Keeping the SCIENCE in Climate Science
There are so many reasons for urgency on climate change, why make up new ones?
I don’t know about you, dear readers, but I love it when I get to update my views, especially when I realise that I may have held them through inertia or received wisdom. This conversation with Michael Liebreich and Dr Roger Pielke taught me a lot and revealed some gaps in my understanding of climate modelling that I didn’t know I had. Pielke is an academic working at the nexus of science and policy and is the author of the Honest Broker blog.
It seems to be an unfortunate feature of any movement agitating for social change, that correction of the most outlandish claims in support of that movement are not corrected from within, but are left to those resisting the change. It is even more unfortunate when the movement seeks to silence and / or ostracise those from within who do push back against the extreme narratives. As such, I’m really delighted that Liebreich did this show. The topic is quite technical, so whilst I did my best to fact check, inevitably it was non-exhaustive. As always, I’m very happy to hear from readers so I - we all - can continue to update our views based on the best information.
The reaction of the climate community (i.e. flat rejection, calls for sacking) to the now infamous presentation by the (former) head of responsible investing at HSBC is revealing of the community’s instinct to silence heterodoxy. Although both Liebreich and Pielke agree that the presentation was tone-deaf and he was mistaken on several points, it did highlight that there are a number of inconsistencies within the climate narrative that need to be grappled with.
For example, within the IPCC work there is a “GDP paradox” – predicts that GDP will be higher, driving more emissions, but also that there will be large sections of the world that will be uninhabitable. There is also a physical infrastructure paradox – scenarios where coal plants are modelled as belching out CO2, but while also modelled as being underwater. The scenarios should be plausible and consistent.
There are also valid discussion points around balancing development and emissions. Whilst development tends to increase emissions, it drastically reduces the human impact of extreme weather, as evidenced by weather-related deaths declining by orders of magnitude over the last century, even as the global population has quadrupled. [Of course the worst effects of climate change lie ahead of us, so it would be optimistic to expect no reversal of this long-term trend.]
Where policy makers or institutions are using different scenarios to form policy or recommendation, there is very little transparency as to which scenarios they are using and why. (Pielke’s work on that and more here.)
Nobody in the IPCC is in charge of identifying which scenario is most plausible. Work done by Pielke and others suggests that the most plausible scenario currently is around 2.2 degrees of warming, not 4 degrees or higher.
The IPCC persists in using RCP 8.5 (SSP5 8.5) as a high-case baseline, even though it is no longer credible given technology advances since it was first conceived. (Note: “8.5” refers to 8.5 watts of radiative forcing per m^2, not 8.5 degrees of warming or anything else.) Work has shown that arriving at an 8.5 scenario wildly implausible, e.g. would include things like coal-to-liquids. More on the origins of SSP5 8.5 are detailed in this comprehensive Carbon Brief article - the forcing scenario was created independently of a plausible pathway to get there. Although that article does cite scientists that say that, whilst unlikely, the climatic effects envisioned under that scenario are not impossible if the climate system is more sensitive or there are major feedback loops, such as methane release from melting permafrost.
Attachment to the extreme warming scenario is partially driven by the fact that it is easier to do science with an extreme scenario. Additionally, hundreds of papers have already been published using RCP 8.5 as a ‘business as usual scenario’, so there are interests in defending it. To take one terrible example, a study was funded (but not disclosed) by McKinsey that suggested that RCP 8.5 actually was plausible, and then used it to justify a bunch of previous work they had done using RCP 8.5. [Boo! Hiss!]
There is a popular perception that there is an increase in all extreme weather events and that these are driven by human-induced climate change.
In fact, from 1990 till today, there has been no increase in land falling hurricanes in the US. From 2006-2017 there was a notable absence of hurricanes making landfall onto the continental United States (a historical anomaly), so there is a perception by younger people that there has been a sudden and unusual increase.
According to Pielke, using Normalised Damage – adjusting damage from past hurricanes for population change and increasing wealth – finds no increase in the destructive power of storms. We’re just building more stuff in their paths. [Check out Pielke’s blog on that here.] However, there is some work using Area of Total Destruction that finds there is an increase of storms, which is more pronounced for the most destructive storms.
It is difficult going from the extreme weather event to the impact, because the event needs to intersect with people, and vulnerable people / infrastructure in particular.
It isn’t right to talk about extreme weather as a single group. Each thing is a separate phenomenon. IPCC breaks down the assessment of these phenomena by:
Observation: has there been a statistically significant change observed over 30 years?
Attribution: is that change linked to human activity?
Projection: how is it predicted to evolve with higher concentration of GHGs?
These are detailed in the IPCC Working Group 1 report, Chapter 11. The actual text is very long, complex and, frankly, boring, so you can understand why the media and policymakers end up confused. Even the condensed exact wording from the report was too long to include here, but I have to hand if anyone is curious to see it (just ping me). Relative to what it says in the IPCC report, Pielke comes off as slightly dismissive as to the importance of some of these as climate change impacts.
Temperature Extremes: Observation: Very clear increases in hot extremes and decreases in cold extremes (see below chart); Attribution: Extremely likely that increases in temperature extremes are human-induced; Projection: Virtually certain the heat extremes will intensify over the 21st century
Extreme Precipitation: Observation: Likely that frequency of intensity of extreme precipitation has increased; Attribution: Likely that most of the observed increase is human-induced; Projection: Increase will increase non-linearly with temperature increase, especially for rare events, but the IPCC report doesn’t indicate what to expect with smaller increases, high-confidence at 4 degrees.
Floods: Observation: Low confidence of increases on a global scale although there has been some increase in some regions [note that the IPCC take river flow as a proxy for floods, because difficult to gather data on all varieties of flood - common sense would suggest that stronger observation of extreme precipitation might indicate more floods which aren’t detectable by river flow alone]; Attribution: Low confidence of human-induced changes in peak river flows; Projection: Medium confidence that, on balance, more areas will be affected by river floods.
Droughts: [A complicated one due to the different factors] Observation: High confidence in increased water stress at times of low precipitation due to accelerated evaporation of groundwater, particularly during dry seasons, medium confidence of some regional increase in precipitation deficits (parts of Africa and South America); Attribution: Medium confidence that human activity has contributed to accelerated evaporation, medium confidence that it has contributed to some specific instances of meteorological drought (lack of rainfall), but not a general increase in meteorological drought; Projection: High confidence that climate change will increase instances of agricultural and ecological droughts with some areas being affected even with modest warming (1.5 degrees).
Extreme Storms: Observation: Whilst there hasn’t been an increase in the number of storms, it is likely that a greater proportion of those storms are category 3-5 and that the instances of ‘rapid intensification’ have increased; Attribution: It is very likely that some regional increase in frequency is due to human activity, but from aerosols (particles in the atmosphere), not from greenhouse gases; there is limited evidence that humans have caused an general intensification of storms, but it is highly likely that climate change has increased the amount of rainfall in storms; Projection: Whilst the total number of storms isn’t expected to increase, it is very likely that peak wind speeds will increase and the proportion of category 4-5 storms will increase and that average rain rates will increase.
On the science of attribution for specific events: it is done by modelling how the world would be if climate change hadn’t been occurring, then run a separate model with climate change, and then see the different risk for an event. But the IPCC is cautious on it, science is very early. Sometimes we read that a given event was “30x more likely” because of climate change – if that was the case, we should see 30 times more of that event in realised data.
The Galveston hurricane of 1900 killed over 10,000 people and flattened the island, 1000's of homes destroyed but cost minimal since houses cost less than $500 each back then...