Let’s try this again: It is nonsense to say that we don’t know whether a specific weather event was affected by climate change. It’s not just wrong, it’s nonsensical.

This post is by Phil Price, not Andrew.

If you write something and a substantial number of well-intentioned readers misses your point, the problem is yours. Too many people misunderstood what I was sayinga few days ago in the post “There is no way to prove that [an extreme weather event] either was, or was not, affected by global warming” and that’s my fault.  Let me see if I can do better.

Forget about climate and weather for a moment. I want to talk about bike riding.

You go for a ride with a friend. You come to a steep, winding climb and you ride up side by side. You are at the right side of the road, with your friend to your left, so when you come to a hairpin turn to the right you have a much steeper (but shorter) path than your friend for a few dozen feet. Later you come to a hairpin to the left, but the situation isn’t quite reversed because you are both still in the right lane so your friend isn’t way over where the hairpin is sharpest and the slope is steepest. You ride to the top of the hill and get to a flat section where you are riding side-by-side.  There is some very minor way in which you can be said to have experienced a ‘different’ climb, because even though you were right next to each other you experienced different slopes at different times, and rode slightly different speeds in order to stay next to each other as the road curved, and in fact you didn’t even end up at exactly the same place because your friend is a few feet from you.  You haven’t done literally the same climb, in the sense that a man can’t literally step twice in the same river (because at the time of the second step the river is not exactly the same, and neither is the man) but if someone said ‘how was your climb affected by your decision to ride on the right side of the lane rather than the middle of the lane’ we would all know what you mean; no reasonable person would say ‘if I had done the climb in the middle rather than the right it would have been a totally different climb.’

You continue your ride together and discuss what route to take where the road splits ahead. One road will take you to a series of hills to the north, the other will take you to a series of hills to the south. You decide to go south. You ride over some hills, along some flat stretches, and over more hills. Three hours into the ride you are climbing another hill, the toughest one yet — long, with some very steep stretches and lots of hairpin turns. As you approach the top, your riding companion says “how would this climb have been different if we had gone north instead of south?”  What is the right answer to this question? Here are some possibilities: (1) “There is no way to prove that this climb either was, or was not, affected by our decision to go south instead of north.” (2) “The question doesn’t make sense: we wouldn’t have encountered this climb at all if we had decided to go north.” (3) “This climb was definitely affected by our decision to go south instead of north, but unless we knew exactly what route we would have taken to the north we can’t know exactly how it was affected.”

1 is just wrong (*).  If you had gone north instead of south you might still had a steep climb  around hour 3, maybe it would have even been a steeper climb the one you are on now, but there is no way it could have been the same climb…and the difference is not a trivial one like the “twice in the same river” example.

2 is the right answer.

3 is not the right answer to the question that was asked, but maybe it’s the right answer to what the questioner had in mind. Maybe when they said “how would this climb have been different” they really meant something like, if you had gone the other way, “what would the biggest climb have been like”, or “what sort of hill would be climbing just about now”?

I think you see where I’m going with this (since I doubt you really forgot all about climate and weather like I asked you to).  On a bike ride you are on a path through physical space, but suppose we were talking about paths through parameter space instead. In this parameterization, long steep climbs correspond to hurricane conditions, and going south instead of north corresponds to experiencing a world with global warming instead of one without. In the global warming world, we don’t experience ‘the same’ weather events that we would have otherwise, but in a slightly different way — like climbing the same hill in the middle of the lane rather than at the side of the lane — we experience entirely different weather events — like climbing different hills.

The specific quote that I cited in my previous post was about Hurricane Katrina. It makes no sense to say we don’t know whether Hurricane Katrina was affected by global warming, just as it would make no sense to say we don’t know whether our hill climb was affected by our decision to go south instead of north. In the counterfactual world New Orleans might have still experienced a hurricane, maybe even on the same day, but it would not have been the same hurricane, just as we might encounter a hill climb on our bike trip at around the three-hour mark whether we went south or north, but it would not have been the same climb.

No analogy is perfect, so please don’t focus on ways in which the analogy isn’t ‘right’. The point is that we are long past the point where global warming is a ‘butterfly effect’ and we can reasonably talk about how individual weather events are affected by it. We aren’t riding up the same road but in a slightly different place, we are in a different part of the territory.

(*) I’m aware that if you had ridden north instead of south you could have circled back and climbed this same climb. Also, it’s possible in principle that some billionaire could have paid to duplicate ‘the same’ climb somewhere to the north — grade the side of a mountain to make this possible, shape the land and the road to duplicate the southern climb, etc.  But get real. And although these are possible for a bike ride, at least in principle, they are not possible for the parameter space of weather and climate that is the real subject of this post.

This post is by Phil, not Andrew.

49 thoughts on “Let’s try this again: It is nonsense to say that we don’t know whether a specific weather event was affected by climate change. It’s not just wrong, it’s nonsensical.

  1. > we can reasonably talk about how individual weather events are affected by [global warming]

    And what can we reasonably say? That if there had been more or less or somewhat different global warming those individual weather events would have been different. Also, if there had been more or less or somewhat different solar activity, or more or less or somewhat different volcanic eruptions, or more or less or somewhat different nuclear explosions, those individual weather events would have been different.

    • Carlos, I’m not sure what you’re saying here. What I said is “we are long past the point where …we can reasonably talk about how individual weather events are affected by [global warming’].” I don’t think it makes sense to talk about whether or how an individual weather event was affected by global warming. Are you disagreeing with me? You think it does make sense to ask that question?

      • The point is that we are long past the point where global warming is a ‘butterfly effect’ and we can reasonably talk about how individual weather events are affected by it.

      • My mistake, I parsed your comment as:

        > The point is that [we are long past the point where global warming is a ‘butterfly effect’] and [we can reasonably talk about how individual weather events are affected by it].

        • Hahaha maybe I should edit it…although to be fair (to me), the way you parsed it wouldn’t make any sense. We are in strong agreement.

  2. Discussion of “causality” always generates endless discussion like this. It is just like statistically significant differences/correlations in that I have never once seen something of value result. That should tell you there is something wrong with at least one fundamental premise behind such ideas.

    • Anoneuoid, I’m not sure what ideas you are disagreeing with.

      Do you think it does make sense to talk about whether a specific weather event was affected by global warming?

      • Any specific weather event was effected by everything that happened earlier. More warming than actually occurred, less warming… any different past would result in different weather events.

        • Yes. And because of that, it makes no sense to ask whether a specific weather event was affected by global warming. In general that weather event wouldn’t have happened at all in a counterfactual no-global-warming world. That’s why I find it so irritating that people say “no one can say if this weather event was affected by global warming.” What do they even think they mean by that?

        • They think of observed weather events as samples from a distribution of possible events. Hurricane Katrina would not be an unusual observation if sampled from whatever distribution they have in mind for the “no warming” world either. Since these two “models” are both consistent with a Katrina, it cant be used to distinguish between them.

        • This is more or less what someone said about the last post. I think you are (both) probably right that that’s more or less what people mean. But it is not what they say, and I do not think what they say is helpful. What they should say is what you just said.

  3. What if my friend and I go home and simulate our bike rides, sit down with our highly detailed topographic maps, and simulate a few thousand hypothetical bike rides along the two different routes. Sure these wouldn’t be perfect replications caused we’d be missing all sorts of details, like an erratic headwind, or the squeak from your bottom bracket that you can’t seem to make go away and gets really, really, REALLY annoying around mile 73. But we can get a good amount of details correct, for example, we’d have a range of elevation profiles and expected wattage expended.

    Suppose we take that universe of simulations along the southern route and compare it to the data from Strava of our actual rides (assuming the measurement error is negligible, which if you’ve ever used Strava you know that’s a bad assumption). We will probably find a good set of cases in that universe that more or less match what we observed on our actual ride. We will also have a set of cases that with more brutal climbs and some with easier climbs. We can state relative to our simulations how likely it would be to experience a climb at least as brutal as the one we actually experienced. Critically we can also see where that our climbs ranks among the climbs in the simulations from the route we did not travel. I would say, that given this exercise, we actually could say something about how our choice of route affected the climb we experienced because we can say with some degree of (un)certainty how probable a climb of that intensity was on the southern route and compare it to the probability of experiencing a climb of similar intensity along the northern route.

    “But wait!” my friend says, “I don’t trust your model. When we actually went on our ride, I got a flat and I wasn’t able to reinflate my tire all the way with this stupid flimsy emphysemic handpump and when we were huddled on the side of the road, me pumping furiously for tiny gasps of inflation and you picking your Lycra wedgie, that big red F250 with the Confedarate flag bumperstickers rolled coal on us. Surely these things affected the intensity of our climb and they’re completely missing from our model.” So, being the credulous and conscientious modeler that I am, I go back and collect data on the historical likelihood of flats tires and the density of yokels and I spend a lot of time improving the model. Unbeknownst to me, my friend has a ulterior motive, you see there is this fish and chip shack on the southern route that he really enjoys and he knows he should give it up because it’s bad for his health, but those are worries for the future, the fish and chips are there now! So when I present my new results he says: “But wait! What about the increase in air resistance from my leg hair growing back in over the course of the ride?! I still don’t believe your model! We still can’t say that the southern route made that climb more likely!”

    So does the questions still not make sense? Or is my friend, not my friend?

      • So the question does make sense? Or at least we can phrase it to make sense?

        How much more likely was a storm of the intensity of Hurricane Katrina striking the Gulf Coast in a simulated world with CO2 at 380 PPM versus a simulated world with C02 at 300 PPM? That question makes sense. And the answer to that question is very far from irrelevant.

        How much more likely is to die from a heart attack if you take up smoking versus taking up cycling? Sure we can’t point to any particular heart attack and say it was CAUSED by smoking or by cycling, but we can certainly say one of these past times increased the likelihood of heart attacks. If you and your friend both smoked, and your friend dies from a heart attack, and a doctor tells you “Smoking X cigarettes a day as your friend did made a heart attack of this severity 80 – 200% more likely,” and you say “But you can’t say smoking CAUSED that particular heart attack that claimed my friend’s life at Crazy Willies Fish and Fries!” then the rhetorical you is an idiot.

        • Dalton, yes, all the questions you pose are important.

          Maybe you didn’t click through to read the initial post that prompted this one. Basically I am sick of the fact that every news item in the past fifteen years about a catastrophic weather event has had some sort of statement like “it’s impossible to say whether this specific event was influenced by global warming” (or was ‘due to global warming’ or was ‘affected by global warming’). That statement doesn’t mean anything. It’s closer to being nonsensical than it is to being wrong, for reasons discussed in this post.

          There are tons of important questions about climate change. This post isn’t about any of them. It is just an extension of my previous post in which I vent about how frustrating it is to see people spouting nonsense, even people who really really should know better.

        • Phil,

          I suspect that you and I mostly agree on the right way to characterize the relation between climate and any particular weather event. And we probably agree that most reporting gets it wrong. But we disagree on how it gets it wrong. You remember every news item saying “It’s impossible to say that this was influenced by global warming.” I remember a lot of reporting saying “This is all because of global warming” with someone very, very occasionally being brought on to say “It’s not quite that simple”.

          One person’s infuriating nonsense can be another person’s background noise.

          MP

    • Critically, the question you’re asking is about risk rates “how often would we have had to put out this much wattage in mile 30 if we choose left at mile 10 vs if we choose right at mile 10”

      These are not questions about “how would choosing to go left at mile 10 have altered the road rash I got when I fell on mile 20 near the fish and chips shack?”

      The answer is you wouldn’t have fallen near the fish and chips shack because you wouldn’t have been anywhere near the fish and chips shack.

      • Daniel, true enough. But the fish and chip shop is at mile 60. Nobody eats fish and chips at mile 20. That leaves you with a loooong way to go on a stomach full of grease. Which is a coy way of saying, sometimes getting caught up in the fact that we live in a unique world and that time is unbroken arrow is simply a beer-battered red herring.

        This has relevance to climate change and event attribution. We can look at the SOME of the measurable particulars of a given event and use that as a benchmark. With a sophisticated enough model, we can say how much more likely it is that we observed those particular measurables in a world with climate change versus a world without climate change. We can’t say that Hurricane Dorian exactly as we observed it was caused by climate change. We can say something about the likelihood of category 5 hurricanes in mid-September in warming world.

    • “big red F250 with the Confederate flag bumperstickers rolled coal on us”. Hmmm, quite the lurid description and stereotype. All part of your nice, descriptive metaphor of course. Just on a serious note, though, as a Cat 1 roadie, I’ve spent thousands of hours riding bikes on rural roads in the South. I actually found that the majority of times that I had a flat, people driving by stopped to ask if I needed help.

      • Cat 1! I’m impressed.

        I used to do some riding around Lexington, Ky. Lots of narrow roads with no shoulder over rolling hills. In general were great about not overtaking on blind hills and curves, way better than here in the Bay Area. It seemed to me that people towing trailers (often horse carriers, in that area) were especially great about it. It didn’t seem like they were seething in frustration as they waited for a chance to pass, more like “I know what it’s like when your speed and acceleration are limited, don’t worry about it, I’ll be as patient with you as I wish people were with me.”

        But the outliers were terrible, very hostile and threatening. I am especially not a fan of the ones with confederate flags. Some stereotypes do represent accurate informative prior distributions, you just have to remember that the distribution is rather broad and the stereotype is just the central tendency.

  4. I’d like to expand on the bike riding metaphor, I like it a lot. Here’s the concept of sensitive dependence… a “butterfly effect”.

    Suppose you start at the top of a hill and you start descending the hill, your friend starts right behind you in basically the same place but just slightly moved to the right. As you guys bomb the hill you gain so much speed that it’s hard for you to track through the turns in a similar way, you start to oscillate so that you are taking one turn tight, and the other wide, your friend is maybe wide on the first one, tight on the next, you’re weaving back and forth bombing the hill on the edge of control. All of a sudden as you reach about half-way there’s a fork in the road… You are in a place where it’s impossible for you to take the turn to the left without killing yourself by hitting a tree, so you go right, your friend, who is on a different track can’t take the right turn, and he goes left. Instead of slowing down and catching back up to each other you just continue taking your paths… twenty minutes later you’re miles apart on different sides of the hill, one of you is in bright hot sun on a desert slope that’s rocky, the other is in the shade of an oak forest on the north side shaded by the mountain…

    the tiny difference in where you started got amplified through time so that eventually you’re in totally different places that are completely un-alike…. that’s the butterfly effect. It isn’t so crazy that suddenly you found yourself riding a horse in a clown suit, and your friend was on a unicycle dressed in riding jodhpers, it’s still bounded by the reachable possibilities of “riding a bike down this hill” but as far apart as you could get within the space of “riding down the hill”… you achieved that separation even though you started in basically the same place and the same time.

    • Daniel,
      Thanks for noting the metaphor, it took me a long time to think about it, and I hadn’t thought about the fact that it can be used to talk about the butterfly effect too. Your example of that does work, but although not a stretch to people who are used to bombing downhill on a mountain bike through rough terrain I think it’s not something most people can relate to very well. But that’s a fault of using bike riding in the first place, you did what you could with it! And as I say, it does work.

  5. I think a better analogy would be this: You have been driving along a paved road for 100 miles, and you got one chip out of your paint from a stone that flew up. Now you drive off the pavement onto a gravel road. In the first ten miles, you get ten new chips.

    You can’t say that any one of the new chips was the result of getting onto the gravel road. After all, you did get that one chip in 100 miles of paved driving. But you know darn well that the (change in the) aggregate effect – lots more stone chips – *was* the result of driving onto that gravel road.

    You know this because the statistics and the explanation (the theory) both tell the same story – and you probably saw a lot of stones flying around, too.

    Coming back to weather events and the climate, people may *say* “wow, strong hurricane, must be climate change”. They don’t mean anything very technical or precise by it. They just, by and large, mean that they think they’ve gotten onto the gravel road and wow, look, lots of flying stones. There’s really nothing much more to parse out.

    • Well, being a resident of Houston I offer the following.

      In 2005 Hurricane Rita made its way from the Atlantic into the Gulf of Mexico. Katrina not long before had of course set everyone on edge (at least those who have never been to New Orleans and seen the walls holding back water levels up to 8′ above peoples’ back yards) and the media whipped (almost, not us, we stayed home and hosted numerous would-be evacuees) everyone into a frenzy and put most of south and southeast Texas on the roads fleeing angry Mother Nature who it was said would no longer be fooled by either margarine or CO2.

      The hurricane itself took just a handful of lives but in one incident alone 23 people burned to death in a bus on Interstate 45 which had been turned into a parking lot by all the people who heeded the talk of imminent environmental apocalypse.

      For the next decade NOAA serially issued predictions of impending hurricane doom and yet things have been happily silent but for Hurricane Harvey which, had it not been for the idiots manning the dams who didn’t open the flood gates until they were almost over-topped and thereby promptly flooded everyone downstream, would have been a NBD.

      Thus for a decade NOAA warned that we were flying down a gravel road and yet we got fewer chips than when on asphalt. What do you make of that?

      • You have just switched from the original subject to another one. Apparently, the new subject is that NOAA has supposedly cried wolf for a decade. I sense that’s supposed to show that there really isn’t any climate change going on, although that leap is quite a stretch in itself. Anyway, it’s a larger subject, and I’m not engaging with that here.

        But you might remember the outcome of the fable…

        • there really isn’t any climate change going on

          I don’t think this is any better than “you can’t say Katrina was affected by climate change”. There is always climate change going on…

        • I don’t think Thanatos is intending to say that climate change isn’t going on, his point seems to be more that we have very little ability to predict any kind of specific outcomes on the basis of our level of knowledge of climate change, and so it’s hard to make decisions on this knowledge (ie. you might decide to tell everyone to evacuate a region, but it actually leads to substantially more deaths on the highways than the storm would have caused).

          Just because we know that climate change is happening, and even maybe that *warming* is happening, and even maybe that a certain range of warming is happening, doesn’t mean we know what to do about it.

        • Yes. The warming trend appears obvious to me and the properties of CO2 undeniable. But they are not the only causes of climate effects and yet too many well-intentioned people treat it as a promptly solved “toy problem” whose QED is promptly followed by WTF as the particular predictions entailed serially refuse to manifest.

        • Thanatos: “The warming trend appears obvious to me and the properties of CO2 undeniable. But they are not the only causes of climate effects ”

          Daniel: “we have very little ability to predict any kind of specific outcomes…so it’s hard to make decisions on this knowledge”

          Absolutely. 100% on both counts.

  6. I’m always a fan of bike stories!
    Nice metaphor.
    I was tempted to ask about that 1 extra Carbon atom on the side of the tire on the friend’s bike in Daniel Lakeland’s extension of the metaphor…but I’ll take a different fork in the road:-D

      • I have! I remember it being very well-written. Perhaps I should re-read it though, as it has been many years. If I remember correctly, the entire book is essentially the thoughts and experiences of a rider while he is riding a race. The perspective, or voice, that it was written from is what made it so good, if I am thinking of the same book.

        • Yes, that’s what happens in The Rider. Worth another read, I think. It’s what got me into riding, in a way.

          My whole adult life, I’ve ridden my bike for transportation. Biking to classes when I was in school, commuting to work by bike once I had a job. For the first three years. Lived in the Bay Area I didn’t own a car. And of course I occasionally rode for recreation, had even done a few centuries, biked up our local mountains, etc. But recreational riding was definitely the exception rather than the rule.

          But in 2009 I read The Rider and loved it. The next year I went to visit Andrew while he was on sabbatical in Paris and I decided that I after The is it I would go to the Massif Centrale, rent a bike, and ride the route of the Tour de Mont Aigoual as described in the book. Unfortunately I sprained my Achilles while doing a track workout a couple of months before the trip, so I couldn’t train at all. I’ didn’t even start riding again until a couple of weeks before my trip. I did it anyway, though: took the train to Avignon, rented a bike and a car and drove to Meyruis (the start town), and the next day I did the ride. Absolutely loved it. I came back home and bought a road bike and now I do recreational rides all the time.

        • Wow, that is quite the read if it inspired you to make the effort to complete the same ride! That’s a cool story. I’ll have to read it again.
          I just ride for fun and exercise now, but I loved racing. It’s much more complex and dynamic than a casual observer unfamiliar with the sport might think. It is extremely tactical and team oriented. I was never great, but I was a pretty good amateur, so I was able to ride a few top US amateur events (Cat1 and Cat2 at NRC races), as well as a few of the pro races on the NRC circuit when the top US pros used a few for training before going to Europe. It was an interesting experience.

    • Since you hinted, I’ll explain. Not all deviations have the same Lyapunov exponent. If there is more than one deviation, then whichever one has the largest exponent will be the one that dominates the deviation.

      the effect of the single carbon atom could only be thought of by cloning the world and re-running the bike ride with just that one carbon atom missing… Suppose that the Lyapunov exponent for this deviation is 1/million-years. Then on the time scale of 4 or 5 million years the deviations get large (exp(4) or exp(5) times the size of the original deviation), but on the time scale of 1 day they remain negligible exp(1/365/1000000) ~ 1 times the size of the original deviation.

      In any real world scenario there are always multiple things different between two runs of a macro-scale experiment. Whichever one has the largest Lyapunov exponent will be the one whose effect dominates.

      • Thanks. I mainly said that for Phil’s benefit, considering that this was a “Let’s try this again” post;)
        This sounds more along the lines of how I was thinking in the previous discussions. I maintain some skepticism about those previous examples, but certainly I can see how the math works in simulations.
        As always, I appreciate the thoughtful replies to questions on this blog!

        • Yeah, a single atom perturbation is most of the times likely to have a small associated lyapunov exponent (or maybe even a zero or negative exponent… so there’s no instability). But in any real-world system the smallest perturbations we can do are literally quadrillions of atoms or more… and there are many many many of them, so that not only is the tire different, but the grains of sand on the road are different, the trees have grown, leaves have fallen off, animals are or are not out foraging along the road, etc etc. Whichever of these causes the fastest growing kind of perturbation (largest lyapunov exponent) will be the one that dominates, the others made a difference, but as time goes on that difference doesn’t matter, in the same way that exp(10*t) + exp(2*t) might as well be exp(10*t) once t is bigger than about 0.5 or so.

  7. Phil: if you can’t detect an effect, it isn’t there. That’s reality. That’s “sensical”. The simple fact is that the temperature change hasn’t been large enough to have an identifiable effect on tropical storms.

    I don’t know why you find this so unbelievable.

    • if you can’t detect an effect, it isn’t there. That’s reality. That’s “sensical”.

      This would explain a lot of of academic hubris it is a widely held belief. Just spend more money to increase the sample size or you can just increase the alpha threshold… then you get to redefine reality. Sounds great!

    • Jim, I find it unbelievable because it isn’t even close to being true. I have a hard time believing things that I know to be false. The Red Queen has tried to teach me, but I’m a poor student in that way.

      • Your argument strikes me as the same kind of argument that’s used to justify all manner of nonsense in education: “x kind of teaching is so awesome it can’t possibly be detected by any kind of testing or measurement!” Wow, that *IS* good! :)

        In the end, the evidence in education is similar to the evidence regarding storms and climate: the larger system of development overwhelms most teaching interventions, just like the larger system of weather overwhelms the small changes in temperature when it comes to tropical storms (so far, at least).

        In teaching, just like in climate, there are also *so many* people *so committed* to beliefs that the few things that do work are rejected for not conforming to beliefs (e.g., any kind of repetition like direct education is odious to the teaching community, just as nuclear power and natural gas are odious to greens).

        • Once again, whoosh Phil’s point goes zooming over others heads…

          Phil has never once made any complaint about statements regarding the rate or severity of hurricanes. He is complaining that journalists talk about SPECIFIC events and claim it’s impossible to know how climate change affected the event. It isn’t even close to impossible. we know for CERTAIN that specific events are sensitive to just about EVERYTHING. Katrina WOULD NOT HAVE HAPPENED if for example a forest fire had erupted a year earlier in Sumatra, or a coal power plant had a mechanical failure and had to shut down in Toledo Ohio…. or any number of things.

          that’s just how the weather is, it’s sensitively dependent on everything.

        • Yes to Daniel’s summary of my point, no to his specific examples. (I’m not sure a forest fire a year earlier in Sumatra would have affected the next year’s hurricane season in the gulf enough to prevent ‘Katrina’, although it would have been a different hurricane in the “different path up the same road” way. Or maybe I’m wrong and a forest fire in Sumatra would have been enough, propagated forward a year.

          But the phenomenon of global warming is much much bigger than a single forest fire in Sumatra. We are definitely in a different region of parameter space than if we hadn’t already experienced some anthropogenic global warming.

          Whatever. jim, you think I am making an argument I am not making. I am not saying any particular hurricane season stands out so far from those of the past that one can discern the influence of global warming from the statistics. Try re-reading this post and see if you can figure you what I _am_ saying. If you can’t figure it out, I can’t help you, I’ve said it about as clearly as I can.

        • Of course I’m just guessing about the size of the perturbation after a year, but given that we can’t predict the weather more than a week out, and that ensemble forecasts of the same hurricane differ dramatically after a week, I’m guessing a year of dynamics is enough for any forest fire or major power plant to dramatically alter the patterns of individual storms. But as you say, climate change is way bigger than a single forest fire.

  8. Saw this article and thought it might be of interest to some:
    T. Knutson, S.J. Camargo, J.C.L. Chan, K. Emanuel, C. Ho, J. Kossin, M. Mohapatra, M. Satoh, M. Sugi, K. Walsh, and L. Wu, “Tropical Cyclones and Climate Change Assessment: Part I. Detection and Attribution”, Bulletin of the American Meteorological Society, 2019. http://dx.doi.org/10.1175/BAMS-D-18-0189.1

    This is the first climate report I have read and it was eye opening (not in a good way.) If this is the standard practice, their field could really use some updates in statistical methodology. The reliance on expert polling to form summary assessment is weird to me too.

Leave a Reply to Anoneuoid Cancel reply

Your email address will not be published. Required fields are marked *