What Just Happened

Somebody needs to write a straight news article about Summer-In-March, climate science, and climate change. No snark, no trickery, no exagerration, no personality. Just the facts, most important first. Herewith, my attempt.



The widespread warm temperatures in the eastern half of North America were accompanied by an astonishingly strong anomaly centered in the upper midwest. Record daily temperatures were attained in 25 states. All time monthly records fell in many places. This can be directly attributed to unusual behavior of the jet stream.

There are scant precedents for this situation in the United States observational record. The main directly comparable event occurred in Russia in Summer of 2010. (Somewhat similar events occurred in Texas and northeastern Mexico in 2011 and, in Australia in the southern summer of 2008/2009.) The techie/journalist Brad Johnson has called this “sticky weather”.

Some meteorologists are now identifying this “stickiness” as a new feature of the atmosphere. There is no prediction of anything like this emerging to be found in IPCC reports. If these extraordinary events keep occurring or even get more extreme, they cannot be taken as a success for climate science, which did not, in fact, predict them.


The Globe Continues to Warm Gradually

It is important for people caught in the “bulls-eye” on the map, where the most unusual warmth occurs, not to conclude that the entire world has warmed comparably.

In fact, large warm anomalies in one place are always accompanied by large cold anomalies in other places. Global temperature is much more sluggish to change than regional temperatures, because it is constrained by the energy balance between the sun, the Earth, and space. This balance changes relatively slowly as greenhouse gases accumulate.

The temperature of the whole world swings back and forth year-over-year because of internal variation, notably the el Nino oscillation which exposes or hides warm water on the equator. A gradual warming is superimposed on these variations, but it takes decades for that signal to emerge unmistakably from the noise.

So in fact at the peak of the warm anomaly, very cold temperatures in the Pacific states were just upstream of the very warm temperatures in the midwest. On the average, the world was not much warmer than it has been of late. So how should we put this event into context?


An Anomalous Anomaly: The Extraordinary Nature of the Warmth

Monthly anomalies in excess of 15 F were noted. Much larger two week anomalies were widespread.

These warm anomalies were unusual not only in their strength but also in their causation.

Unusually warm days in March at the latitude of the anomaly normally arise when a location is briefly in the warm sector of a large scale storm system.


The above shows a typical early spring heat wave, from April 1, 2008. It shows the vigorous dynamics that normally accompany a warmish day in early spring in the great lakes region, with temperatures briefly hitting the 60s (F). Commonly such days are followed nearly immediately by snow and sleet, and in the case shown, the abrupt cooling is passing through Chicago.

(I show these because I found this pair of images, not because this was an extraordinary event. Generally, heat records at this time of year in this year are set by similar events.)

Such events are usually followed in short order by a vigorous cold front trailing behind. Hence, people in northern climes faced with a pleasantly warm day during the transitional seasons often are heard to say “We’ll pay for this later.”

But this year, the bill never came due.

Extraordinary warmth, with anomalies in excess of 18F (10 C) persisted over a large area for weeks. Wisconsin was the bullseye, with an astonishing two-week anomaly of 25 F (14 C).

Clearly, this was a different sort of a thing. But what sort of a thing was it?


Understanding the Jet Stream

The world’s weather can be understood as a battleground between cold air and warm air. The tropics get much more energy from the sun than the poles do, and the air, being a fluid, resists this difference and tries to mix it out. (This is fortunate, since without such mixing the habitable zone of the earth would be very narrow.) However, the rotation of the earth actually adds a twist. There are a few ways of looking at it, but in the end they result in a situation where the mixing is inefficient because winds are steered off to the side by the Coriolis force. This causes the characteristic large swirls we are familiar with on satellite infrared views.

This struggle to mix warm air and cold air ends up creating distinct air masses, and aloft at the tropopause a strong stream of eastward flowing air marks the boundary. (It has to. The mathematics of this are interesting, but out of scope for this article.) Because the world round, the stream, which we call “the jet stream” in a meteorological context, or rather “a jet stream” (after all there are two of them on earth, not counting those on other planets) is approximately closed, forming a wobbly ring around the pole.

The big weather systems that affect us middle-latitude dwellers are usually associated with those meanders in the jet. Then there are the persistent anomalies. Traditionally called “blocking events”, we can choose to describe them with the more intuitive name “jet stream stickiness”. They were called “blocking” by the first people who saw them on weather maps, who would find themselves saying things like “something seems to be blocking this meander from moving off to the east”.

When the jet stream gets “stuck” unusually warm and dry weather is persistent where it bulges northward, and unusually cool and wet weather where it bulges southward. Which brings us, finally, to last month’s picture:

Notice that this is a picture of winds and temperatures aloft (probably at 300mb, i.e., at 0.3 atmospheres of pressure, or a bit above the top of Mount Everest), not of surface conditions. But it is closely correlated with what is happening at the surface. And what it shows is that the middle of the country (this was toward the end of the month so the bubble had shifted eastward a bit) is firmly in a tropical pattern. The jet stream does not ordinarily retreat that far north in March.

Having achieved an extraordinary configuration, the atmosphere would ordinarily revert to a more normal configuration in short order. However, as with summer 2010 over Russia, it did not. Day after day showed extraordinarily warm temperatures, due to a persistent and extreme jet stream anomaly.

Let me quote the summary by Steve Skolnick of Capital CLimate:

U.S. daily record high temperatures continued to surge into the end of March, with new heat records overwhelming cold records by the incredible ratio of 35.3 to 1. The total number of heat records was 6,182, nearly double the number in the sweltering month of August last year. The monthly count for March was also more than the entire year’s total of cold records in either 2011 or 2010. The year-to-date ratio is now at 22 to 1, nearly double last summer’s pace.

Included in the total were 340 new all-time March high temperature records vs. only 8 low temperature records, a ratio of over 42 to 1. Over 110 official climate reporting locations (ASOS stations) in 27 states also reported record high average temperatures for March. This represents roughly one-fourth of the total number of such stations.

What About 1910?

There’s some talk that this event was unprecedented. It turns out that this is a matter of degree, or literally, a matter of degrees. If you look at this map, also via Steve Skolnick, you see that March of 1910 was unusual in a similar way:

This map has to be taken as somewhat, but not entirely reassuring, in comparison with this year’s.

So the pattern does have a precedent; it’s only a few degrees worse this year. Can we attribute this to random variation? Not entirely, because weather is not actually a statistically normal distribution. Unusual warmth has to come from somewhere. A couple of degrees beyond the previous record is suggestive of a warmer and more generous supply of tropical air, combined with a similar variation.

When we look globally, though, it is hard to avoid coming to the conclusion that similar events have been especially common of late. The French heat wave, the Chicago heat wave, the Australia heat wave, the Russian heat wave, the Texas drought… these are all comparable patterns that have done considerable damage in the last few years.

It is hard to know whether selection bias is at work. Are these events just more salient than they used to be? That is, do we just tend to notice them more now that there is a trend for which we are looking? Various leading meteorologists have changed sides on this question. Perhaps the most notable of them is Stuart Ostro, once a skeptic, who has compiled a massive compendium of recent odd meteorological phenomena that he attributes to slow-moving large-scale dynamics.


This is Exactly What Scientists Have Been Predicting All Along, Isn’t It?


People whose support of science is deeper than their understanding of it are taking a sort of smug I-told-you-so attitude to all of this. “Scientists said there would be more and more intense heat waves, see?” In fact, this is wrong, because this is not the sort of heat wave that has been predicted. It is too intense, too regional, too idiosyncratic. It is all about changes in behavior of the jet stream. If you look in the latest IPCC report, you see very little. Searching the literature pretty much turns up a single paper calling for a slight decrease in blocking over Europe.

The fact is that nobody saw this pattern coming.

In an abstract sense, though, there has long been an expectation that climate change would bring surprises, and perhaps abrupt ones. So call this an unknown unknown.

While it is tempting to enjoy an early spring in the midwest, the phenological impacts on wildlife alone are particularly concerning. Especially fruit crops are far advanced for the season, and climatologically, hard frosts can still be expected. The average last freeze date in Madison, as I recall, is May 10.

This image purports to show cherry blossoms in Galena, Illinois, around March 24.

That said, all should be glad that this did not occur in summer, and hope that it doesn’t recur in summer.

But it simply has to be said that the scientific community as a whole didn’t expect this pattern to emerge. If some individual predicted it, we would be well-advised to find that, but so far as I can tell at present there isn’t even that. Post hoc explanations are to be expected, and have to be suitably discounted.

The fact is, while we understand jet stream dynamics quite well, our models do not capture the blocking events, and the efforts to develop a theory of blocking have not been entirely successful.


So Is It Global Warming?

As Kevin Trenberth explains, all weather now occurs in a changed context, so unusual weather does as well. When we see unusual patterns of weather, we are hardly being unreasonable in associating them with the changed context.

This is a long way from statistical attribution. But thinking of extreme events statistically is already leaning on the wrong abstraction. The rarer a thing is, the less useful statistics will be in getting a handle on it. We’d better hope that a formal pure-statistical, physical-inference-free attribution remains far in the future.

Nevertheless, it is possible to form beliefs. Formally, this is modeled as a Bayesian process. We see the world being disrupted, and we see an increase in odd events. We must choose between coincidence and causality. Nobody can be sure.

The trouble is that while this is an unimportant question from a scientific point of view, it is an important one from the point of view of policy. The costs of mitigation have to be balanced against the costs of adaptation. If we think the sorts of weather we are seeing are likely to be disruptive, we need to weigh that disruption against the disruption of a policy response.

Unfortunately, the name “global warming” confuses the matter horribly. Is it reasonable to suggest that the indistinguishable difference in temperature between this year and last “caused” an enormous and persistent anomaly in the jet stream? What could that even mean?

The fact is that we have altered the environment in which weather develops, on a global scale, in various ways, some directly through greenhouse gases, but also through pollutants and land surface changes. If we see something odd in the present context, it cannot be independent of that context.


Uncertainty is Not Your Friend

When I say “uncertainty is not your friend”, I mean just this sort of thing. We did not see this increase in blocking coming. It cannot be taken as a success of climate science. It is the system giving us a surprise. But by the nature of surprise, these surprises will be events for which we are unprepared.

So far, the warm outbreak seems to have been relatively benign, certainly in comparison with the events in Russia in summer 2010 and Texas and Mexico in summer 2011. But these events are all in the same class, and if they are becoming more common, the anticipated cost of climate change needs to be adjusted upward.

We cannot expect pleasant surprises. By its nature, the climate we want is the climate we are already adjusted to. This is why failures of climate science should not be celebrated by people who don’t want to change our infrastructure. These failures of climate science will tend to correspond to missed mechanisms of climate disruption.

We may not know what the new normal means, but the old normal is already vanishing. The key question at hand is how much worse we are willing to let it get. It’s important to understand that any gap in our understanding is unlikely to make matters better.

UPDATE: See (or hear) also Kevin Trenberth on the subject.


  1. Thanks, an excellent summary that I found very helpful.

    Others have probably also seen a recent study in GRL especially after it was posted by Climate Central and Joe Romm, which seems to be relevant to this discussion. It argues that Arctic amplification is slowing down the movement of the jet stream, leading to longer weather patterns and so greater extremes as wet/dry, hot/cold stick around for longer. (For some reason, the study itself seems to be down at the moment.)

  2. Yes, it appears none of this sea ice causes blocking appeared until after the IPCC was complete. Jennifer Francis (the person whose paper has been discussed frequesntly) submitted a paper in early 2009. This paper discussed the atmosphere "remembering" the sea ice loss across seasons.

    Later in 2009, Vladimir Petoukhov submitted a paper that was covered on RC, which was treated as bleeding edge science.

    I assume this lack of awareness was due to a lack of data on large sea ice loss seasons and lack of extreme events caused by blocking. The IPCC link you gave seems to show that the models were not accurate to carry out any useful predictions.

  3. Well, clearly the question was being at least asked prior to the Russian event, although note the expectation of "more numerous, weaker and long-lived" events.

    Lupo (as a NIPCC co-author someone whose results should perhaps be examined more carefully than usual) has a blocking event page that may contain some useful information.

  4. As a system undergoing change snd exhibiting spatio-temperal Chaos, it would be nearly impossible to predict this kind of detail (i.e. more blocking weather events) as more heat is added to the overall climate system. We have heard for years that as we push the climate system toward some new equilibrium point, we can expect "nasty surprises". That isn't very scientific, but might be about the best we can hope for. Trenberth is right is saying it is all Anthropocene weather now, and it seems likely more such unpredictable nasty surprises are in store. The Holocene was good to us. It remains to be seen if the unpredictable Anthropocene will be.

  5. OK, this 2008 conference paper is more informative about blocking events generally. Skimming the first bit, it says a 1997 paper "suggested that a warmer climate would lead to an increase in the number and duration of blocking events." So the idea has certainly been in development for a while.

    TAR Chapter 9 mentioned the 1997 paper, barely, but said the results should be treated with caution.

    In any case none of Lupo's stuff seems to mention Arctic sea ice.

  6. "Atmospheric blocking occurs between 20-40 times each year and usually lasts between 8-11 days, Lupo said. Although they are one of the rarest weather events"

    So it's so rare that it happens around 300 days per year?

    Aargh, press releases. Thanks for the clue, though.

  7. Pingback: The Climate Change Debate Thread - Page 1198

  8. "This image purports to show cherry blossoms in Galena, Illinois, around March 24."

    Michael, the two cherry trees just around the block from me were in blossom by March 26, and I'm on the north shore of Lake Ontario.

  9. "This image purports to show cherry blossoms in Galena, Illinois, around March 24."

    Those are Eastern redbud (Cercis canadensis) blossoms. It's a well-known early spring bloomer, but March 24 is pretty early for Galena.

Leave a Reply

This site uses Akismet to reduce spam. Learn how your comment data is processed.