The pace of global warming
April 8, 2013 8:22 AM   Subscribe

A consensus is emerging that in the past decade or so global surface temperatures have plateaued at a recorded-breaking level, not increasing. In fact the world's oceans can absorb up to 90% of all extra heat so global warming has not stalled, it is heating the pool. Predicting ocean heat is tricky, but one scientist's model got the past decade right (in retrospect). Her model shows that by 2020 or so, the ocean may begin to circulate heat back into the atmosphere and things will pick up for us on land. Maybe. Fred Pearce explains.
posted by stbalbach (38 comments total) 16 users marked this as a favorite
Heat isn't the only thing getting sucked up by the ocean.
New York Times Story Bizarrely Downplays Impact of Ocean Acidification
posted by the man of twists and turns at 8:30 AM on April 8, 2013 [3 favorites]

Given today, I think I'll leave here these words from the late Maggie Thatcher:

But the threat to our world comes not only from tyrants and their tanks. It can be more insidious though less visible. The danger of global warming is as yet unseen, but real enough for us to make changes and sacrifices, so that we do not live at the expense of future generations.

Our ability to come together to stop or limit damage to the world's environment will be perhaps the greatest test of how far we can act as a world community. No-one should under-estimate the imagination that will be required, nor the scientific effort, nor the unprecedented co-operation we shall have to show. We shall need statesmanship of a rare order...

posted by ocschwar at 8:44 AM on April 8, 2013 [8 favorites]

Why don't scientists get this right for once and for all so we can keep on polluting.
posted by weapons-grade pandemonium at 9:00 AM on April 8, 2013 [2 favorites]

The ocean's kind of like a great big thermal energy storage battery. Some of Earth's atmospheric systems are like that, too.

It's not that the pace of warming has actually slowed, as the original post correctly frames it; it's like we're now storing a lot of the excess heat energy in a pumped storage system (the ocean).

This is old news to most of us, I know, but it can't be stated often enough that Global Warming does not mean and never has meant steady increase in local ambient temperatures. It means there's more heat energy going into the system, in its totality. Just as refrigerators and air-conditioners use heat to generate cold, what goes into the system (heat) doesn't necessarily come out in the same form. The heat energy that goes into the global climate can come out as kinetic energy in the form of stronger winds, or it can collect in heat sinks like the ocean where it contributes to producing more energetic weather systems, or it can result in any number of other physical phenomena other than warmer weather.
posted by saulgoodman at 9:33 AM on April 8, 2013 [20 favorites]

all the octopi are so giving us the finger right now
posted by angrycat at 10:04 AM on April 8, 2013

That was a nice way to put it saulgoodman. Many folks don't consider the basic math behind it either. You can have anomalous cold, but if the anomalous warm spells are even more extreme, then long term means will increase. Take March for example. The average March temperature last year in Minneapolis was 16 degrees above the long term mean (one of the most extreme departures I could find). That is truly an unprecedented monthly departure. This March in Minneapolis was the coldest in over a decade with a departure of about 6 degrees below the long term mean. The National Climatic Data Center updates the long term mean every 10 years to include data for the previous 30 years. When the next recalculation takes place in 2020, data for March 2012 will have a greater impact on the long term (30 year) mean compared to the cold March of 2013. The long term mean for March will go up as a result of that single warm month (unless we have many more cold springs in the next few years---but I would not bet on it).
posted by Seymour Zamboni at 10:39 AM on April 8, 2013

IANAClimateScientist, but presumably more heat and energy stored in the oceans means, in the short term, more violent storms systems for those of us living on the coast.
posted by salishsea at 10:46 AM on April 8, 2013

Whoops...on post-view, whjat saulgoodman said. I shoulda called saul.
posted by salishsea at 10:47 AM on April 8, 2013 [4 favorites]

Heat isn't the only thing getting sucked up by the ocean.

This is exactly right, and it has serious climate implications. The ocean has acted so far as a buffer for air temperature as well as atmospheric CO2 levels. While ocean acidification is the main consequence of CO2 in the ocean, it can't keep absorbing CO2 forever. We will be emitting way more CO2 in the coming decades and the projections of oceanic CO2 absorption indicate the absorption rate isn't just going to stay steady but is going to decrease, meaning way more atmospheric CO2 and a stronger greenhouse effect. We've been lucky to have an ocean up to this point, but it can't regulate the climate forever.

I think the new IPCC report is due at the end of this year/start of the next. If it follows the trend of the previous reports it is not going to be optimistic, to say the least.
posted by edeezy at 10:59 AM on April 8, 2013 [1 favorite]

So would storage of that much heat impact the entire water column? And then the question becomes what happens if ocean temperatures warm sufficiently that deepwater methane clathrates begin to melt? There's an estimated 500-2500 gigatons of carbon bound up in those deposits (for comparison, there is approximately 700 gigatons of total carbon in the atmosphere).
posted by CosmicRayCharles at 11:06 AM on April 8, 2013

one scientist's model got the past decade right

Talk about burying the lede.
posted by Tell Me No Lies at 11:17 AM on April 8, 2013

The ocean is pretty stratified, so added heat and also absorbed CO2 are generally kept in the top 1000m of the water column. But there are areas where surface water cools enough to downwell, forming deep water on a decade/century timescale. I can't look up any diagrams right now, but there have been measurements made that found the heat and CO2 signature of the last century following the standard circulation pattern, showing the beginnings of significant warming in the North Atlantic Deep Water and Antarctic Bottom Water. I'll try to find some good diagrams when I get home.

I did find a recent article that goes into some more detail.
posted by edeezy at 11:19 AM on April 8, 2013 [1 favorite]

one scientist's model got the past decade right

Talk about burying the lede.

Well, not really. I'm pretty much a reformed climate skeptic these days, and I'm happy to believe that this is good research (though I couldn't get to the original article). However, one thing that still bugs me is people using their models on historic data, getting the right results for the present day, and claiming that this is 'proof' of their model's viability. What they have done is the minimum required to show that their model is not wrong or useless; in order to show that it's convincing, make some solid predictions with it and check back in five years. For the record, I think that this is happening, hence the 'reformed'... though I think a dose of skepticism about anything is healthy, provided you aren't a dick about it.
posted by Ned G at 1:44 PM on April 8, 2013 [1 favorite]

"Retrospective prediction"? Seriously? I think it's just a fancy word for "making a model".

I'm not saying that this hypothesis about heat energy being stored in the ocean is wrong (I have no idea), but anyone who has ever made a model of anything more complex than a few atoms knows that there are usually so many (unknown) variables and assumptions involved that it is all just a matter of fitting those to the available data. In other words, you just tweak the (dozens or hundreds of) knobs so long until the "prediction" fits the data. Of course, an entirely different position of the knobs might fit the data as well.

So there is no way to say how well the model extrapolates into the future. Maybe your current model says the takeup of heat in the oceans will continue for 10 years, or maybe another position of the knobs will say something entirely different but was discarded because it just didn't sound reasonable.

Anyway, there is something about the term "retrospective prediction" that just rubs me the wrong way.
posted by sour cream at 1:45 PM on April 8, 2013 [1 favorite]

Intellectually i read all this stuff and feel... Well, nothing you might call 'positive' or 'good', more something along the continuum between 'despairing' and 'despondent.' Yet the articles themselves are so dry, I think that people are deaf to the message because it isn't coated in an emotional narrative.

And I don't just mean it needs to be shrill - remember 'Bowling for Columbine'? that message didn't sink in very far, did it?

The message needs to be deep, subcutaneous. Needs to bring about change the way smoking went out of favor and blind greed came in.

Its really starting to look like if we don't change we are toast.
posted by From Bklyn at 1:50 PM on April 8, 2013 [1 favorite]

sour cream, this can be offset somewhat by using out of sample testing. It won't prove that the model will continue to work in the future, but it's a bet better than turning a few knobs (aka curve fitting).

Basically, out of sample testing means you create your model with a period of time excluded. After you develop your model, you compare it against the period you didn't look at. That's a bit over simplified, but if you're interested, I'm sure wikipedia has plenty on the subjects.
posted by Crash at 2:04 PM on April 8, 2013 [1 favorite]

Sour cream, a better term for it is "retroactive coherence" meaning that things that were guesses before make sense now. The danger of that clarity is believing that you can get it right going forward, emergent phenomena don't work like that, or as they say in mutual fund circles, "past performance does not guarantee future results."
posted by salishsea at 2:15 PM on April 8, 2013

Reminds me of a recent tweet conversation I had with a friend in Europe:

@them Wishing that global warming actually made the globe warm. #brr

@me It's only cold right now because the ice cubes are melting furiously. Give it a bit of time. It'll be plenty warm.

posted by markkraft at 2:19 PM on April 8, 2013 [2 favorites]

I don't know what is happening with the climate, but here in Ireland and the UK we've had the coldest March on record, with snow drifts in late March.

Even now in a southern part of Ireland, we are having days where the temp is barely getting over 5 degress centigrade.
posted by BlueMarble72 at 4:07 PM on April 8, 2013

Ever push a swing?
posted by the man of twists and turns at 4:13 PM on April 8, 2013 [5 favorites]

I don't know what is happening with the climate, but here in Ireland and the UK we've had the coldest March on record, with snow drifts in late March.

Us in England 2.0 got your Gulf Stream water hanging out offshore.
posted by ocschwar at 6:37 PM on April 8, 2013 [1 favorite]

While at the same time we were having the hottest summer on record right across Australia, BlueMarble72.
posted by panaceanot at 7:15 PM on April 8, 2013

I don't know what is happening with the climate, but here in Ireland and the UK we've had the coldest March on record, with snow drifts in late March.

Here's a model that predicts that Europe will actually get colder as the globe warms - and did so at least a decade before the recent cold winters. IIRC, it was created by Dr. Wally Broecker in the late 1990s.
posted by BillW at 5:26 AM on April 9, 2013

BillW: I recall that same prediction being made on the news in 1989.
posted by ocschwar at 10:51 AM on April 9, 2013

"Retrospective prediction"? Seriously? I think it's just a fancy word for "making a model".
In my understanding (I am not a climate scientist, but I am a scientist and I do currently work on model building in an area related to climate change & energy use), this is more like n-fold cross-validation, where you make models with many subsets of the training set and check their ability to predict the remainder. Cross-validation is one defence (and without the ability to run actual experiments, the only defence) against charges of over-fitting, which is a common criticism of the use of models.

If all you have is a big training set, the only way you can compare hypothesis without waiting it out is something like cross-validation (and perhaps a physically motivated understanding of what the models are doing, in this case). From an empiricist's perspective cross-validation is just as convincing as successful prediction of the future, as it doesn't really matter whether you had the test set written down when you did the training or not (and indeed, if you suffer from Hume's paradox, no amount of successful prediction can ever convince you that the next prediction is any more likely to be correct).

Anyway, for climate change, the probable consequences are so severe that waiting to confirm them would be exceedingly unpleasant for a lot of people, and possibly irreversible, which in my eyes changes it from an interesting debate about epistemology into a matter of considerable urgency.
posted by larkery at 11:09 AM on April 9, 2013 [1 favorite]

Facts. Yes, the planet is warming, at a surprising rate, and yes, we may or not be entirely
to blame but our activity may be the cause. The facts are however that the planet we live on has been both colder and warmer by several degrees, within the last 15,000 years, without our actions. Yes, the world will change, maps may have to be redrawn and we may be forced
to find new sources of food supplies and where agriculture may be successful may be entirely different from today. but, we are not stupid, we are adaptable, we will not only survive but will continue to thrive, but what we must do is learn to control our population.
posted by Lyon1972 at 11:22 AM on April 9, 2013

What they have done is the minimum required to show that their model is not wrong or useless

I was focusing on the "one" in "one scientist's model got the last decade right." Somehow making policy based on the predictions of a bunch of scientists who's models disagree seems premature.

I've never been a skeptic on the existing data, but I have been and barring extraordinary events will always be a strong agnostic on any future predictions. We do not have the ability to simulate something as large and complex as the Earth's atmosphere, period. We simply don't. Computers are not magic, climate scientists have not managed to tame chaos theory, a system with literally millions of inputs does not generate predictable outputs without a hell of a lot more current input data than we're capable of feeding it.

I say "strong" agnostic because I am not only certain that I don't know the outcome of the current situation but because I am certain that no one else does either. And frankly between the purges, the inquisitions, and the millions of people exercising blind belief in something that maybe 500 people actually claim a deep understanding of, I see all the hallmarks of human religious behavior.

Not that it matters much. Like theist religions it's a belief that often drives people to beneficial ends. I like the reduced smog in cities. I like the reduced incursion on the absolutely gorgeous wonders of nature. I love the reduction of reliance on resources that have had a long and very bloody history.

So shine on you crazy diamonds. People doing the right thing for suspect reasons are still doing the right thing. Good enough for me.
posted by Tell Me No Lies at 11:47 AM on April 9, 2013 [1 favorite]

I like the reduced smog in cities. I like the reduced incursion on the absolutely gorgeous wonders of nature. I love the reduction of reliance on resources that have had a long and very bloody history.

The weather must be lovely inside that frame of reference that doesn't take global measurements into account, out here alteration of ecosystems and resource depletion have been accelerating without a hitch and the weather has been insane around the planet for several years running.

It's very difficult to determine empirically how many angels can dance on the head of a pin, but significantly easier to measure things like sea level rise, wildfires, droughts, and hybrid hurricane/blizzards flooding major cities. Climate science is at best somewhat less prone to human tendencies toward ideological blindess, but the fact that I can personally see the global climate changing in front of my damn face leads me to think they have a point.
posted by pleurodirous at 1:34 PM on April 9, 2013

We won't be able to predict the details, butĀ if the ensemble of models predict a range of behaviours which, whilst not identical are somewhat consistent and are in particular consistent on the property of being unpleasant, it seems somewhat more justified than religious faith to conclude that there's a good chance putting lots of extra energy into the system is sending us to a bad place in the phase space rather than an ok one or even a nicer one than we have been occupying so far in the holocene.

The basic science is pretty unquestionable, and can be confirmed with simple equipment. Once you have that, it's a back of an envelope calculation to estimate how much extra solar input we are capturing. The fine detail of what that will do is very hard to predict, but so is the fine detail of what happens when you hit a plate with a hammer. If one model says all the bits go over here and the other says they go over there, doesn't mean we should throw our hands up and say "oh well, it's impossible to know if hitting plates with hammers causes broken plates".
posted by larkery at 1:48 PM on April 9, 2013

all the octopi are so giving us the finger right now

At least the penguins are happy.
posted by homunculus at 3:40 PM on April 9, 2013

The better term for me is "global weirding."
posted by salishsea at 4:33 PM on April 9, 2013

The basic science is pretty unquestionable, and can be confirmed with simple equipment.

The basic science is almost always unquestionable. That's why it's basic science.

Wind tunnels are my usual example for this. There's not much complicated about a wind tunnel. You put an object in, turn on the wind, and see how the air moves.

And the answer is: very differently than your computer simulation. People have been trying to do airflow simulations for years (hell, my Dad was doing them by hand for NASA fifty years ago), and no one has ever arrived at a reliable model.

And that's just wind blowing past a five foot piece of plastic in a completely controlled environment.

It's a long way from basic science to real world application. A long long way.
posted by Tell Me No Lies at 10:47 PM on April 9, 2013 [1 favorite]

Again, we can't predict turbulent flow, at least not when the fluid has a high Reynolds number, because the equations are intractable and the behaviour is chaotic. We can predict the mean direction of the airflow, and all the conflicting fd models will roughly agree on that whilst disagreeing on the location of a given eddy.
posted by larkery at 11:11 PM on April 9, 2013

I think Tell Me No Lies is on the ball.

Think about it, how far can a Meteorologist predict the weather? Mostly a week, and if they are lucky they can see roughly 5 days ahead at best.

Let alone a 100 year simulation.

I'm think the argument on Climate change is over politicised from both left and right, and the the answer of whatever will occur is sacrificed to ideology.
posted by BlueMarble72 at 10:06 PM on April 10, 2013

It's one hell of a challenge trying to get the predictions right and its no suprise their models have a hard time accurately prediciting the effects. When you look at the possible effects of climate change it is quite a list. Sea surface temperature increase and the resultant thermal expansion, increased air temperature, more active storms and increased rainfall, loss of sea ice, glacial loss, etc.

It's a bit like Super Size Me. If you asked your doctor what a diet of McDonalds would do to you he'd say obesity, cholesterol, heart disease, and diabetes. It's hard to say in what order the symptoms would appear but it's certain that your caloric intake is way too high and things won't end well.
posted by karst at 6:51 AM on April 11, 2013

« Older Pounce!   |   Find a separating hyperplane with one weird kernel... Newer »

This thread has been archived and is closed to new comments