characteristics (airline, weather, route, direction, location, aircraft, etc) - before you take class into account!

>That's not a bold claim at all; it's just irrelevant.

It's absolutely a bold claim to 99% of people. Nearly everywhere, from large newspapers to "calculator" sites to social media, people do not realize that the emissions for a flight are very individual to that specific flight. If you asked random travelers, I bet nearly none would expect things like seating configuration or weather to strongly affect emissions. If you told some random people that by choosing a densely seated plane they could strongly reduce their emissions I bet they'd be interested!