Don’t blame hurricanes Irma and Harvey on climate change?

Says Alan Reynolds in Newsweek, although originally at Cato. But Michael Mann says he wants you to let @Newsweek know what you think about them running Koch-funded Cato Institute climate denial propaganda. But I only care because someone called "Lawrence Torcello"1 Tweeted Quoting Popper against climate science signals Pseudointellectualism.

Obviously, quoting Popper against any genuine Science must be wrong. Although equally obviously it can't possibly, of itself, signal pseudo-intellectualism; it could simply signal stupidity or error. But the more extensive implicit claim - that quoting Popper against anything said about climate science must be bad - is obvious nonsense. But enough generalities, what of the present case?

Before going on, it might be helpful to read, say, Neptune's revenge by mt, who cannot possibly be accused of being Koch-funded or anything of that ilk.

The core of the Newsweek article is attributing today’s extreme weather to “climate change” regardless of what happens ( maybe droughts, maybe floods ) is what the philosopher Karl Popper called “pseudoscience.” If some theory explains everything, it can’t be tested and it is therefore not science. (Popper’s favorite examples of pseudoscience were communism and psychoanalysis.) [The article also contains various stats and numbers that I wasn't very interested in and didn't trouble myself to check; and some other rather dubious assertions that are uninteresting but which would earn my ire if I could be bothered to analyse them.]

I think it is interesting to ask if Popper's viewpoint has any explanatory power in the ever-widening debate about how to say something meaningful about the relationship between hurricanes and GW. So for example I think it was inevitable that Harvey and so on would be "blamed" on GW, but that if they hadn't happened, that wouldn't have cast doubt on GW. But that latter point is uninteresting, because there are so many lines of evidence for GW that it isn't possible to honestly doubt it. The Newsweek article, though, is too un-nuanced in its accusations to be much use. I've read lots of pieces about the connection between Harvey etc and GW, and almost all of them have been pretty equivocal about the causal connection. But is there any testable theory that Harvey would be evidence for or against? Remember, whatever the theory is, it must be proof against the absence of Harvey last year, or the year before that. Perhaps More on Bayesian approaches to detection and attribution is relevant.

Anyway, I wondered (sea-ice betting having rather faded out due to lack of diverging opinions) if we could translate this into a bet. Does anyone think that "GW caused Harvey (or Irma, or whatever)" translates in any way into a meaningful prediction ability for next year, 2018? Or if the storm season for 2018 reverts to normal, will everyone be completely unsurprised? I'll take the "revert to normal" side of the bet, of course. If anyone has $1,000 or above for the "Ha! 2018 will make 2017 look like a picnic!" side, we can discuss terms. Actually, you don't need to be that extreme, I'm sure something far more moderate would do.

Incidentally, if 2018 did turn out anything like 2017, it might be time to revive a piece of pure speculation I made on sci.env perhaps 20 years ago now. I was talking about tornadoes in the Southern US, but the concept works as well for hurricanes: how much stronger / worse would they have to get, to make the entire area economically uninhabitable?2 People will rebuild from one year, and perhaps even from one year a decade, but not from every year.

Notes

1. Ah. He's a Mann co-author and climate alarmist.

2. Eli catches up.

Refs

* Bonus Quotation of the Day at CH: on new ideas.
* Eli says that climate change increased the DAMAGE from these storms.

More like this

It isn't. Well, it is a little, but not totally. OK, it is, but actually, it is complicated. First, you are probably asking about the Atlantic hurricane season, not the global issue of hurricanes and typhoons and such. If you are asking world-wide, recent prior years were worse if counted by how…
"Hurricane Katrina overwhelmed levees and exploded the conventional wisdom about a shared American prosperity, exposing a group of people so poor they didn't have $50 for a bus ticket out of town. If we want to learn something from this disaster, the lesson ought to be: America's poor deserve…
Um. sorry folks, don't blame me, blame Eli. 'twas the now-aged lagomorph who attempted to interest me in the good old days of sci.env when we were all young and bushy-tailed. And indeed that thread does make for interesting reading: the present-day switch to blogs doesn't encourage that style of…
There's a thread on twitter, started by "@JacquelynGill" noting "The Day After Tomorrow", "@ClimateOfGavin" replying that "it was that movie and lame sci community response that prompted me to start blogging", and continuing "Spring 2004 was pre-RC, Scienceblogs, etc. Deltoid was around, Stoat, @…

Irma seems unable to decide whether to descend on Trump's winter White House- the bottom of whose swimming pool is well below sea level. or strike out at Palaeorepublican waterfront property lovers along the normally placid Gulf side of Florida.

Either way, somebody is gong to throw a hell of a hurricane party.

Charley Cat 4 struck Florida in 2004, 110mph winds at landfall
Wilma Cat 5 struck Florida 2005, 120mph at landfall
Place is not identical, but not all that different either (and maybe Irma won't be far off either and Andrew 1992 exited Florida nearby as well. These are all very costly hurricanes all top 10?). 14 and a bit months between them so not much more than a year.

So, 'but not from every year' presumably means expected return time rather than a one off actual return time. I assume that for any location that is well over 20 years for 100mph windspeed hurricane hit.

Don't think anyone (sensible) is talking about doubling the frequency let alone increasing frequency by 20 fold.

Can't see how you are going to get any takers for your suggested bet, but you never know.

How risky is that stretch of coast? Anyone want to suggest how much more risky that is compared to nearby similar length of coastline? Maybe highest and lowest factors people suggest can formulate a bet between them?

The Guardian
Sunday 10 September 2017 00.05 BST

Bob Ward is policy and communications director at the Grantham Research Institute on Climate Change and the Environment at the London School of Economics and Political Science

" Climate change cannot be blamed for the hurricane count in any single season, nor for the occurrence of any single storm, but there are three ways in which it is making the consequences worse.

First, although the intensity of a hurricane depends on many factors, warmer seawater tends to promote stronger storms. Average sea surface temperatures have been rising, and some parts of the North Atlantic and Gulf of Mexico are warmer than average at the moment, which is a key reason why both Harvey and Irma became so strong so quickly.

Second, a warmer atmosphere can hold more water vapour, which can result in heavier rainfall. That is true not only for hurricanes but also for weaker storms across the world. Even relatively mild tropical storms can cause great damage by dropping huge volumes of rain over one area.

Third, apart from strong winds and heavy rainfall, hurricanes cause damage through storm surges as their winds push seawater ahead of them. Storm surges can inundate extensive low-lying coastal areas, sweeping away everything in their path. Sea levels have been gradually rising globally, making storm surges bigger and deadlier."

By Rod LeBlanc (not verified) on 09 Sep 2017 #permalink

Unless Karl has a couple of placebo Earths stashed out beyond Pluto his POV is pretty well useless for this.

[I think you've misunderstood what he is saying; because what you've just said is irrelevant to his point1 -W]

Besides which the question is badly formuated. How about that climate change increased the DAMAGE from these storms and that the refuseniks refusal to acknowledge climate change CAUSED more damage because nothing was done in anticipation,

[It seems reasonable to me, as I've said, that GW increased the strength and thus the damage potential of these hurricanes. Though as I also said, it's associated CO2 comes in part from work that has made the USA more resilient to hurricanes, so the overall balance becomes unclear -W]

[1. Adding: we're not talking - at least, I'm not talking - about the kind of strict Popperianisn that would make, say, observational astronomy non-science because it makes no predictions. I'm talking about the idea that, as was quoted, a theory that can "explain" any observation explains nothing -W]

By Eli Rabett (not verified) on 10 Sep 2017 #permalink

>"climate change increased the DAMAGE from these storms

> [that GW increased the strength and thus the damage potential of these hurricanes]

It seems to me that 'damage potential' is a lot more coherent than 'DAMAGE from these storms'.

We know there is a butterfly effect why would this not apply to emission of the marginal ton/Kg/g of CO2 (more than a couple of months ago)?

Popper is useless when dealing with theories that are constructed from components, each of which is tested and accepted, but which combined extend the range the individual components beyond the realm of possible testing . That includes both climate and observational astronomy unless you happen to have a few planets or a galaxy or two in your shed.

[I don't think that makes sense. For one thing, all theories are made up of components. Not is it true that GW is untestable -W]

By Eli Rabett (not verified) on 10 Sep 2017 #permalink

Crandles, given diffusion the butterfly effect is useful for argumentation but not real. Sort of like all the oxygen in the room moving to one corner.

[Why would you believe that? There is no evidence that diffusion removes the effect -W]

By Eli Rabett (not verified) on 10 Sep 2017 #permalink

How many tower cranes were in Miami, FL at the time of Irma?

"As Hurricane Irma threatens to pound Miami with winds of mind-boggling power, a heavyweight hazard looms over the city's skyline: two dozen enormous construction cranes."
http://abcnews.go.com/US/wireStory/huge-construction-cranes-loom-miami-…

How much has Miami, FL spent on flood control?

Short term (near term) about half a billion $$$.
http://www.miamiherald.com/news/local/community/miami-dade/miami-beach/…
http://www.miamiherald.com/news/local/community/miami-dade/miami-beach/…

I'm kind of pi$$ed off that Miami, FL didn't get a direct hit from a Category XI hurricane though.

By Francis E Sargent (not verified) on 10 Sep 2017 #permalink

>"I’m kind of pi$$ed off that Miami, FL didn’t get a direct hit from a Category XI hurricane though."

Poor Miami, OTOH I could feel that about Mar-a-Lago.

I really don't follow why people think the butterfly effect is not real. On what scale does it not work?

Do you accept browning motion creates random movement of a grain of pollen? Does this work in the atmosphere? Does a grain of pollen sometimes cause an animal to sneeze? Does an animal sneezing sometime reveal that animal to another animal. Once you have altered animal behaviour, won't there be follow on effects that will lead to stampedes of animals at different times or in different directions? Is this big enough yet to talk about unstable convection in the tropics?

On what scale does the butterfly effect break down?

Any one particular path is remarkably unlikely. However if there are animals that do sneeze as a result of a grain of pollen lodging itself in an awkward position, then in due course all of these situations are affected and with a large number of such affected situations at least one of them is going to lead on to the next stage of up-scaling in the size of the difference.

[I've had this weird argument that "diffusion" elsewhere, but can't find it now. See-also http://mustelid.blogspot.co.uk/2005/10/butterflies-notes-for-post.html and http://www.realclimate.org/index.php/archives/2005/11/chaos-and-climate/ -W]

"observational astronomy unless you happen to have a few planets or a galaxy or two in your shed"

But, in fact, observational astronomy leads to improvements in theoretical astronomy which leads to predictions which lead to observational tests all the time... especially as technology improves and enables tests we did not previously have.

[But it doesn't make it science. A leads to B doesn't give A the properties of B. Making better instruments may lead to science, but may "only" be engineering -W]

Part of it is, of course, in creating the appropriate framing for the questions. E.g., for Harvey, the question might be "is the frequency of category 3+ hurricanes in the Atlantic changing" or "average rainfall of a cat 3 hurricane changing" or "is sea level rise increasing damages from a given strength hurricane" or "is there a change in landfalling hurricanes as a fraction of all Atlantic hurricanes" etc. etc., each of which is testable in part or in whole through some combination of observation, paleo-climate work, modeling, and theoretical analysis.

"[It seems reasonable to me, as I’ve said, that GW increased the strength and thus the damage potential of these hurricanes. Though as I also said, it’s associated CO2 comes in part from work that has made the USA more resilient to hurricanes, so the overall balance becomes unclear -W"

I grant that if we had never built a fossil-fuel fired power plant in the US, we'd likely still have 19th-century infrastructure & would have seen far more mortality impact from recent hurricanes (not property damage, because we wouldn't have had as much property to be damaged!) I'd still be interested in your thoughts on the question of "if we had started a reasonably climate mitigation policy at some date - say, the Carter administration - of a carbon tax & rebate system, increasing over time, would the net effect of such a climate regime have been to increase or decrease vulnerability to storms?" I would personally assign slightly more probability to more resilient infrastructure/society rather than less, but I could see arguments on either side.

[That one is much harder; you're asking what would have happened if we were a saner society, and that's hard to see the shape of -W]

From crandles: I really don’t follow why people think the butterfly effect is not real. On what scale does it not work?

How many times have you waved a fan and produced a high wind?

[This is a commonplace misunderstanding of the butterfly effect -W]

On the scale of a butterfly, the added energy is dissipated. It is a good question about when real change can happen.

The thing about the butterfly effect is that the butterfly's wing flap in Tokyo results in a tornado in Texas, as opposed to a typhoon in Taiwan or whatever else might have happened. It doesn't directly cause the bad weather, it just shifts its location and timing around in ways that are not predictable on time scales longer than a week or two. I'm simplifying here--air has nonzero viscosity, so dissipation will happen--but that was the point of Lorenz's paper.

The claims that you can't blame global warming for a single storm miss the point in a similar fashion. It's technically true that you can't attribute a particular storm to global warming. What global warming does do is shift the distribution, so that, for instance, you get a Hurricane Harvey in 2017 instead of a Tropical Storm Allison as you did in 2001. The former gives you 1300 mm of rain instead of the 1000 mm that you got from the latter. Either way your flood control systems are likely to be overwhelmed. You do get more water that has nowhere to go, so some places that didn't get flooded by Allison did get flooded by Harvey. (Greater Houston's land use practices didn't help, but that's a minor perturbation here.)

The problem is that the estimated frequency of certain events, which for anything on time scales of 100 years or more will almost always be extrapolated (there is a separate question of whether the extrapolations are valid because they will be sensitive to the frequency model you assume) based on the unchanged event distribution. Already, if your house is in a once-every-hundred-years flood zone, you have about a 26% chance of being flooded out during the course of a 30-year mortgage: p = 1 - (0.99)^30. If the distribution has shifted such that you are dealing with a 30 year frequency, that probability rises to 64%: p = 1 - (29/30)^30. Which means that flood insurance, if you can get it at all, is probably not as expensive as it should be. And without insurance, no banker is going to underwrite a mortgage on such a property.

By Eric Lund (not verified) on 11 Sep 2017 #permalink

"I think it was inevitable that Harvey and so on would be “blamed” on GW, but that if they hadn’t happened, that wouldn’t have cast doubt on GW."

The words after "but" are a bit of a non-sequitur. If people are suggesting that Harvey proves GW then you could suggest that its absence would disprove GW.

Whether Harvey is attributable to GW is debatable, but even if not (and from what I understand, probably not), that couldn't cast doubt on GW.

From 2007 IPCC
https://www.ipcc.ch/publications_and_data/ar4/wg1/en/ch10s10-es-13-trop…
Results from embedded high-resolution models and global models, ranging in grid spacing from 100 km to 9 km, project a likely increase of peak wind intensities and notably, where analysed, increased near-storm precipitation in future tropical cyclones. Most recent published modelling studies investigating tropical storm frequency simulate a decrease in the overall number of storms, though there is less confidence in these projections and in the projected decrease of relatively weak storms in most basins, with an increase in the numbers of the most intense tropical cyclones.

Doubt these numbers mean much but
Atlantic:
Season_______ ACE______TS_____HU____MH

1950-2016___ 101.16____ 11.25___ 6.22__ 2.66
1979-2016___ 103.32____ 12.26___ 6.39__ 2.61
2006-2016____ 98.09____ 14.09___ 6.36__ 2.73
2005-2016___ 110.75____ 15.25___ 7.08__ 3.08

East Pacific:
Season ACE TS HU MH

1971-2016 129.96 16.52 9.30 4.48
1979-2016 132.55 16.82 9.32 4.58
2006-2016 129.36 17.55 9.73 4.91
2005-2016 126.58 17.33 9.50 4.67

If anything numbers of Tropical storms and hurricanes seem up and even more debatably ACE possibly down.

So IPCC possibly looking wrong on both counts?
Yeah ok, back to these numbers probably don't mean much, if anything.

Still slightly surprising that numbers seem to show increase in number and little change or decrease in ACE?

crandles: 66-years is not enough time to ferret out a climate signal in acute weather events. I've been looking over the literature on paleo hurricane proxy studies. Some report natural cycles on 300 and 600 year frequencies. Others clearly show increasing # and strength with higher temps for data spanning the MWP-LIA-2000. Until we get a more complete multi-proxy study (the ones I've read admit the data is still thin) of paleo storms and intensity going back 2000-years, we only have the overly pixelated GCMs to make guesstimates. That's my take on why NOAA and IPCC aren't claiming attribution yet.

The stupendous hurricane season of 1780. In Wikipedia.

Hmmm...

By David B. Benson (not verified) on 11 Sep 2017 #permalink

More for completeness than any insight these numbers may offer:

Pacific Typhoon seasons look more in accordance with IPCC predictions:
SeasonTDTSTySTy

1971-201640.9625.8514.374.11
1979-201642.5325.6814.034.45
2006-201639.0923.4512.455.27
2005-201638.5823.5012.505.08

Still adding same strength categories for above basins I get
SeasonTSH/Ty

1971-201654.1529.80
1979-201654.7629.74
2006-201655.0928.55
2005-201656.0829.08

Still shows slightly more Tropical Storms and less Typhoon/Hurricanes.

300 or 600 year cycle seems unlikely to make much difference to the analysis but a 20 to 80 year cycle certainly could change the way these numbers look. 11/12 years is certainly not enough to draw a conclusion, but I still thought it worth taking a look to see how things seemed to be going.

Mitch wrote
"How many times have you waved a fan and produced a high wind?

On the scale of a butterfly, the added energy is dissipated. "

Every high wind is a consequence of trillions of interacting effects.

If the added energy is dissipated presumably you believe there are no high winds ever???

Some have suggested that the track of Harvey (the fact that it stalled) may have been due to climate change. If you want a testable hypothesis to bet on then this is maybe an easy one.

[You're joking, I hope. I suspect that teasing out exactly what "caused" Harvey to stall would be very difficult -W]

Trenberth said that this was done for Sandy (though I can’t find the papers) and found that the track was probably not influenced by climate change:

“One of the examples might be super storm Sandy where it recurved in an unusual direction back into the Jersey shore. There have been a number of experiments done with numerical models that have replicated the storm very well, and they’ve changed the sea temperature, they’ve changed the environment in ways that would be happening with climate change. The track of the storm is identical just about in every case.”

[That sounds very strange; I would expect the track to vary with SSTs; how could it be otherwise? -W]

>"Some have suggested that the track of Harvey (the fact that it stalled) may have been due to climate change. If you want a testable hypothesis to bet on then this is maybe an easy one.

[You’re joking, I hope. -W]"

I have seen this:

"Finally, the more tenuous but potentially relevant climate factors: part of what has made Harvey such a devastating storm is the way it has stalled near the coast."

That seems to be an opinion piece by Michael Mann
https://www.theguardian.com/commentisfree/2017/aug/28/climate-change-hu…

[I wasn't disputing the stall; only that it was an easily testable hypothesis -W]

Don't ask me, but that Trenberth guy seems to know a thing or two. I'll try to find the paper.

This isn't quite it: http://www.pnas.org/content/110/38/15211.full

[Trenberth knows what he is talking about. But the day-to-week shape of the upper level winds is not fixed, especially if you play with teh SSTs, so I find it hard to understand how the track could be unchanging. Unless of course you're only running short-term sims with fixed initial conditions -W]

"The stalling is due to very weak prevailing winds, which are failing to steer the storm off to sea, allowing it to spin around and wobble back and forth. This pattern, in turn, is associated with a greatly expanded subtropical high pressure system over much of the US at the moment, with the jet stream pushed well to the north. This pattern of subtropical expansion is predicted in model simulations of human-caused climate change."

I don't think anyone is disputing the actual stall. The 'tenuous' description certainly indicates this explanation is not entirely certain or complete and could change with further research.

so "teasing out exactly what “caused” Harvey to stall would be very difficult" no quibble from me on that.

The part I am interested in is whether blaming cc for worse Harvey damage/cost is coherent. Is it possible that increased stalling means fewer landfalls such that cc means fewer such disasters but worse when they do occur. Could this climate induced change actually be good rather than bad and we were just unlucky with Harvey?

I agree it is difficult to test either the explanation for more stalls or whether good or bad. In the absence of evidence as to whether good or bad, I would probably tend to cautiously fear it was bad. Is that enough for such an opinion piece pushing towards blaming Harvey on climate change with this explanation?

I guess a bet on number of landfalls or frequency of stalling if that could be defined would need to run a long time before it is in any way interesting and that length of time then makes it uninteresting as a bet.

The Newsweek article, though, is too un-nuanced in its accusations to be much use.

Since the article's function is propaganda, this isn't much of a criticism.

"because there are so many lines of evidence for GW that it isn’t possible to honestly doubt it."
AGW or GW.
There is GW and GC, undebatable.
There is GW in a test tube planet with an increase in CO2,
Then there is the earth with seas, clouds and water vapour and ice.
Which comes first? The CO2 or the temp rise.
CO2 must rise with a warming planet naturally.
Adding some in should make it a tetch warmer.

" the relationship between hurricanes and GW."
Some say more, some say less.
Increasing damage, such a shonky measure.
A bit like someone hearing a tree fall in the woods if no one is there.
Measuring the strength of a hurricane in terms of human "damage" is ridiculous.
If no humans were there there would still be damage.
The more humans you put there the more human damage.
How many humans in Houston 200 years ago, and now.
Increased human damage of course. Meaning? None.
For the strength of the Hurricane that is.
It is a baloney after the act, how bad can we make something sound, puerile factoid.
-
Worse we all know this but persist in arguing.

crandles 2017/09/11
"I really don’t follow why people think the butterfly effect is not real. On what scale does it not work? Do you accept browning motion creates random movement of a grain of pollen? Does this work in the atmosphere? Does a grain of pollen sometimes cause an animal to sneeze? Does an animal sneezing sometime reveal that animal to another animal. Once you have altered animal behaviour, won’t there be follow on effects that will lead to stampedes of animals at different times or in different directions? Is this big enough yet to talk about unstable convection in the tropics?
On what scale does the butterfly effect break down?
Any one particular path is remarkably unlikely."

Love your argument, very well put. Sort of how I thought about it but could not articulate.

However there are other considerations in blame games. One of which comes to mind is that the butterfly is not unique or individual but it itself is only part of the mosaic of events that are all intertwined.
Saying that a pathway attributable to the butterfly and the butterfly only ignores all those other exiguous causes [did I make this word up] which also impacted on the hurricane.
For instance every other butterfly , person etc in the world also happened and the same contiguous line must be drawn for all those other effects.
I think the butterfly effect is a real correspondence but a fake cause.

Thank you and yes absolutely it isn't a blameable cause because a) you can't predict it in advance and b) there are a trillions of other movements that are also needed to get the one path that does play out to occur c) any different motion would still cause a similar frequency and severity of tornadoes in tornado alley and so on just a different timing sequence.

It only *appears* to be a traceable cause where you consider the world with a butterfly flap and without it. In reality, you should equally consider each one of the trillions of effects and end up concluding they are all needed.

So I think I end up saying it is only a joint cause with trillions of other such causes and certainly does not carry any responsibility.

So:
"How many times have you waved a fan and produced a high wind?"

Every high wind ever (after a month after my birth) is caused by me but don't blame me I have no responsibility and it is only joint causation with trillions of other causes.