This is a post that I’ve been thinking of writing, but have been somewhat reluctant to start. It’s mostly because it relates to something said by Roger Pielke Jr, so could end up being a bit of same ‘ol same ‘ol, which I’ve become rather tired of. However, it’s an interesting issue, and I’m not entirely sure that I’m right, so I’ll have a go.
As part of Roger’s attempt to withdraw from the public climate debate, he publishes a monthly newsletter. In his most recent one, he discusses extreme weather and climate change and says
Research keeps accumulating that shows that so far at least, the rising costs of weather disasters are not a result of weather extremes that have become more common or intense due to climate changes resulting from the emission of greenhouse gases
Now, I think the above is probably strictly correct. We probably have not been able to definitively demonstrate that the rising cost of weather disasters can be formally attributed to the human emission of greenhouse gases. However, I think this is mostly the wrong way to consider this, and I’ll try to explain why.
Let’s start at the beginning. We’re emitting greenhouse gases into the atmosphere. These greenhouse gases accumulate in the atmosphere. Their increase reduces the outgoing longwavelength flux, producing a planetary energy imbalance. This means that energy accrues in the climate system, causing it to warm.
Now, even if we do observe warming, we can’t necessarily claim that it is due to our emission of greenhouse gases. However, as this Realclimate post explains, we can do attribution studies to determine the most likely causes. However, it’s not as simple as observating that it is indeed warming; it requires some kind of model and also requires considering how what we expect from the different possible causes compares to the actual observations. In particular, the different possible causes produce different spatial, and temporal, patterns of warming, which can then be used to determine the most likely cause of the observed warming. Having now done this type of analysis, we are pretty confident that most of the observed warming is anthropogenic.
What are the consequences of this anthropogenically-driven warming? We obviously expect an increase in the frequency and intensity of extreme heatwaves. We also expect the hydrological cycle to intensify, which essentially means more evaporation and, consequently, more precipitation. In particular, we expect an increase in the frequency and intensity of the most extreme precipitation events. We also expect an increase in the frequency and intensity of some of the more extreme weather events, such as tropical cyclones.
Can we actually formally attribute changes in some of these extreme events to anthropogenic emissions? I think this is actually quite tricky, partly because the events are extreme and, therefore, rare. We don’t therefore have enough suitable data to make formal attribution claims (I may be wrong about this, so happy to be corrected). However, I think we do have evidence for an increase in the frequency and intensity of extreme precipitation events and heatwaves. We also have evidence for an association between increased sea surface temperatures and an increase in the frequency and intensity of tropical cyclones (and an understanding of why). So, the observations appear consistent with what would be expected.
In some cases, climatic events do damage that we can quantify in terms of how much this would cost to repair. Typically, these are the rare extreme events, not all of which do damage, which makes the events that do damage even rarer. Hence, formally attributing this to anthropogenic emissions is very difficult. Furthermore, there are many factors that could influence the scale of this damage. It could increase partly because there are now more valuable assets in the way. On the other hand, it could be lower than it might otherwise be because the infrastructure is now more able to withstand the impact of these events.
However, does this really matter? Do we need a formal attribution study to be quite confident that an increase in the frequency and intensity of extreme precipitation events will likely impact flooding? That increased evaporation and warmer weather will likely impact droughts and wildfires? That increased sea surface temperatures will intensify extreme tropical cyclones and make these more damaging if they impact populated coastal areas?
I don’t really think so; these don’t seem to be a particularly controversial inferences (they probably are, but it’s not clear why). Of course, that we expect our emission of greenhouse gases into the atmosphere to ultimately influence the cost of weather-related disasters doesn’t immediately tell us what we should do. However, recognising this seems more relevant than highlighting that we haven’t yet formally attributed increases in weather related damage costs to our emissions of greenhouse gases, especially since the latter could be true even if anthropogenically-driven warming has indeed lead to an increase in the cost of weather related disasters.
...and Then There's Physics 
I don’t understand this. You seem to be equivocating about whether AGW is responsible for the observed increase in extreme weather events… isn’t this established fact by now?
Speaking of extreme weather events…,
It’s official: 2017 was the costliest year on record for natural disasters in the United States, with a price tag of at least $306 billion.
The National Oceanic and Atmospheric Administration, which tracks billion-dollar disasters, reported in January that the record total came from 16 separate events with damages exceeding $1 billion.
Megadisasters devastated America in 2017. And they’re only going to get worse. by Umair Irfan & Brian Resnick, Energy & Environment, Vox, Mar 26, 2018
Fischer+Knutti (2016): “heavy precipitation intensification is now emerging in the observed record across many regions of the world, confirming both theory and model predictions made decades ago”
https://www.nature.com/articles/nclimate3110
I’m of the opinion that extreme events should NOT be measures in human terms (per GDP, people killed, buildings/infrastructures damaged). I think extreme events MUST be measured with environmental metrics that are not anthropocentric (like temperature, precipitation, humidity, wind speed, sea level rise, wave heights, etceteras).
RPJ says: “Research keeps accumulating that shows that so far at least, the rising costs of weather disasters are not a result of weather extremes that have become more common or intense due to climate changes resulting from the emission of greenhouse gases.”
And ATTP says: “Now, I think the above is probably strictly correct.”
I doubt the above is strictly correct: research failing to show X is not the same as research showing not-X (and single papers concluding not-X on the basis of underpowered or premature analyses are not the same as an accumulating body of evidence showing not-X). Can you point to any research conclusively showing that rising costs are not due to AGW?
ATTP this is a rats nest. Say, for example, Tycoon X, has his little private island destroyed by a
hurricane, damge is 50M. He decides not to rebuild. A more intense hurrican comes along and destroys nothing, damage zero. Suppose he fortifies, and it costs, 100M. next hurrincane causes zero damage. what metric captures the impact of the worsing weather?. Suppose it wipes him out again, what is the right metric.? you need more than a climate model to ask the right “what if” questions.
I like
“I’m of the opinion that extreme events should NOT be measures in human terms (per GDP, people killed, buildings/infrastructures damaged). I think extreme events MUST be measured with environmental metrics that are not anthropocentric (like temperature, precipitation, humidity, wind speed, sea level rise, wave heights, etceteras).”
But for policy we do want to cash out the impact of worsening weather and untangling that from continued growth in vulnerable areas isn’t straight forward.
For some viewpoints from risk assessors serving the insurance industry, check out some pages from AIR Worldwide:
* http://www.air-worldwide.com/Models/Climate-Change/
* https://www.verisk.com/csr/statement-on-climate-change/
* http://w3.air-worldwide.com/AIR-ABI-Climate-Change-and-UK-Windstorms-Research-Paper
* http://www.air-worldwide.com/Publications/Infographics/Rising-Waters–The-Growing-Need-for-Proactive-Flood-Risk-Management-in-the-UK/
* http://www.air-worldwide.com/Blog/How-Will-Climate-Change-Impact-Extreme-Weather-/
* http://www.air-worldwide.com/Blog/Climate-Change-and-U-S–Hurricane-Landfall-Probability/
* http://www.air-worldwide.com/Blog/Climate-Variability-vs–Climate-Change%E2%80%94What%E2%80%99s-the-Difference-/
These folks are actuaries, and, so, pretty good statisticians. Given a bevy of climate scientists or these people, I’d trust them more.
Munich Re has similar assessments.
No further comment.
This has been a topical debate in Australia in the last couple of weeks triggered by a late season bushfire (in a lengthening fire season) that devastated a NSW South Coast town. Malcolm Turnbull, PM of Australia claimed in response to the Greens that “You can’t attribute any particular event, whether it’s a flood or fire or a drought or a storm — to climate change,”
There was almost immediate push back from climate scientists including Dr Sarah Perkins-Kirkpatrick from the University of New South Wales and Dr Andrew King of the University of Melbourne.
e.g. Dr King
>Disappointing to see @TurnbullMalcolm not up-to-date on attribution science. Yes, we can’t say climate change caused these events but we can look at how it’s changing their intensity and frequency.
This is a very good article which teases out some of the issues with attribution science in a discussion with the two.
http://www.abc.net.au/news/2018-03-21/does-climate-change-cause-extreme-weather/9566168
And of course there is the annual issue of BAMS, the Bulletin of the American Meteorological Society which focuses on attribution science, a series that started in 2011. So it is not new although still a developing science.
what about this stuff https://www.theguardian.com/environment/2018/mar/21/flooding-and-heavy-rains-rise-50-worldwide-in-a-decade-figures-show
pedantry,
I don’t know if this is indeed true, or not, but certainly the increases are entirely consistent with what we would expect. Personally, I do think that AGW has resulted in an increase in extreme events.
However, I think it is probably true that we haven’t yet attributed increases in damage costs to AGW. However, I don’t really think that is the right way to consider this. A intensified tropical cyclone that encounters a populated area is almost certainly going to do more damage than one that is less intense.
Steven,
Indeed, which is why I think that formal attribution is difficult because you need model that takes all these factors into account.
I completely agree with the latter part of your comment. We should be focusing on the physical properties of these events. That will tell us whether or not they’re being impacted by AGW (as many are). We can then decide the implications of such changes.
[Edit: Sorry, I didn’t realise that the latter part of Steven’s comment was actually a quote from EFS. So, yes, I agree with EFS that we should really be assessing extreme event using changes in their physical properties, not whether or not there is some indication of an increases in damage/costs.]
Leto,
You have slightly caught me out there. Indeed, I was giving Roger the benefit of the doubt. What is probably true is that we have not yet managed to formally attribute increases in costs to anthropogenic emissions. As you say, I do not think there is any evidence that our emissions have had no impact (I did somewhat imply that at the end of the post). In fact, one of the papers he uses to support what he asserted, doesn’t appear to say what he suggested. However, I’m somewhat tired of pointing out that Roger’s source doesn’t say what he claims it says.
JH and Brigitte,
Thanks, I hadn’t seen those articles. In Roger’s post he complains about some group of climate scientists always wanting to associate damage from extreme events with AGW, and then links to an article. The article itself, however, seems fine to me. It simply discusses how various factors can influence these events.
Chubbs,
Thanks, I hadn’t seen that article. Something I realised after writing this (and I should probably have mentioned this in my response to pedantry) is that I think there is more and more work being done in attribution and there are stronger and stronger indications that we are indeed influenced these extreme events.
hyper,
Thanks, those are interesting links.
What’s the goal?
Is it to convince doubters that climate change is really, really, happening, because all the physics to this point isn’t sufficient? A fool’s game.
I completely agree that damage costs are a very poor metric, but characterizing the physical variables is a valuable scientific endeavor. Will we get to the point of more localized predictions (of climate, not weather) through sophisticated statistical analysis soon enough to guide adaptation? I don’t know; perhaps changes will be driven by other considerations before that happens.
hyper beat me too it with the insurance links. The people who have to put their money where their mouth is disagree with Roger.
From the Insurance Information Institute
World natural catastrophes by overall and insured losses, 1980–2017. Quite spiky as you’d expect with a few large first-world losses dominating. But a clear upward trend. Interestingly their customers also seem to disagree with Roger, with a distinct increase in insured as a proportion of total losses. More people buying insurance? People buying insurance with larger payout limits? Or unchanged limits being breached more often?
But maybe it’s all a function of GDP growth. So how about the Number of world natural catastrophes, 1980-2017? With a very low threshold (one fatality or $100k – $3M, depending on country wealth) which gets round the statistics-of-infrequent-events problem. Nice upward trend, tripled since 1980. And, oh look, there’s even a bit of a dip during the “pause”. And it took off like a rocket in the 2010s. Does that remind you of anything?
World weather-related natural catastrophes by overall and insured losses, 1980-2017? Spiky again but upwards.
World weather-related natural catastrophes by peril, 1980-2017? Upwards again, tripled since 1980, driven particularly by hydrological events although droughts, heatwaves and forest fires have the biggest proportional increase.
And there’s a convenient archive page which shows a similar but less obvious pattern in data available in 2014. The big recent surge says that any 2014 paper (by RP Jr or others) is out of date.
Everett F Sargent says: March 27, 2018 at 11:06 pm
“I’m of the opinion that extreme events should NOT be measures in human terms (per GDP, people killed, buildings/infrastructures damaged). I think extreme events MUST be measured with environmental metrics that are not anthropocentric (like temperature, precipitation, humidity, wind speed, sea level rise, wave heights, etceteras).”
–
Hear, hear.
well expressed, sensible, commonsense approach.
[sorry to agree with you Everett, it will probably make you want to change your mind].
–
Steven points out one reason for not using human wealth as a measure of disaster.
The other is also simple. Assets multiply in value by both inflation and population growth.
Dollar value is no indicator to severity over time.
“What are the consequences of this anthropogenically-driven warming?
We obviously expect an increase in the frequency and intensity of extreme heatwaves.”
Not sure why the frequency should go up, the heatwave is an increase on the background heat which will be higher but this does not presage frequency, merely intensity.
A heatwave in cold environments would not be unwelcome to people.
–
” We also expect the hydrological cycle to intensify, which essentially means more evaporation and, consequently, more precipitation.”
Yes. More water could be good in some places. Obviously dry areas move polewards with the Hadley cells.
–
“In particular, we expect an increase in the frequency and intensity of the most extreme precipitation events.”
If there is more water in the air logically there should be more rain. Intensity against the old baseline? The more extreme an event is classed as the less often it can occur. Frequency by definition would have to stay the same.
–
” We also expect an increase in the frequency and intensity of some of the more extreme weather events, such as tropical cyclones”.
Treading on questionable grounds there. I thought that an increase in temperature is expected to reduce the frequency tropical cyclones but make them more likely, when they occur, to be more intense? Sure some of the weather guys here can correct me if wrong.
Stefan Rahmstorf says we must reduce emissions 100% in 20 years to stay below 2° C.
James Hansen says 2° C = DISASTER
Kevin Anderson says we have a 5% chance of staying below 2° C.
Collapse Data Cheat Sheet
[A note to bypass the automagical code. – W] https://www.reddit.com/r/collapse/comments/311m7d/collapse_data_cheat_sheet/
In 2017, the myth of powering the world with 100% renewables has started to crack
http://energyforhumanity.org/en/climate-energy/2017-myth-powering-world-100-renewables-started-crack/
EUROPE GETS 60% OF ITS “RENEWABLE ENERGY” BY BURNING TREES
https://www.newscientist.com/article/2114993-europes-green-energy-policy-is-a-disaster-for-the-environment/
The EU is emitting way more greenhouse gases than it says
https://qz.com/528491/the-eu-is-emitting-way-more-greenhouse-gases-than-it-says/
UC Davis Peer Reviewed Study: It Will Take 131 Years to Replace Oil with Alternatives (Malyshkina, 2010)
http://energyforhumanity.org/en/climate-energy/2017-myth-powering-world-100-renewables-started-crack/
At this rate, it’s going to take nearly 400 years to transform the energy system
https://www.technologyreview.com/s/610457/at-this-rate-its-going-to-take-nearly-400-years-to-transform-the-energy-system/
University of Chicago Peer Reviewed Study: predicts world economy unlikely to stop relying on fossil fuels (Covert, 2016)
https://www.aeaweb.org/articles?id=10.1257/jep.30.1.117
Solar and Wind produced less than one percent of total world energy in 2016 – IEA WEO 2017
Click to access KeyWorld2017.pdf
Fossil Fuel Share of Global Energy since 1990 – BP 2017
https://imgur.com/k7VecMq
Renewable energy ‘simply won’t work’: Top Google engineers
http://www.theregister.co.uk/2014/11/21/renewable_energy_simply_wont_work_google_renewables_engineers/
Top scientists show why powering US using 100 percent renewable energy is a delusional fantasy
http://energyskeptic.com/2017/big-fight-21-top-scientists-show-why-jacobson-and-delucchis-renewable-scheme-is-a-delusional-fantasy/
IEA Sees No Peak Oil Demand ‘Any Time Soon’
https://www.wsj.com/articles/iea-sees-no-peak-oil-demand-any-time-soon-1488816002
The Curse of Energy Efficiency
https://thetyee.ca/Opinion/2018/02/26/Energy-Efficiency-Curse/
No Soil & Water Before 100% Renewable Energy
https://lokisrevengeblog.wordpress.com/2016/01/24/no-soil-water-before-100-renwable-energy/
Vaclav Smil: “The great hope for a quick and sweeping transition to renewable energy is wishful thinking”
https://blogs.scientificamerican.com/the-curious-wavefunction/vaclav-smil-e2809cthe-great-hope-for-a-quick-and-sweeping-transition-to-renewable-energy-is-wishful-thinkinge2809d/
The Long Slow Rise of Solar and Wind – Vaclav Smil
Click to access scientificamerican0114-521.pdf
Is Renewable Energy Renewable?
https://ozziezehner.com/2013/04/03/is-renewable-energy-renewable/
Global Energy Demand & Carbon Emissions Increase In 2017
World energy consumption
Humans are sleepwalking into a mass extinction of species not seen since the demise of the dinosaurs | The London Economic
https://www.thelondoneconomic.com/news/humans-are-sleepwalking-into-a-mass-extinction-of-species-not-seen-since-the-demise-of-the-dinosaurs/23/03/
Black Carbon Aerosols Cause Global Dimming But Overall Warming – Paul Beckwith 15 min
angech,
Yes, this is correct. We may see a reduction in the number of tropical cyclones, but an increase in the frequency and intensity of the strongest tropical cyclones.
Everett says:
I’m of the opinion that extreme events should NOT be measures in human terms (per GDP, people killed, buildings/infrastructures damaged). I think extreme events MUST be measured with environmental metrics that are not anthropocentric (like temperature, precipitation, humidity, wind speed, sea level rise, wave heights, etceteras).
It’s already been repeated a couple of times, but I think it’s worth repeating again. Also interesting that I agree with both Steven AND angech w/r/t Everett’s comment.
Prolly a first.
angech says:
The other is also simple. Assets multiply in value by both inflation and population growth.
Dollar value is no indicator to severity over time.
This relates to an issue that I brought with RPJr. a number of times (back in the day), with no response. Everett’s comment notwithstanding, it is entirely possible, say, for damages as a % of GDP to stabilize or decrease even as damages in absolute values increase – and in particular damages such as deaths. It’s also possible for there to be patterns in outcomes such as a decrease in deaths concurrent with an increase in deaths or other forms among suffering, in disproportion, among some people as compared to others.
In short, damages as a function of GDP may be a useful metric in some ways, but to focus on it so heavily, as RPJr. does, it seems to me, fits in with his larger plausibility deniability gambit, that he uses to personalize the public debate and play the victim and cloak his stealth advocacy.
> We should be focusing on the physical properties of these events.
Here’s one recent study:
https://www.sciencedirect.com/science/article/pii/S0921800916309752
Adding “and coastal development” somehow suffices to contain decades of Junior’s efforts to monopolize attention.
angech –
A heatwave in cold environments would not be unwelcome to people.
What do you gain from such comments that so lack specificity that they are completely worthless (with respect to a reasoned discussion – they may have value in winding people up, I suppose)?
Perhaps not, but under climate change the expectation that temperatures drop as magnitude of latitude increases is often turned on it’s head. I don’t think people in the far North would dig a month of 80 degree temperatures, as comfortable as that sounds. There’s infrastructure which just isn’t used to it. And which needs to tolerate much colder as well.
Related to angech’s vapid comment:
https://www.vox.com/energy-and-environment/2018/1/17/16851398/cold-snap-heat-wave-deaths-human-health
> damages as a function of GDP may be a useful metric in some ways,
It drowns the signal into GRRRRROWTH noise.
The basic idea is that we can’t directly correlate extreme events damage loss without taking into account that there’s more coastal properties than before. Normalizing is not too obvious either: as Eli is wont to recall, better building codes help mitigate the damage costs:
https://ascelibrary.org/doi/abs/10.1061/AJRUA6.0000947
Even if we accept Junior’s shenanigans, we’re still stuck with the problem of dealing with properties vulnerable to the whims of very powerful elements:
https://www.sciencedirect.com/science/article/pii/S0095069616302315
In other words, our insurance policies will be established whether or not extreme events are caused by AGW, or if we can mitigate some of the risks. The risks that matter are independent from Junior’s vainglorious triomphalism.
All this reminds me an episode between Sam Harris and security expert Bruce Schneier on airport profiling:
https://twitter.com/kjhealy/status/978816406803369984
A security system needs to be simple, otherwise it’s riskier, therefore more expensive. SamH suggests something that may make intuitive sense (we know where the terrorist threat comes from), but does not make any sense from a policymaker’s perspective (we need explicit TSA guidelines), at least insofar as our policymaker is interested in doing something, not just in throwing snowballs in Congress.
Willard –
better building codes help mitigate the damage costs:
RPJr. does claim evidence shows otherwise. I thought his argument in that regard not conclusive, but eventually he stopped engaging with me on the topic. .
Junior does claim all kinds of things depending upon the audience, Joshua.
Let me remind you of how NG dealt the Junior-JamesA ClimateBall episode:
https://www.sciencedirect.com/science/article/pii/S0095069616302315
Make sure to read back JamesA’s posts. If you want the spoiler:
http://julesandjames.blogspot.com/2011/08/roger-rabbit.html
Let’s hope Junior only plays Poker with fungible money.
I’m not sure how Junior could survive posturing himself as some kind of IPCC metacritic while failing to grasp the main modality by which the IPCC bundles its deliverable. Yet here he is, again and again.
“These folks are actuaries, and, so, pretty good statisticians. Given a bevy of climate scientists or these people, I’d trust them more.”
they find zippo about hurricanes.
“We consider policy options such as climate change abatement, policies which encourage adaptation, reducing subsidies for coastal development, and disaster relief policies, which distort insurance markets. ”
sound like frickin libertarian luckwarmers.
kill them.
Yeah, but they know how to pull tiny signals from among the chaotic mess of big distorted ones.
That hurricanes haven’t had a signal detected yet, if you read their literature, doesn’t mean they aren’t expecting these to emerge. People forget that expecting signals from land-based damage is extreme sampling bias. Ditto coastal damage. Most hurricanes pound oceans with their energy, and do most of their damage from precipitation inland.
New England’s long term risks coastal risks come from longer-lived and bigger nor’easters, not hurricanes.
What’s wrong with that?
> sound like frickin libertarian luckwarmers.
Try reading the sentence that follows, emphasized earlier:
Doesn’t sound luckwarm anymore.
> they find zippo about hurricanes.
They didn’t find anything about unicorns either. They did find however that hazard mitigation buries the normalized loss signal over the whole domain:
https://www.sciencedirect.com/science/article/pii/S0921800916309752
Speaking of sameosameo:
No wonder FrankB specializes himself in sensitivity studies these days.
When looking at insurance numbers, let’s bear in mind that they don’t cover what lies beyond insurance coverage.
https://link.springer.com/article/10.1007/s10584-014-1141-0
I eargerly await the “but Malibu” retort.
While being sensitive to the risk of sameosameo…
… I don’t recall RPJr. seeing RPJr. speaking directly to the limitations of looking at extreme weather damages as a function of GDP.
Anyone seen evidence otherwise? A link perhaps?
My latest algorithm for reading expository writing is to skip to the limitations section as a starting place. Unfortunately, so much expository writing has no limitations section.
zippo
http://www.air-worldwide.com/Blog/Climate-Change-and-U-S–Hurricane-Landfall-Probability/
like i said.
“We consider policy options such as climate change abatement, policies which encourage adaptation, reducing subsidies for coastal development, and disaster relief policies, which distort insurance markets.
What’s wrong with that?”
because looking at any other solutions than a carbon tax is a crime against humanity.
Steven,
In what way is it like you said? That link seems to suggest that Hurricane activity will increase if the Atlantic continues to warm. My point is that we should look first at how our emissions will impact these events and then consider how this would impact us directly (how much will this impact damage/costs). That link seems to present things in pretty much that way.
@-“It drowns the signal into GRRRRROWTH noise.”
Is that not the salient point, rather than an error ?
We can (and do) measure the physical changes in the environment in terms of increased heatwaves, drought, floods, stronger landfalling storms (W.Pacific) etc.
But it is difficult to find the signal of that in the cost of damages. The increase in population, wealth and infrastructure relegates the costs of adaption and climate damages to an insignificant proportion of the total ongoing cost.
Therefore the damages and costs of climate change are a minor inconvenience on the economic front, not an existential threat.
Willard, re the quote
Isn’t that overstating it? When it comes to building infrastructure, you want a good forecast of the average and peak conditions, but don’t care much whether the peak happens on a Monday or a Thursday, or the hurricane hits this year or next year. Yes if it’s a movable piece of infrastructure like the Thames Barrier, but most is fixed and is there all year round. Saying that something could go up or down with equal probability is reliable but useless, but saying it will go up, with high confidence, is reliable and useful. Especially for infrastructure which is more expensive to build wrong then fix than it is to build right in the first place.
Actually the IIS numbers do make an estimate of uninsured losses as well. Although they will obviously be subject to more uncertainty than the insured losses (I presume some are solid, e.g. turned-down claims when someone tried to claim on an excluded risk, official rebuilding cost of public assets which the state has self-insured).
Mosh
Because they’re rare and the impact is very contingent on where they land. To within a few km. But when you consider all weather-related catastrophes, not just hurricanes and not just the expensive stuff, they’re up. A lot. Notably floods, which kill many more people than hurricanes.
izen
But there’s more to consider than economic inconvenience. Mass fatalities, for example.
FWIW –
I asked Roger this:
Hi Roger –
Some questions.
1) Do you think that in balance, wthout significant resources devoted to adaptation and/or significant steps taken towards mitigation, the scientific evidence suggests that continuing/increasing anthropogenic emissions of CO2 are likely to play a causal role in increasing damages (e.g. costs, lives lost, dislocated populations, etc.) due to increases in imfrequency/intensity of extreme weather events?
2) A related question. Could you speak to the limitations of using damages as a % of GDP as a metric for evaluating the risks of extreme weather in association with ACO2 emissions?
3) A related question. Do you think there is a downside potential of relying too heavily on damages as a % of GDP as a window for evaluating the risks of increases in frequency/intensity of severeweather events in association with continuing/increasing ACO2 emissions?
He answered thusly:
Some replies:
1. Yes, see IPCC. Also see our papers on this related to TCs.
2. Don’t use disasters/GDP as a metric of extreme weather. The metric is important (It is included as a target indicator under the UN SDGs) but it is not a proxy for climate or weather data. If you want to understand risks of extreme weather, look at weather and climate data.
3. Again, don’t use economic data to evaluate weather frequency/intensity, use weather and climate data directly.
For more depth see my book on this, which goes into gory detail on all this. Thx
His answer #2 is interesting given the reaction to Everett’s comment above.
Dave,
I think Willard’s point (or John N-G’s actually) is a bit more general. If I predict a 20% chance of rain next month and it rains on 2 days out of 10, then my prediction was reliable, but rather unhelpful. It would be better to be able to make a somewhat higher resolution prediction. On the other hand, if I go to very high resolution and end up being mostly wrong, then that too is unhelpful. In the case of infrastructure, then the optimal may be reliability over resolution. When it comes to sporting events, I’d probably prefer resolution over reliability.
Joshua,
I saw your comment and Roger’s response. It does seem odd given that he seems to regularly highlight the link (or lack thereof) between extreme weather and damage/costs, yet suggests that it shouldn’t be used as a metric. He may not actually be doing the latter, but I’m pretty sure that what he says is interpreted by many as a metric which indicates that extreme weather events are not being influenced by anthropogenic emissions.
IIRC, in his Manichean Paranoia talk he suggested that people should be careful to ensure that what they say is not misinterpreted. Advice he appears not to really follow himself.
Indeed ATTP. Sometimes it’s relevant. But it can be co-opted into the we-don’t-know-everything-so-we-can’t-do-anything trope.
I was also going to add a comment to the effect that if I want to know whether or not to take an umbrella, I just need to know it’s going to rain. Not how heavy. Unless of course it’s dangerously so.
Of course if I live in the Sahara I don’t need a brolly. Actually that’s not true, I’ve been rained on twice in the Sahara but didn’t have a brolly. But the locals all drive around with no windscreen wipers, they just stick their head out of the window if it’s too bad, or stop. So they don’t benefit from forecasts, even if they’re accurate.
Dave,
Indeed, it can be co-opted in that way. In some sense, this was the context; Roger’s rather confused intepretation of the IPCC’s probabilities.
Anders –
It seems par for the course (which, I believe, is French for sameosameo).
IMO, Roger’s response can be justified with a legalistic defense, but IMO lacks accountability and rests on a foundation of plausible deniability.
In line with his response, we can could say that the reason he focuses on damages as a % of GDP (despite his indirect acknowledgement that there are limitations to the usefulness of the metric) is because he thinks it’s important to refute attempts to misleadingly focus on damages as a measure of the mechanistic impact of emissions on weather events.
Of particular note (IMO) is that he didn’t directly respond to my question – which was to address the limitations of the metric on which he focuses so much of his attention. Instead, his response focuses on the limitation of another metric. Non-sequiturs are useful rhetorical tools.
IIRC, in his Manichean Paranoia talk he suggested that people should be careful to ensure that what they say is not misinterpreted. Advice he appears not to really follow himself.
And yes, I have asked for Roger to address how easy it is to anticipate how what he writes will often be interpreted within the polarized context of the public climate change discussion; I have yet to get him to do so. I would follow-up along those lines in the comments section of his most recent post, but I’m rather sure it would be fruitless.
> The increase in population, wealth and infrastructure relegates the costs of adaption and climate damages to an insignificant proportion of the total ongoing cost.
If that was the case, normalizing would show a trend down. I don’t think that’s what the data shows. Otherwise one could argue that losses are anti-correlated to AGW. But then somebody would point out that the normalizing process abstracts away the fact that infrastructure investments have been done with increased landfall intensity in mind.
Beware double accounting.
Perhaps a more important ClimateBall move can be illustrated by this quote from a citation above:
The first part of that claim echoes AT’s point. Insisting on attribution is a way for ClimateBall players such as Junior to bypass it.
AT’s point should not go without saying. It should be repeated. Again and again. It is what we know. What we know matters.
> When it comes to building infrastructure, you want a good forecast of the average and peak conditions, but don’t care much whether the peak happens on a Monday or a Thursday, or the hurricane hits this year or next year.
NG is the Texas State meteorologist, DaveG:
https://en.wikipedia.org/wiki/John_Nielsen-Gammon
So he’s overseeing things like the drought index:
https://www.drought.gov/drought/states/texas
How one defines resolution and reliability should be relative to the intended purpose.
> Actually the IIS numbers do make an estimate of uninsured losses as well.
Of course. It’s how the insurance industry can delimit its future market. What I was referring to was figures such as this one:
https://twitter.com/nevaudit/status/979047511183720448
This shows the money insurers got versus what they paid in reclamations. It also shows the uninsured losses, but what’s most striking is how selling insurance has been profitable until 2004. The whole industry lost money this year in that market.
This is not normal. Even if we couldn’t rule out that 2004 was a fluke, the simplest explanation is that AGW increases the odds of extreme events. That we can’t predict when extreme events happen is not good news for insurance brokers and buyers alike. The insurance industry tries very hard not to lose money. Premiums will increase every year.
Washing our hands over this doesn’t solve the problem.
@-W
“But then somebody would point out that the normalizing process abstracts away the fact that infrastructure investments have been done with increased landfall intensity in mind.”
And otters would point out that building better infrastructure may confer better resilience even if a specific increase in landfall intensity is not a factor.
We may build, and benefit from brick houses even if a huffin’ & puffin’ wolf never appears.
Beware of double accounting general improvement as a cost of adapting to AGW.
Given the flexibility of economic arguments a case could probably be made that the extra resilience built into power distribution to withstand AGW is a small part of the general improvements that can be applied to new and replacement systems. It could be regarded as a Keynesian stimulus rather than a deficit.
(grin)
> Given the flexibility of economic arguments a case could probably be made that the extra resilience built into power distribution to withstand AGW is a small part of the general improvements that can be applied to new and replacement systems.
I like that idea. We could push it further by selling more conservative counties that investing in infrastructures would prove that the cosmopolitan elites have been wrong all along about AGW.
Speaking of wolfless brick houses, let’s bear in mind that building codes set up minimal standards:
https://icce-ojs-tamu.tdl.org/icce/index.php/icce/article/viewFile/3751/3434
In considering “the rising costs of weather disasters ” one should recall that the rate of rise of global radiatve forcing , temperature and sea level have in many developed nations been vastly exceeded by rates of urban growth and sprawl.
Let me give a formative example. CO2 stood at barely 325ppm when I first was driven South for the winter as a child. An unforgettable fact of that journey was the seemingly endless distance between Southern coastal towns, and the unending tracts of palmetto wilderness that separated them. Outside of a handful of cities, there were virtually no high rise buildings along the Atlantic coastal strip from the Carolinas to southern Florida , then, or for that matter , on the arc of the Gulf Coast .
Repeat that deive today, and you will seldom be out of sight of condominium towers- a conurbation of coastal suburbs and high rise construction is now the norm where only live oaks and palm trees stood two generations ago.
A storm surge or flood that would have uprooted a million trees in the 1950’s might damage or demolish as many homes or commercial buildings today.
OK. This is getting serious:
Russell –
In considering “the rising costs of weather disasters ” one should recall that the rate of rise of global radiatve forcing , temperature and sea level have in many developed nations been vastly exceeded by rates of urban growth and sprawl.
Can you walk and chew gum at the same time?
@-Joshua
“Can you walk and chew gum at the same time?”
Perhaps the point being made is that the ‘walking’ (development) is of much greater impact on the problem than the gum-chewing (AGW).
It may be the main reason that insurance payouts have risen.
Rare extreme events continue to occur at either the same frequency, or an increasing one. But the impact is dominated by the expansion in exposed population and infrastructure/assets.
Mortality rates, given the massive increase in population exposed have gone down in those regions rich enough to adapt.
And where local governence is effective.
But it is the general communal infrastructure that seems to determine mortality, and damage, in extreme events whatever their probability or intensity.
I know of no extent or theoretical Libertarian, or Free market model for a fire service that can respond to forest/bush fires.
There are two unstated implications of JPR’s ambiguity.
The undetectability of a change in extreme event incidence in economic indicarors is evidence there is no change in incidence.
That may be attractive to the outright AGW rejectionists, and may be why he attracts that fringe.
But the plausible deniability resides in the implication that impacts of extreme events are dominated by the development of the physical infrastructure and social organisation. The economic ‘cost’ of imposing adequate building codes is subsumed in the general Grrrowwth!
> The undetectability of a change in extreme event incidence in economic indicarors is evidence there is no change in incidence.
Well put.
***
> [T]he plausible deniability resides in the implication that impacts of extreme events are dominated by the development of the physical infrastructure and social organisation.
Indeed. If only men were truly from Mars, and stayed there. A similar argument can be made about insurance payouts: they increase first and foremost because people buy more insurance. Getting rid of insurance would solve that problem.
A less radical idea would be to shoot for the middle ground:
Here would be a proposal: every odd year, the state covers all your insurance; every even year, you’re uninsured. A true climate roulette. Let’s not call it Russian, it’d sell less.
@-W
An inverse correlation with AGW.
https://www.climatechangepost.com/news/2018/2/26/number-people-dying-storm-surge-floods-decreasing/
“The occurrence of very substantial loss of life (>10 000 persons) from single events has decreased over time, especially since the 1960s. This is remarkable since world population has approximately doubled since the 1960s, and increased six-fold since 1900, with much of this increase occurring in the coastal regions. Storm surge mortality, the fraction of people exposed to the storm surge flood that are killed, has decreased consistently for all global regions, except South East Asia.”
Atlantic basin hurricane activity has increased. Whether or not that is attributable to AGW or not is not known. But if you look at all counts of all extra-seasonal hurricanes, the increase is significant and there is a trend. Note I am not qualifying these to be land-falling. Thr restriction to extra-seasonal is there to get adequate statistical power.
The highlighted portion is a non sequitur. It assumes that the only way of forecasting damages and costs of climate change are through a retrospective assessment of cost of damages. There could be many ways to do this. That proves the non sequitur.
In particular, the estimate of the expected value of damages and costs of climate change should be done conditional upon many sources, including modeling of future trends, whether by climate models, informed extrapolations, or knowledge of physical science. To do otherwise is driving down a road by only looking in your rear view mirror.
On reliability versus resolution, I’d say that’s fine, except that the standards for climate forecasts ought not be tougher than those to which other standards or estimates are held, most having forecasts implicitly built-in. In particular we have building codes for earthquakes and building codes for flooding, at least in Massachusetts and Texas, and possibly more than I do not know about. That the expectations of damage are bigger than those for earthquakes and other things and, so, imply greater prevention measures and costs ought not, logically, be a reason to hold them to higher standards. If the expected damages are bigger, but prevention and adaptation too expensive to embrace, that ought to be stated transparently so, with some probability, people can prepare to lose everything. That this might have a dampening effect on present day prices is good, as it should.
izen –
Perhaps the point being made is that the ‘walking’ (development) is of much greater impact on the problem than the gum-chewing (AGW).
Yes, perhaps. And if so, then the point is part of a productive discussion.
But absent some greater measure of specificity, with mention of caveats and context, then I see no particular reason to refrain from asking if walking and chewing gum at the same time is beyond our capabilities.
Rare extreme events continue to occur at either the same frequency, or an increasing one. But the impact is dominated by the expansion in exposed population and infrastructure/assets.
Perhaps – but with the use of present tense there, then maybe that’s like shooting a moving target in the same manner that you’d aim at a stationary target. And so, again, I think that context and caveats are important. In planning for the future, it seems to me that it isn’t enough to say what dominates the signal, currently.
Mortality rates, given the massive increase in population exposed have gone down in those regions rich enough to adapt.
I added bold to emphasize important context – context of a sort that is frequently left out in the discussion. Which is problematic because that context is very important.
And where local governence is effective.
More important context. What (from what I’ve seen) often gets lost in the trading of rhetorical scoring points is the important context that these problems are not either/or local/ regional/national/global, but some combination of them all.
But it is the general communal infrastructure that seems to determine mortality, and damage, in extreme events whatever their probability or intensity.
I know of no extent or theoretical Libertarian, or Free market model for a fire service that can respond to forest/bush fires.
Yes. And that is why when a libertarian starts to talk about adaptation I look to see caveats and context w/r/t discussion at some level beyond theoretical, where funding is addressed beyond the invisible hand of free market fetishism.
I believe that hyper was just making a related point (albeit, more coherently).
izen –
But the impact is dominated by the expansion in exposed population and infrastructure/assets.
A thought that occurs to me regarding that point.
I imagine that growth in population and correspondingly growth in infrastructure/assets exposure, are somewhere between linear and exponential (keep in mind, I’m math illiterate so grant me some leeway in understanding and vocabulary).
But growth in damages from severe weather, it seems to me, runs the risk of extreme increase (a kind of step change increase?) even if the growth in frequency/severity is smoother. Consider even a linear growth in average severity which, at some point, renders Miami or New York inhabitable.
…uninhabitable…
Concerning extreme events, Curry thinks she can predict ENSO :
Far far more importantly, and only secondarily dependent upon population, global consumption of stuff is increasing exponentially. Every person on the planet currently consumes 10.5 metric tonnes of raw materials per capital per annum, about 30% higher than a sustainable rate.
Suggested references:
* Wiedmann, et al, “The material footprint of nations”, 2015
* Davis and Caldeira, “Consumption-based accounting of CO2 emissions”, 2010
* Hoekstra and Wiedmann, “Humanity’s unsustainable environmental footprint”, 2014
@-Hypergeometric
“To do otherwise is driving down a road by only looking in your rear view mirror.”
That is one problem with this approach. If all you see in the mirror is a straight road, you may not allow for non-linear impacts.
But the view forward is not clear.
It can, and is informed by modelling and knowledge of physical science.
http://impact.ref.ac.uk/CaseStudies/CaseStudy.aspx?Id=37295
How well this is done depends on the effective actions of local governence. It is a communal economic process.
@-Joshua
“But growth in damages from severe weather, it seems to me, runs the risk of extreme increase (a kind of step change increase?)”
Yep, some of that modelling and projection indicates that climate can throw a curve ball.
A sea wall protecting the subway system against a 12ft storm surge stops working at 14ft.
I am not sure this clarifies whether adaption can be separated out as an independent expense, or gets neutralised as an inherent cost/benefit of Grrrowwth.
T J Bass makes a convincing case that the carrying capacity of the planet is around three Trillion. With a little GM.
Joshua, while loathe to chew gum in public, I have on occasion won small sums by keeping a cigarette lit while body surfing.
Russell –
I guessing it might have been one of those funny cigarettes?
Food for thought…
Though we’re starting to feel the effects of climate change, those effects are not dramatic enough on a day-to-day basis to convince the majority of Americans that climate change should be taken seriously, says Magali Delmas, a professor at the Institute of Environment and Sustainability at UCLA. Human beings aren’t great at dealing with situations that are high-risk but don’t happen that often. Think of earthquake insurance, for instance. Though there’s a 99 percent chance that there’s going to be a magnitude 6.7 earthquake in the next 30 years in California, the probability that it’s going to happen in the next year and is going to affect me is much lower. As a result, only 13 percent of people in California have purchased insurance from the California Earthquake Authority, Delmas says.
About half of Americans don’t think climate change will affect them — here’s why by Alessandra Potenza, The Verge, Mar 29, 2018
izen
Which is why I linked to this graph Which shows a big increase in natural catastrophes defined by fatalities (>1) and a cost metric from $10k to $3M depending on the country’s income group. That’s low enough to encompass one condo in Florida and a pier or motorised fishing boat in a poor country. Or a decent-sized field, forest or plantation owned by a farmer, even in a poor country. So should be less sensitive to suburban sprawl than the $$ cost. Those virgin forests might have been uninsured, but I bet the cotton or sugar-cane fields, were, or at least the farm buildings and equipment. And worth more than $3M, in replacement cost and/or lost production. Putting more people in harm’s way might increase the fatality count, but better building codes and early warming systems should decrease it. So it’s not clear that development inherently increases that metric. If you’re a “Because Grrrowwth!” person you might even argue that making people richer and putting them in brick houses mitigates the climate risk to life better than cutting emissions, making Westerners poorer and reducing their demand for palm oil so the workers have to live in shacks (see, anyone can play sceptical economist! 😉 ).
But can RP Jr plausibly justify picking GDP as a metric when there are better ones publicly available (literally the first Google hit for “catastrophe losses by year”). And that’s before you even get into damages not counted in GDP, rebuilding costs counted as a GDP addition and lost production not counted if some other plant takes up the slack, unless it’s overseas in which case it still makes it into global GDP.
It’s reminiscent of the way homeopaths and other fringe scientists somehow forget all they learned about experimental design and conduct underpowered experiments with a high risk of false positives built in. E.g. feeding GM food to mice specifically bred to get cancer, so you guarantee the control group and the test group will both get cancer. Or feeding raw GM potatoes to rats in the knowledge that (a) raw potatoes are as poisonous to rats as they are to humans and if humans eat raw potatoes, whether they’re GM or not is the least of their worries; and (b) the control group and test group are both guaranteed to get ill. Why not use mice with a normal cancer incidence and cook the potatoes?
I agree Willard. And had guessed who NG was. But the original quote read like an overly sweeping statement. “reliability without resolution is not useful” makes sense for a weather forecaster and perhaps for a State meteorologist. But not for a state hydrologist or urban planner writing building codes. If you’re designing for the 100-year storm you literally don’t care whether it comes in year 1 or year 99. Only about getting the magnitude right. The IPCC’s purpose is not weather forecasting, so its intended purpose is different and it’s allowed to use different definitions and trade-offs. The various CMIP models are all about reliability not resolution* but are not useless.
*For metrics like global temperature, but not for metrics like how well model El Niño temperature excursions match actual El Niño temperature excursions. And even that is about post-dictive resolution to validate the model, not predictive resolution to guide public actions.
Reads like a bit of a straw man, perhaps unintentional. The IPCC could do that but it doesn’t. Because physics. And because observations. We’re not in the a priori world any more and haven’t been for decades.
Scary. Not because I have a vested interest in insurers, but because I want them to pay out when I need them. Remember, Lloyd’s of London almpst went under in 1993. Which was tough on the investors who had to raise new capital, but it would have been tougher on customers if their insured losses had tuned overnight into uninsured losses.
I’ll bet that nowadays there’s so much reinsurance spread opaquely across the big companies that if a Lloyd’s or a Munich Re went under, we’d be in for a 2007-8 domino effect. A too-big-to-fail bailout moment coinciding with a call for governments in the affected areas to act as insurers-of-last-resort.
> I’ll bet that nowadays there’s so much reinsurance spread opaquely across the big companies that if a Lloyd’s or a Munich Re went under, we’d be in for a 2007-8 domino effect.
If we’re lucky. Insurers are the ones who helped us get out of that mess:
Source: http://www.oecd.org/finance/insurance/45044788.pdf
Where insurers fail, the government needs to take over. Not exactly a good thing for nobody. More public debt leads to more tabloid shrieking, which favors reactionary and populist attitudes:
Source: https://www.ceu.edu/sites/default/files/attachment/event/14668/economic-correlates-populist-attitudes-eva-anduiza.pdf
@Willard,
Well, despite disagreeing with them, I can see where the populist sentiment comes from. The placement of assets in very risky places and a system for reimbursement when the risk is realized is just moral hazard, and when such are socialized, I can see a justice in the people bearing the cost of the failure complaining and worse.
Trouble of course is that the people who profit from these moral hazards, whether coastal real estate owners or banks, elect local reps who support them and support the maintenance of the hazard.
Joshua, they’re clove actually, and i suggest you check them out, as climate modeling has become more than a two-pipe problem.
Heroic splif consumption better explains this chap’s view of Charney feedback:
https://vvattsupwiththat.blogspot.com/2018/03/marylebone-batsman-declares-victory-in.html
In fairness to the UK rescues at least, where the Government used the cash injection to buy shares, the shareholders did lose out, by equity dilution as well as by the share price fall. In the case of RBS, by about 80% relative to the share price when they finally settled a Court challenge in 2017, by 92% relative to the rescue price, and by 99% relative to the pre-collapse price. And the executives all lost their jobs (although they and the trading-floor gamblers got to keep their bonuses).
The price of making the shareholders lose that last 1% by allowing the bank to fail would have been Granny losing her life savings.
But, @Dave_Geologist, the same reasoning could be applied to coastline properties on the Eastern shore, where equity is all a retiree has, say. This would apply whether or not the home was a loss or subject to something like a 40% of assessed value buyout program as they have in Texas. Admittedly, the effect of such plans or actions upon coastal real estate values would probably be severe.
In places like Scituate, MA, at town meetings addressing the problems, people with houses at risk beg for au
(Sorry, was entering from my phone, and got cut off. Continuing.)
In places like Scituate, MA, at town meetings addressing the problems, many people with houses at risk beg for any action that will save their properties, typically in emotional appeals. Most of these are not second homes.
In fact, seeing the response of people on the shores is discouraging because it suggests that even if direct impacts of sea level rise and climate change begin occurring frequently and at scale, people could be so overwhelmed they would stop considering why this was happening. It’s like the standard line which local and regional authorities give when asked by journalists about possible connections between storm flooding and climate: “Now is not the time to consider these things, but, rather, get people the help they need”, etc. Of course, no time arises when these things are considered.
This is why it is a mistake to continue to allow people to rebuild in these areas. There is no winning the matter.
The Town of Sandwich on the Cape is considering a very expensive beach front hardening, with armoring of the beachfront and a line of such rocks offshore. Of course they’ll lose a lot of their beach, and the interruption will starve nearby beaches of rebuilding sand. Moreover, the project is so expensive, they will seek Commonwealth and federal funds to do it. The object is to save the marsh and then the old Town. However, there are places which simply cannot be saved directly:
Here’s why:
Cape Cod is the poster child for this kind of change.
The southeast Cape in 1984:
The southeast Cape in 2017:
Woods Hole Oceanographic Institution (WHOI) is working on moving their lab and machine shops from the Iselin (“Main”) Dock in Woods Hole Village up the hill to protect these assets from sea level rise:
Willard and hyper:
I did rather focus on the bank moral hazard. The UK has introduced a flood insurance moral hazard by setting up a flood reinsurer, Flood Re, which is funded by a levy on all insurance policies. That is supposed to allow homeowners at risk of flooding to get cheaper insurance with lower excesses. But at the expense of everyone else.
At least it excludes homes built after 2009 (presumably the argument is that by then we knew we were in a new world risk-wise), businesses and landlords. But although some of the properties that benefit from it were no doubt built in good faith when their location was less vulnerable than it is today, a better long-term solution might be to abandon them the next time they flood and use the money to build a new house somewhere else.
hypergeometric, your aerial photo of Sandwich looks uncannily like Sussex’s Manhood Peninsula, apparently the site of Europe’s largest managed retreat and ‘one of the most sustainable projects the [UK] Environment Agency has ever delivered’:
https://www.bing.com/maps?osid=52f35272-b40a-4786-8d64-8f04cb5b24aa&cp=50.742208~-0.813176&lvl=16&style=h&v=2&sV=2&form=S00027
https://www.ice.org.uk/knowledge-and-resources/case-studies/managed-realignment-at-medmerry-sussex
HTH.
> I can see a justice in the people bearing the cost of the failure complaining and worse.
That’s still how insurance works usually. This reminds me of Paul Ryan’s gem:
http://www.msnbc.com/rachel-maddow-show/paul-ryan-flubs-the-basic-idea-behind-insurance
One might as well argue that insurance can’t work because of the way it works.
@Vinny_Burgoo,
That photo came from here. It should be no surprise. That these similar systems look similar is partly the reason why @Dave_Geologist has a profession.
Indeed, my beautiful bride and I are taking a holiday this summer to Bar Harbor, ME, and Acadia National Park having spent some time on the “other side” of the Iapetus Ocean from Northern Ireland (and, roughly speaking, Ireland generally) which we visited in August of 2017. While the dating of outcrops is only approximately the same, the similarity of the Antrim Coast to Acadia is striking to someone who’s been to Acadia many times, even though the basaltic dikes in Acadia (Cadillac intrusive system) are Silurian whereas, for instance, the Giants Causeway is Tertiary. I’m sure @Dave_Geologist could chime in much better. My knowledge of this this is the Wikipedia article and
D. L. Kendall, Glaciers & Granite: A Guide to Maine’s Landscape & Geology, 1987, Down East Books
In particular Kendall has a figure on page 167 taking from Maine Geological Survey Bulletin 26, 1982:
There’s a similar presentation from Wikipedia:
From the other side,
which is from:
D. M. Chew, “Grampian orogeny”, chapter 4 in C. H. Holland, I. A. Sanders, The Geology of Ireland, 2nd edition, 2009, Dunedin Academic Press.
I read these things as a kind of a hobby, even if I’m pretty weak on Mineralogy. In fact, I’m working on a post for my blog at the moment concerning the physics of columnar jointing which I hope to get done by the weekend’s end.
My Comment above contains a mistake …
That is not at all true and it isn’t what I meant. I just got to typing away. The outcrops differ in age roughly 390 million years, even if specific rocks and locations have ages which vary from a nominal. I was going to make a point about processes and rock types being similar despite vastly different ages, but I never got there, having gotten busy looking up figures in books.
Apologies!
I have always though “attribution” was a bit of a mugs game tbh
if you have a pack of 52 cards and you introduce another Jack of Clubs into it – you have changed the odds in turning over any one card – whether that is a Jack of Clubs or not. In fact if even if you draw a Jack of Clubs you have no real way of knowing whether it was the original one or the extra one you added into the pack
all you can say is that you would expect, over the long term, an increase in the number of times a Jack of Clubs is drawn – and that increase is consistent with adding a extra Jack of Clubs into the pack
I just view climate change/weather extremes as adding that extra Jack of Clubs, no extreme weather event comes with a handy label that says “courtesy of climate change”
DG
“A priori, anything could go up or down with equal probability, so the IPCC could make perfectly reliable but useless predictions by saying for all weather and climate variables have equal chances of going up or down. Reads like a bit of a straw man, perhaps unintentional. The IPCC could do that but it doesn’t. Because physics. And because observations. We’re not in the a priori world any more and haven’t been for decades.”
–
Is this correct?
My understanding would be that when making new a priori assessments, they are made from a basis of observations and occurrences up to that time hence the probability should be equal of an event or trend going either way from it’s then present course.
If no new knowledge or expectation is to be brought to bear in making a prediction.
@tadaaa,
What makes it more complicated is that we don’t really know the makeup of the pack to begin with, having but estimates. As many have suggested, if it’s that important to know whether or not this is happening, in order to assess how much to spend to mitigate or adapt, the really logical thing to do is dump a bunch of money into finding out more about the pack, or roulette wheel, if you like. But, of course, noooooooooo, what’s done is the amount spent is cut. Why? Oh I don’t know, because, as emeritus Prof Richard Lindzen of MIT has said, to “Cut the funding of climate science by 80% to 90% until the field cleans up”.
I am so glad I’m not a Christian ….
More Curry strategy. Example of climate science research entering the IP realm — all of CFAN’s hurricane and ENSO predictions are guarded secrets. No need to share because that’s not where the money is
angech
My point was that the IPCC (and its predecessor scientists who got nations and the UN sufficiently concerned to set up the IPCC in the first place) does have access to “new knowledge or expectation”.
Specifically, the laws of physics, as observed, for example in the radiative properties of CO2 measured since the 19th Century, observations of CO2 content directly since the 1950s and indirectly over millions of years, measurements of solar radiation, knowledge of stellar processes and measurements of upgoing and downgoing IR radiation, material balance (conservation of mass) calculations involving emissions, the biosphere, atmosphere and oceans, and the physics implemented in GCMs.
So, for example, the planet could only not warm if God intervened to stop it. Period. Therefore the IPCC can with high confidence project future warming based on a particular emissions path. Temperatures are much, much more likely to go up than down. An a priori assumption of equal probability is only appropriate if you’re prepared to deny physics.
The GCMs are sufficiently well constrained that we know what to expect by way of increases in extreme events. That attribution is just beginning to be possible, and that some attribution is now possible, is a function longer records while the climate continues to change, giving more signal to detect against a constant background noise.
Remember, scientists wait for 95% certainty before making the call. If we were content with as-likely-as-not or 50/50, many events would probably have passed that threshold already. And the converse (95% certain it’s not happening) has never even been in the race. That’s one reason the faux pause was faux. Even if you ignore the cherry-picked, too-short timespan, the fact the the lower 95% C.I. bound (just) encompasses zero trend doesn’t mean we’re 95% confident there’s zero trend. It means there’s a 5% chance there’s zero or negative trend. Hence a 95% chance it’s not,. And looking at the full C.I. range, it’s obvious that all of that 95% sits in the warming half of the graph. (I’m aware that my usage here of confidence and certainty is colloquial and not strictly correct in professional statistics.)
The reality of global warming is the reason AGW deniers are on an even stickier wicket than evolution deniers. The signal for evolution is mostly there already (no big breakthrough since genome sequencing), so they only have to face incremental increases in the evidence and can hang onto their old arguments and beliefs. Even though they’ve been thoroughly debunked for decades, the level of debunkedness has remained fairly constant. So the degree of suspension-of-disbelief required to deny the evidence hasn’t changed much either.
Global warming really only took off in the last 40 years. So every decade adds a half, a third, a quarter, a fifth, etc. more signal against a constant background noise. That alone is enough to guarantee that detection and attribution studies will become more confident over time. Add the fact that not only is the signal duration getting longer, the signal strength is getting bigger, decade by decade, and it’s a slam dunk.
Evolution is analogous to a drug trial where you repeat the experiment with the same sample size each time.
AGW and its impacts are like a drug trial where you progressively increase the sample size. One thing to look for there is whether the signal gets stronger with increasing sample size. If the initial result was a false positive (Type I error) you’d expect the signal to get weaker with increasing sample size. Guess what we find with global warming? And with extreme event attribution?
tadaa
I agree that if scientists did attribution the way you describe it, it would be a mug’s game*. I’m sure the scientists doing attribution studies would agree with both of us.
Fortunately that’s not how scientists do event attribution. I’ve not read a single scientific paper which says a particular extreme event was entirely down to climate change. Perhaps you can point me to one? And not to a newspaper article by a journalist, or a University press release, or a scientist speaking in his private capacity (where you may find one or two but certainly not the hundreds or thousands working on attribution). Plenty which say AGW has made the likelihood of this event occurring higher, or attributing a percentage of that event to AGW (in the sense that if it’s made the likelihood of a particular event twice as high as before, you can attribute half of the death or damages to AGW and half to natural variability).
Your statement is equivalent to saying “we can’t tell if smoking causes lung cancer because some people who’ve never smoked get lung cancer”. We can’t tell if it was responsible for a particular lung cancer death, but we can say with confidence that it’s responsible for most lung cancer deaths.
A similar approach can be and is applied to extreme weather events. Attribution of Extreme Weather Events in the Context of Climate
Change. You can download a pdf for free.
See also my response to angech. Not only is the signal getting stronger because it’s getting longer, it’s getting stronger still because the departure from noise is getting bigger.
I prefer to use a pair of dice as 52 cards understates the scale of the problem. If we assume that a heatwave is 12°C warmer than a normal summer, we used to be playing with two dice with sides labelled 1 to 6. One of the dice is now labelled 2 to 7. Even if we meet the Paris target, by 2100 both dice will be labelled 2 to 7. Our chance of throwing 12 or greater will have quadrupled. And we’ll be seeing previously unprecedented 13 or 14s with three times the frequency we used to see 12s.
And it’s actually worse than that because of the fat tail of the extreme event distribution. 1°C can make a particular extreme event 3 or 4 times more likely. That’s how statements like “90% of the heat in this heatwave was due to natural variability, only 10% to AGW” can be consistent with “AGW made heatwaves of this magnitude three times more likely”
*If you mean it’s a mug’s game rhetorically, because bad actors can attack it with specious arguments and extend them to criticism of climate science in general, that may be true. But
(1) Bad actors will act badly whatever scientists say. If we applied the same logic everywhere, stores wouldn’t sell steak-knives and utilities wouldn’t pipe gas to homes.
(2) It’s really only in the U.S. that AGW denial has political traction, and anyway the purpose of the scientific papers is not to persuade the public or politicians that AGW is real. It’s to further basic scientific understanding, and to provide planners with the information they need to save life and property. Even in the U.S., Federal and State agencies can get on with that as best they can. And if they’re stopped for four years, there will be an evidence base there to give them a running start when they resume. And of course the rest of the world can ignore the U.S. and reap their own benefits.
(3) Scientists have a duty to present science without fear or favour. If they’re concerned it may be misused, they should be careful about how interviews and press releases are handled, but should never censor their results.
I partly agree with tadaa’s point about attribution, although maybe for slightly different reasons. Clearly it is scientifically interesting to see if we can do such analyses. However, it’s also partly to answer those who ask “how do you know it’s us?”, or “how can you be sure?”. Most people who ask that are mainly doing so to sow doubt, so answering their questions won’t really convince them; they will simply move on to the next question, or misrepresent the response. We’ve already seen this with claims that the attribution studies indicate that we’re only responsible for 50% of the observed warming, even though it’s clear that they’ve demonstrated that we’re responsible for at least 50% of the observed warming, with the best estimate being slightly more than all of it.
ATTP
I’m probably a bit more cynical than you about “how do you know it’s us?”, or “how can you be sure?”. If people don’t believe the mountain of evidence that’s already there, some of it quite simple, they’re not going to be persuaded by even more abstruse actuarial and statistical analysis. And they won’t believe “but Harvey” because climate scientists will honestly give the same sort of complex answer I gave above. And use the complexity as an excuse to say we’r hiding that we don’t really know.
So I think the attribution studies are only useful to the people and organisations who are listening. In part because there is still genuine uncertainty about timing and magnitudes, so they can in good faith ask for confidence limits and demonstrations of predictive power before making large investment decisions. Not “is it really us” but “how much, when and where”. And in part because some of them will be driven by actuarial decisions. Road improvements, pedestrian crossings are authorised on the basis of calculations which say “we’ll spend £1M to save a schoolchild’s life but only £200k to save a commuter’s life”. The numbers are from faulty 15-year old memory, so wrong in detail, but local authorities do make that sort of decision. I was involved in my former employer’s attempt to get a pedestrian crossing installed between the office and car park, crossing a busy road congested with trucks and warehouses as well as cars. And it did emerge that they’ll put a pedestrian crossing in front of a school but not an office because the public cares more about child than adult deaths. I’m not challenging the decision – I’d probably agree with it – but simply making the point that infrastructure investments aimed at saving lives have a cost-per-life element in the calculation. And of course investments to save financial loss have a direct cost-benefit calculation.
Dave,
That was essentially what I was suggesting. In some respects, attribution studies are futile attempts to convince the unconvinceable. Of course, they do have scientific merit, but I think they were also an attempt to present a robust analysis of why it is us. However, even this won’t convince some, many of whom have ended up misrepresenting these results (“it’s only 50%” as opposed to “it’s extremely likely to be more than 50%”).
angech
In re-reading my posts in light of ATTP’s , I realised that I might have misinterpreted yours. I read the original J N-G comment as meaning equally likely to go up as down. Yours could be interpreted as equally likely to increase at a faster or slower rate.
If the latter, I would still argue for the use of physics-based models which allow the central estimate to be faster or slower than previously. That’s be cause there’s no reason any particular increase should be linear. In the case of mean surface temperature, the long-term trend has been approximately linear for decades. But that’s just because, by coincidence, exponential emissions increases (Because Grrrowth!) approximately balance a logarithmic temperature dependence on pCO2 (because physics). If emissions trends change, the slope will increase or decrease, as is reflected in the various RCP projections. And even the logarithmic dependence is only an empirical approximation, which holds over a restricted range of pCO2.
ATTP
I also think attribute studies are important for people who are not in denial, but are complacent. Who think it won’t happen to them, or it won’t be bad for fifty years, or that only poor countries will suffer, rich ones will engineer their way out of the problem. Look, it’s happening now, not just in Pakistan or India but in Moscow, Paris and Houston. And yes, it’s happened before, but it used to be once a century, now it’s once a decade. In some ways that may be more powerful than a bogeyman which they’ve never experienced. People haven’t forgotten Paris 2003, but most probably don’t expect it to happen again in their lifetimes. For young people at least, it’s not only likely, it’s virtually certain. If they grasped that, they’d lobby harder for mitigation, both on the emissions and investment front.
Dave,
Yes, that is a good point. I agree, it can help people who acknowledge AGW to recognise that anthropogenic factors are already influencing climate events today.
The Onion does consumer (market) choice:
Thanks DG.
As extreme events go, it’s hard to beat the Falcon Heavy launch, which established Elon Musk’s 2008 Tesla roadster as the vehicle with the world’s worst fuel economy.
Don’t worry- its carbon footprint can’t be undone. but its MPG economy has already fully recovered.
Pingback: Disasters and Climate Change – part 2 | …and Then There's Physics
Pingback: 2018: A year in review | …and Then There's Physics
Pingback: Extreme weather event attribution | …and Then There's Physics