Thread by @rustneversleepz

*Much*-abbreviated sister🧵to 100-tweet monster, so the highlights aren't lost!

•*Noted* climate physicist personally dives into cost-benefit Integrated Assessment Models (IAM's) - DICE and others.

•Writes own simplified version! Derives *intriguing* - provocative? - results!

The physicist is Myles Allen, and the investigation is detailed in a short 4-page 2016 paper, and a 2015 lecture (⬋, ⬊) about the same, pre-publication.

Links to the lecture and paper are in this tweet.

In the monster🧵, I...
podcasts.ox.ac.uk/how-hot-will-i…
nature.com/articles/nclim…

... spend much more time detailing why Myles Allen undertakes this investigation and how he goes about it.

And I *strongly* encourage people to watch the lecture to hear him describe his motivations and game plan in his own words.

(link to monster-🧵👇)

https://twitter.com/rustneversleepz/status/1562566447225643009?t=AOuovCC15v5IpHtlhuCpPQ&s=19

Because I think there's a real lesson for people there in the way Myles rolls up his sleeves, rather than simply throwing up his arms, or playing a tedious telephone-tag game of casually repeating CBA-IAM critiques they've encountered, without really understanding the models.

In any event, what prompts him to investigate (remember, this is 2015) is an intuition he has that the modules in the CBA-IAM’s used to convert emissions to temperatures were "sluggish" because they weren't using the 2008-14 emerged scientific consensus that "Transient Climate...

... Response to Cumulative Emissions" (TCRE) was now the agreed way to estimate this. (Images above👆from the IPCC AR5 WGI, 2014.)

Reasoning that delayed temperature response to emissions would propogate errors through the IAM's via damage function to Social Cost of Carbon to...

... mitigation response, he wants to test the sensitivity of the models to his preferred TCRE methodology.

To his credit, Myles Allen decides to build his own simplified IAM to test this intuition.

He finds the IAM's *do* indeed underrepresent the speed of the TCRE response...

... *BUT*, to his initial surprise, this doesn't have the relative knock-on impact he'd expected on the damage functions, SCC, etc. - the IAM's seemed "less sensitive to climate sensitivity" than he'd expected a priori!🤔

In the sister-🧵, I drill more into why Myles finds...

... this muted in his own model, but I gather this particular result then further piqued Myles Allen's curiousity, so he set out on a "murder mystery" (his phrase) to use his own new mini-IAM to investigate what exactly these CBA-IAM's *ARE* sensitive to (or not! or if they're...

... even sensitive to anything at all!🤔).

Unsurprisingly, he finds/confirms his model is - as other IAM's are well understood to be - *very* sensitive to the specification of the function converting *temperatures* to damages, as well as to the choice of discount rates used...

... (although, notably, Myles chooses a relatively low (👍) 1.5% discount rate for a "pure rate of time preference" (prtp) as the base case for most of his sensitivity runs).

But *then* Myles Allen turns his sensitivity analysis attention towards two *other* variables in his...

... model which are "policy relevant":

• the *marginal* price at which a "backstop" abatement technology is deployed which allows us to to eliminate the *marginal* emissions that get us to net *𝙯𝙚𝙧𝙤* emissions;
• geometric avg economic growth *between now and net zero*.

I discuss the relevance of the concept and important role of both a "backstop abatement technology" and its *marginal* price at much greater length in the sister-🧵(as does Myles in the lecture/paper).

But to keep *this* 🧵brief, note ⓐ that the carbon/climate dilemma...

... requires that we eliminate *entirely* our net carbon emissions, so that cumulative emissions do not ultimately stop increasing until we are deploying decarbonization technologies that can eliminate the *last* emissions to be abated, largely irrespective of how vast and cheap

... earlier mitigation occurs; and ⓑ projected *marginal* abatement costs can rise *very* steeply as we move towards "harder to abate" application emissions *𝙪𝙣𝙡𝙚𝙨𝙨* some "backstop" abatement technology(s) emerges that can eliminate much of the remaining emissions with...

... flat or only gradually rising *marginal* costs as it deployed across applications.

(again, more about the important concept in the sister-🧵)

So, to examine these sensitivities, Myles Allen sets constant model various parameters (below), varies economic growth rate...

... and solves for what the implied marginal backstop technology price would need to be in order to achieve various peak temperature targets (in 0.1°C increments) (and hence, peak cumulative emissions as well).

The fixed parameters (2015) are:

•Current real GDP: $75 trillion

•Current temp anomaly: 0.9°C
•"Growth-corrected" discount rate: 1.5%
•Damage coefficient "D" on ΔT: 0.00267
•Damage exponent "γ" on ΔT: 2
•a TCRE value of 0.55°C/1,000 GtCO₂
•a constant fraction used approximate "average" *annual* emissions after model initiation: 0.75

(more on all these👆in sister-🧵)

Two more *very* noteworthy conditions:
•GDP growth rates *only matter for the period from present to the point of net-zero emissions* - not afterwards!
•*NO* temp declines modeled after peak temps - i.e. assumes no atm CO₂ drawdowns by CDR!

Running his model on those👆 parameters yields these plots👇, of peak temps vs marginal abatement cost overlaid with isoquants of "until net-zero" geometric avg economic growth in 0.5% increments from 1.0% to 3.5%.

Before examining these results in detail, recall again that *climate physicist* Myles Allen *initially* undertook his investigation to see how sensitive the CBA-IAM's might be to variations in the emissions→temp module, *𝙣𝙤𝙩* to economic growth and backstop technology cost!

I find this a *very* illuminating, clarifying presentation of his results.

What do they suggest?

First, unsurprisingly, all else equal, lower levels of peak warming are associated with lower costs for the backstop technology.

Secondly, however, assuming backstop abatement...

... measures are available and deployed when socially cost-effective, lower levels of peak warming are associated with 𝙝𝙞𝙜𝙝𝙚𝙧 levels of economic growth.

A further implication: since peak temps are a function of *cumulative* emissions to the point zero-emissions is that...

... *𝙝𝙞𝙜𝙝𝙚𝙧* rates of economic growth would yield *𝙡𝙤𝙬𝙚𝙧* cumulative emissions.

*Some* readers at this point will find this second part of the result *also* unsurprising.

But if I know my #climatetwitter, *many* will instinctively balk at it (or worse!).

So, what...

... is driving this outcome?

Again, I dig in to this in considerably more detail on sister-🧵, but essentially:

• The damage function for a given temp is higher given greater incomes - primarily because more consumption is at risk of damage. Which means we are more willing...

[still some tweets even on the "short thread"... but Robbie needs a walk...]

[Robbie👇]