Thread by @RogerPielkeJr

A preview of updated “social cost of carbon” methods under the Biden Administration—> brookings.edu/wp-content/upl…

While key details are TBA, once again all of the heavy lifting is done by (newly-created) extreme scenarios & damages post-2100

Includes ~12C temp increase in 2300 🤨

The new RFF-SC scenarios (apparently not publicly available) have a median emissions trajectory similar to 3.4/4.5 SSPs, but include a ridiculously wide uncertainty range (from net-zero CO2 ~2050 to ~3x CO2 ~2100)

Even so, SSP5-8.5 is wildly implausible

Though RFF-SC details are not yet available, we can clearly see that the PDF for future emissions is heavily skewed

Of note the emissions distribution is centered on the median result & initial SCC results on the average result, thus increasing the influence of extreme scenarios

Two questions I’ll be asking of the new methodology

1. What fraction of SCC estimate comes from damages post-2100 (or >3C)?

2. What fraction of SCC estimate comes from uncertainty PDF of future emissions vs only using median value?

My expected answers for both = most/all

The illustrative approach in the new paper uses the damage function from DICE
You can clearly see below that almost all damage is post 2100 (>3 C)

Figure NAS: nap.edu/read/24651/cha…

Also from NAS, here are damages (DICE = blue) as a function of T

This explains why having projected Ts go as high as 12C by 2300 are important to a SCC estimate

But DICE is going to be replaced with a new damage function via CIL

And that one is mainly based on … RCP8.5

Tangled web weaved😎

So the new SCC approach will show undeniably that RCP8.5 is utterly implausible and at the same time base almost the entirety of the newly estimated SCC on a damage function based on RCP8.5, all of which will be obscured by massive, impenetrable complexity

/ENJOY
/END