1/ In the reporting on the National Climate Assessment, some articles (such as this by @CNN) have focused on the far end of the risk distribution and say that "the economy could lose ... more than 10% of its GDP" by 2100.
I'd like to push back a little.
edition.cnn.com/2018/11/23/hea…
I'd like to push back a little.
edition.cnn.com/2018/11/23/hea…
2/ The "10% of GDP" claim comes from this figure in Chapter 29: nca2018.globalchange.gov/chapter/29/
It compares the increase in global mean temperature to the projected US economic losses at 2100.
It compares the increase in global mean temperature to the projected US economic losses at 2100.
3/ So yes, the top end of the figure does exceed 10%.
However, the high estimate comes from RCP8.5 (red dots), an emissions scenario which assumes extremely aggressive growth of CO2 emissions.
While possible, I certainly wouldn't regard RCP8.5 as a likely future.
However, the high estimate comes from RCP8.5 (red dots), an emissions scenario which assumes extremely aggressive growth of CO2 emissions.
While possible, I certainly wouldn't regard RCP8.5 as a likely future.
4/ More importantly, a close examination of the GDP figure will notice that there are many red dots associated with RCP8.5. The distribution of red dots reflects an uncertainty in the climate's response to this much CO2, and covers a very large range from 4 F to 14 F at 2100.
5/ This presented range is huge. In fact, it is substantially larger than the response range presented elsewhere in the same report. In Chapter 2, nca2018.globalchange.gov/chapter/2/, a 90% likelihood range of ~4.5 to ~9.8 F is indicated for RCP8.5, with a most likely value of about 7 F.
6/ So, in order to get to a 10% loss of GDP, one has to choose the most extreme emission scenario and then one has to travel far out on the response uncertainty curve, beyond what one considers most of the time.
7/ That doesn't make the result wrong. Studying low probability outcomes is useful and necessary for risk managers. However, focusing on such extreme tail risks is not great for public science communication.
8/ When @RogerPielkeJr and others criticize the 10% GDP numbers as extreme, I think that is entirely reasonable. 10% losses are a vaguely possible outcome, but based on the current report such an outcome is unlikely and many things would have to go wrong to get there.
9/ If science communicators focus on such scenarios, the public can easily misunderstand these low probability risks as likely outcomes. Giving heavy weight to low probability outcomes can potentially make presenters seem to be alarmist.
10/ The emphasis on the "10% of GDP" appears to be coming mostly from the media (scary headline!). As far as I can see, the report itself barely even mentions this possible outcome aside from a point at the far end of one graphic.
11/ Most of the report actually focuses on substantially lower estimates, rather than 10% of GDP (nearly $2 trillion/yr).
For example, they present a $500 B/yr loss at 2090 due to the extreme RCP8.5, and "only" $280 B/yr if the more moderate RCP4.5 emission scenario is followed.
For example, they present a $500 B/yr loss at 2090 due to the extreme RCP8.5, and "only" $280 B/yr if the more moderate RCP4.5 emission scenario is followed.
12/ Potential annual costs to the USA due to climate change of hundreds of billions are still enormous. I don't see a need to exaggerate that further by focusing attention on the extreme tail risks. /end
