Temperatures have soared wildly over Antarctica during the past week, with the eastern part of the continent experiencing temps about 70 degrees above normal (i.e., about 10°F instead of -60°F). That sounds pretty spectacular, but the technical description is even more spectacular:
The warm conditions over Antarctica were spurred by an extreme atmospheric river [that] made landfall on Tuesday between the Dumont d’Urville and Casey Stations and dropped an intense amount of rainfall, potentially causing a significant melt event in the area.
The moisture from the storm diffused and spread over the interior of the continent. However, a strong blocking high pressure system or “heat dome,” moved in over east Antarctica, preventing the moisture from escaping. The heat dome was exceptionally intense, five standard deviations above normal.
A five-sigma event happens only roughly one in a million times. I don't know how many heat domes Antarctica typically gets in a year, but probably not many. That means the current one is a million-year event—or maybe a 2-million-year event or a 500,000-year event.
This is ridiculously unlikely, of course, and a much better hypothesis is that climate change has pushed temperatures up enough that a heat dome like this is now more like a two-sigma event, one that we'll see every few years.
But don't worry. After we've melted Antarctica away, I'm sure we can buy a new one on Amazon for an astonishingly low price and next-day delivery.
We just had one-in-500 year floods in Australia, a year after one-in-100 year floods and two years after one-in-500 year bushfires. The kids these days can look forward to enjoying the rest of their lives free of natural disasters!
Actually journalists are struggling to write articles explaining why a "one-in-100 year event" doesn't mean it only happens every 100 years, which makes one wonder why climatologists use such expressions in the first place when what they mean is there's a 1% chance of the occurrence every year. Probabilities that are being urgently revised.
Now that you've brought Antarctica up, given how atmospheric circulation works, Antarctica may be the safest place to live through a global nuclear war. I'm still drinking myself to sleep if the bombs start dropping, tho.
The models showed warming in the Antarctic could lead to more precipitation. Initially, that would give more snow and potentially build up the glaciers in the middle of the continent while the edges melted.....but rain would be bad....
Here's the NPR write up:
https://www.npr.org/2022/03/19/1087752486/antarctica-record-heat-arctic
Looks like they are not sure if this was just a freakish happening or if it portends bad things ahead. We have enough bad news about Antarctic melting that it is a dire emergency with or without this latest heat wave.
"A five-sigma event happens only roughly one in a million times. .."
I suspect there is no reason to believe these events are normally distributed making this calculation meaningless.
If any random distribution(s) applies in this case it would be to a first approximation Poisson.
Poisson would be a first guess, but there is no reason to assume randomness on any scale. There are all sorts of cyclical and other non-steady-state phenomena in nature. See more below.
Only it is not clear to what the ‘five-sigma’ label is applied. Is it to the temperature of the heat dome? The pressure? Either could very possibly be normally distributed (though describing something as a ‘heat dome’ indicates that only the positive deviations are being considered). It does not appear that ‘five-sigma’ is being applied to the occurrence of the heat dome, as the description of it as ‘exceptionally intense’ implies that there have been others with which to compare it.
Over East Antarctica, no less. That's supposed to be the "secure" one, compared to the less stable West Antarctica.
Now, let's not hear any of those nasty lefties attacking that nice Mr. Manchin for trashing electric cars or joining with Republicans to veto Sarah Raskin's nomination to the Fed for the unspeakable crime of " arguing that financial regulators should pay more attention to the risks associated with climate change."
https://www.theguardian.com/environment/2022/mar/15/manchin-reluctant-electric-cars-biden-climate-crisis-fight
https://www.npr.org/2022/03/14/1086456895/sen-joe-manchin-sarah-bloom-raskin-nomination-fed-climate
I had always assumed that his opposition to BBB was mainly about the climate provisions, i.e. it was about his fossil-fuel donors. Though he had a record of deficit hawkery that he could invoke to cover for his self-interest.
We'll finally find that secret Nazi airbase and have lots of new beachfront property where tag line can be something like "Daylight Surfing 24 hours a day." (at the right time of Summer of course...)
I'm moving to a balmy retirement community in Canada called New Tundra Terrace, once the polar bears are out of the way.
Wheeeee.....
https://climatecrocks.com/2022/03/19/arctic-ice-loss-staggering-in-2021/
It seems like we’re getting a preview of our future ahead of climate change time… pandemic, war, shortages, refugees, soaring prices…
“The war in Ukraine has delivered a shock to global energy markets. Now the planet is facing a deeper crisis: a shortage of food. A crucial portion of the world’s wheat, corn and barley is trapped in Russia and Ukraine because of the war, while an even larger portion of the world’s fertilizers is stuck in Russia and Belarus. The result is that global food and fertilizer prices are soaring.”
The end isn’t near, but the decline is well underway.
Neon shortage!
https://www.vox.com/recode/22983468/neon-shortage-chips-semiconductors-russia-ukraine
We’re going to have to figure out what the economic system looks like in a world with these resource constraints. Climate change will make all of them worse. There will be no soft landing in a green field.
There's plenty of neon in every crummy neon light on Earth.
Our plan is pretty obvious:
1) Melt Antarctica down completely.
2) Reveal the various UFOs that have been locked under the ice for eons.
3) Utilize their advanced technology to fix the climate. Presto!
Five sigma a problem?
Just lean six.
They have 65 years of records - how do they know what a million-year event is, on the scale of weeks? How does anybody know what even a thousand-year event is on this scale? This is guesswork - they have to assume that they know the form of the time-temperature distribution, and then extrapolate way out. Real science requires verification, which they don't have. The same thing applies to claims about the heat wave in the Pacific Northwest and a number of other phenomena - investigators just do not have enough records on this scale to make the statements that they do. Thermometers were only invented a few hundred years ago, so it will be another few thousand before we really know what the time-temperature distribution is anywhere (and if it changes we can't measur what it was). Nor can anyone reliably say how the distribution would change when the mean temperature goes up a couple of degrees. All sorts of global weather patterns will probably be changing.
We do have measurements on the scale of years (not weeks, that I know of) going back hundreds to millions of years with ice cores, tree rings and some other paleotemperature measurements, so conclusions made on this basis would be more reliable. There is plenty of evidence of global warming - we don't need this kind of wild claims.
Good job.
In the first-place, the quoted text does not say anything about million-year events, that was Kevin’s extrapolation. Second: yes, lots of nonrandom factors influence temperature, but the very fact that the number of factors is large makes it more likely that the distribution is Gaussian. Third, if It the authors are saying ‘if present conditions obtained, we would expect an observation this extreme only one time in a million’. then we don’t need nearly as many observations as you think. Just as pollsters can come within a few percentage points by asking a thousand people in a nation of 150 million voters. Mean and variance converge faster than you think, and even distribution fits don’t take as much data as you imagine.
"... Third, if It the authors are saying ‘if present conditions obtained, we would expect an observation this extreme only one time in a million’. then we don’t need nearly as many observations as you think. Just as pollsters can come within a few percentage points by asking a thousand people in a nation of 150 million voters. ..."
Sampling error in polls is known to be normally distributed. I doubt there is good reason to believe that whatever these authors are measuring is normally distributed especially at the tails.
Scientific temperature measurement of air and ocean have been conducted for a couple of centuries. I would assume by now that there is enough data to estimate the distribution quite well. Similar for atmospheric pressure.
The "few percentage points" in polls you refer to are the theoretical *minimum* errors. Pollsters have blown it in the last couple of Presidential elections - the actual results were very improbable according to them. The assumptions that go into real-world measurements are usually not strictly valid. This is why methods of prediction have to be tested empirically, and such tests are absent when it comes to temperature events on the scale of weeks.
Poll accuracy is declining as response rates. Fortunately for the science of meteorology, air and water can’t refuse to have their temperature or pressure recorded.
The problem is not so much in technical statistics but confusion about what is being measured and compared. Presumably the idea is to compare the pre-industrial temperature distribution with the current post-warming distribution or at least events therein. But which era do the 65 years of measurement belong to? Do the first 64 years of this time belong to the pre-warming era and the current year to the post-warming era? Actually the 65 years probably represent mostly the post-warming era, and we are just in the process of measuring and defining it. Even if we had enough measurements to really define the probability of extreme events in the post-warming era, that probability would not apply to the pre-warming era. From the 65 years of measurement we can infer that the current event is improbable, but we do not have the information to determine whether the event belongs to the pre-warming or post-warming distribution - we don't really know either of these distributions.
If we really wanted to know what the pre-warming distribution is we would have to go back a lot more than 65 years and get many, many years of measurement to define the distribution to the point of specifying how improbable an event is. Any real pre-warming statistics in the antarctic would have to come from paleo temperature measurements. I think we do have those at some time scale from ice cores, but not on the scale of individual weather events.
On what basis do you presume this? The paragraphs Kevin quotes describe only a comparison of the intensity of one event with the intensity of earlier events.
Its the old axiom of climate versus weather
It's hot in Australia today and its been hot all this summer. Thats weather
It's been hot for the last 400 years might be climate
Climate changes there is NO DOUBT about that.
Natural versus man accentuation climate change and how fast it's changing is the only argument here.
Weather is weather we have the instruments to measure that
Climate is climate. We have computer models and ice core studies but not actual eye witness accounts
Even with the heat dome temps are still below freezing so nothing is melting
Check out the climate expert over here.
Can I make a request that "sigma" be avoided as an example of dreadfully obtuse jargon that very few people recognize?