Left 1960s: low NAO index, northerly winds, cold European winters
Right early 1990s: high NAO index, westerly winds, warm stormy European winters
The North Atlantic's El Nino New Scientist 31 Jan 98
Oliver Morton is a science writer based in London
IN THE vast open wastes of the South Pacific, El Nino is now reaching its climax. This is the biggest, most dramatic recurring climatic event there is-its effects are felt around the world. It is reasonably easy to grasp how El Nino works, at least in broad outline: a big glob of warm water, normally kept in the west by prevailing winds, breaks free as the winds collapse, and heads towards South America. The El Nino currently under way is the biggest on record, and is already causing havoc: crop failure in southern Africa, storms in Santa Monica failed monsoons across Asia. Climate scientists and meteorologists who specialize in the weather of the Atlantic Ocean might well feel a twinge of envy when they see such an impressive suite of effects. What do they have that can compare? Over the years it has become increasingly clear that the Atlantic has a dramatic event all of its own. And in some effects could be just as important.
World of weather
El Nino's cousin is called the North Atlantic Oscillation. Changes in the NAO correlate with all variables around the ocean, from rainfall in Bordeaux to the amount of Saharan dust that ends up in the Bahamas and the richness of the fisheries off Iceland The NAO affects the circulation of seas at the North Atlantic's margins. It leaves smudgy cyclonic fingerprints all over the northern hemisphere's climate. And it affects temperature of the whole world. An analysis carried out by Jim Hurrell at Center for Atmospheric Research in Boulder, Colorado which was published in 1996, shows that mild winters across Europe and Asia linked to the NAO account for a good chunk of the warming trend in global temperatures seen in the past few decades-with cumulative global effects on the same scale as those of El Nino. This news has been enough to make NAO researchers flavour of the month with the energy lobby in which likes nothing better than natural explanation for global warming. But it may well find that the NAO is not quite the boon it had hoped for. One of the greatest unanswered questions in climatology is the range of the climate's natural variations. How does it vary when left to itself? How long are the variations? How extreme? These natural fluctuations for some of what is seen as greenhouse warming but at the same time they might mask it. They could even be under the control of the greenhouse effect, channelling its impact. In a greenhouse world they could become still more important.
The NAO is one of the most intriguing of those fluctuations. Its long-term swings have become more pronounced over the past century. It had a significant warming effect on European winters in in the early 1990s, and it may now be on course to cool things down for a decade or so. To find out what the NAO will do next means, trying to understand all sorts of subtlties in the ways the North Atlantic's winds drive and are driven by its currents. Which is why more and more climate researchers are getting into boats and putting to sea whatever the weather.
When the wind blows
In terms of measurement, the NAO is a fairly simple thing-it is an index created by comparing the pressures in the Azores and in Iceland. In the first half of the 20th century meteorologists created a number of similar oscillating indexes. The Southern Oscillation, the difference between the pressure in Darwin and in Tahiti, measures the atmospheric component of the El Nino effect, which is why climate moddellers tend to call the whole ocean and atmosphere ensemble ENSO (El Nino/Southern Oscillation)
Recently the NAO has been switching from high to low on a timescale of decades
For the NAO a high index signifies low pressure around Iceland and the reverse off Portugal a situation that gives rise to strong westerly winds In winter - which is when these signals are strongest - these westerlies bring heat from the ocean's surface into the European continent, along with storms like those that this winter saw in the New Year in Britain. So a high index tends to mean relative warmth in northem Europe, while a low index means weak westerly winds and a continental climate dominated by cold air from the north and east. The difference is marked. David Parker of the Hadley Centre for Climate Prediction and Research in Bracknell, part of Britain's Meteorological Office, says that oscillations in winter between high values and low values of the index account for about 50 per cent of the variability in monthly average temperatures in central England. And it's not just the wind and the warmth, it's the rest of the weather too. Mild winters with high NAOs dry out the south of Europe, with a spate of them through the early 1990s explaining everything from poor olive harvests in Valencia to poor skiing in Val-d'Isre. If you could predict how the NAO is going to rise and fall on a scale of weeks and months, then you would be doing all Europe a favour. Unfortunately, you can't. According to Tim Palmer of the European Centre for Medium Range Weather Forecasts in Reading, no weather forecast could predict the exact evolution of the NAO over more than a few weeks. The problem is that the atmosphere over the North Atlantic is chaotic. Forecasts based on almost idenfical initial condifions give similar results in the short run, but in the long run they diverge hopelessly. Computer models make the point neatly-and show up the difference between the NAO and its better-known cousin half a world away. If you run a computer model with the temperatures of the sea surface fixed, says Palmer, you will see something rather like the NAO flip-flopping around happily enough. The Southern Oscillation, though, will shut down completely if sea surface tempera ture remains constant. For ENSO, the ocean calls the shots, and nothing happens without its active participation. So at first sight the NAO appears to be an atmospheric phenomenon, inheriting all the unpredictability that air is heir to. But the story isn't that simple. The NAO doesn't just jiggle about on timescales of weeks and months, it also has slower rhythms whose beats are measured in 8 years, decades or more. These are the trends that interest climatologists, and raise the hope of predicting climate-as opposed to weather. The chart shows over a century of these long-term ups and downs: a dip into the negatives in the 1940s, which saw some of the coldest Eu ropean winters of the century, including the ones that delayed Hitler's invasion of France and defeated his assault on Moscow; a protracted dip in the 1960s, the decade with the consistently coldest winters in Britain since the 1880s; and the odd, long period of very high average NAO states in the late 1980s and early 1990s which corresponded to particularly mild winter weather across Europe.
These long-term changes suggest that something more than atmospheric chaos is going on. Atmospheric effects tend not to persist all that long unless there is something pushing them. Though it is possible to get decade-long ups and downs in the NAO in computer models that mimic the atmosphere alone, many scientists in the field are pretty sure that that's not the whole story. They think the long-term memories that shape the NAO are held beneath the waves, in the cat's cradle of currents inside the North Atlantic. The NAO may not be the highly marshalled double act of atmosphere and ocean seen in El Nifios, but its long-term shifts may wefl be the product of the same players following more complex rules. Hurrell and Mike McCartney, who studies the NAO from the Woods Hole Oceanographic Institution in Massachusetts, have produced data suggesting that the ocean seems to be responsible for the NAO's ability to remember what it has been doing from one winter to the next. They took a set of years with similar winter NAOs and looked at what the NAO was like in the preceding autumn and the following spring, summer and autumn, finding that there is no fixed pattem at all. In the warmer parts of the year the NAO appears to wander around aimlessly. But the clue to the role of the ocean comes when you compare one winter with the next: when winter comes, the NAO tends to go back to the state it was in the previous year no matter what it has being doing in the meantime. McCartney and various other oceanographers are convinced that the ocean reminds the atmosphere what to do from year to year. And since there are long-term patterns in the NAO, it makes sense to look to the ocean to see what might be responsible. There are some intriguing candidates. Donald Hansen of the University of Miami and Hugo Bezdek, who runs the Atlantic Oceanographic and Meteorological Laboratory at the US National Oceanic and Atmospheric Administration, showed in 1996 that big patches of ocean with anomalous surface temperatures-a fraction of a degree hotter or colder than expected for the time of year-slowly drift up across and around the North Atlantic. They follow a path from west to east, rather like that of the Gulf Stream and the North Atlantic Current, though they move much more slowly than the currents themselves. Then, at the same stately pace, they turn in a great circle around the sub-polar gyre of the North Atlantic, ending up somewhere west of the tip of GreenlandMcCartney and his colleague Ruth Curry soon showed that these anomalies are not restricted to the surface-they have deep roots (which get deeper as they make their slow and cooling progress towards the pole). They contain a lot of heat, and represent significant changes in the overall flow of heat from the subtropical Atlantic to the north. McCartney thinks that they may be responsible for the NAO's decade-long swings-they are in the right place carrying the right sort of energy and taking the right sort of time to pass by. As yet, however, there is no clear mechanism to explain quite how such circulation might force the NAO one way or another. Without understanding this in detail, longterm guesses about what the NAO will do next remain impossible. Even though the effect that the ocean has on the air above it is still obscure, the effect that the atmosphere has on the waters below is becoming clearer. Bits of the story have been worked out in studies all around the ocean, including long-term observations of the Labrador Sea coordinated by the US, and pulled together in Britain by Robert Dickson of the Centre for Environment, Fisheries and Aquaculture Science in Lowestoft Over the past few vears, Dickson has studied the changes in the ocean that went along with the shift from very low NAO indexes in the 1960s to consistently high indexes in the 1990s. He has found that these changes in average atmospheric pattern fitted into changes in the Sargasso, Labrador and Greenland Seas. Each of these areas is capable of creating large volumes of homogeneous water, a process which can have effects on the ocean elsewhere, and their propensity to do so seems to be linked to the NAO. In the 1960s, when the NAO was low, the Peak practice: recently the North Atlantic Oscillation has been switching from high to low values on a timescale of decades.
Sargasso Sea was producing the water it is famous for, which oceanographers call 18oC water", and there was deep convection going on in the Greenland Sea, producing cool salty water at depth. But nothing much was happening in the Labrador Sea. When the NAO was at its reverse extreme in the early 1990s, there was deep connection in the Labrador Sea, but little interesting activity anywhere else (see Diagram, pren,ious page). This deep convection in the Labrador Sea makes sense. Water sinks when it is cold and dense. Strong winds passing over the surface will always cool water, and when the NAO is high the westerlies rip across the face of the Labrador Sea. The same pattern of pressures doesn't provide such winds for the Greenland Sea, up above Iceland, and so when the NAO is high the Greenland Sea gets warm; it also gets less salty, which makes it less dense, and that also lessens convection. When the NAO changes to its negative state, though, the tables are turned-high pressure over Greenland brings strong, cold, dry winds from the North Pole down over the Greenland Sea, cooling it enough for convection. The same winds drive relatively fresh water and sea ice down the east coast of Greenland and round its tip into the Labrador Sea, where the freshness and lack of cooling winds make deep convection less and less likely. During the early 1990s, the high NAO drove convection in the Labrador Sea at a cracking pace, producing large amounts of cool water at depth. But in the winter of 1995/96, every thing changed. Straight after the highest winter index of the cen tury came the lowest. The winter after that was ambiguous, and the current one is only halfway 90 through, though it has certainly had its fair share of westerlies recently. What is going on? Is the NAu unciergoing a long-term shift towards the negatiN,e-suggesting a cooling of northern European winters for years to come-or is it simply in the middle of a blip? A long-term shift in the climate towards low NAOs and cold European winters should, if Dickson is right, shift the production of cool, deep water from the Labrador Sea to the Greenland Sea. A European Union programme called ESOP-11 has been coordinating a wide range of studies of the Greenland 5ea over the past three years. Last year, as part of ESOP-11, a team led bv Andrew Watson of the University of East Anglia released an inert marker chemical, SF, below the circling waters of the Greenland Sea. The way this has spread out vertically from its original depth suggests that moderate convection has indeed started again, though it has not reached the depths. And observations made early this winter by McCartney show that there may be twice as much water coming through the Denmark Strait between Iceland and Greenland at depth than there was a couple of years ago, which might also indicate increased convection. There may be a big shift going on, but it's too early to say for sure-the models just aren't good enough. Such shifts in circulation may not just be a result of the NAO's moods, they may be contributing to them. McCartney says that a large mass of cold water that sank in the Labrador Sea at the end of the 1980s and the beginning of the 1990s is now heading south along the eastern edge of America's continental shelf. Its effects have already been detected south of the Gulf Stream at Bermuda, and east of Miami. According to numerical models created by McCartney's colleague Michael Spall, this cold water at depth could make its effects felt in the Gulf Stream passing over it. This could in turn create colder surface temperatures that would then generate anomalies like those Hansen and Bezdek found wandering across the Atlantic. If so, changes in the NAO's output would be leading directly to changes in the factors that force its development. The snake may be biting its tale, producing a closed cycle where the factors which encourage one NAO state lead, through oceanic and atmospheric routes, to changes that encourage the opposite state. But if there is such a closed cycle, it is unlikely to be closed tight. The different processes-the creation of sea surface temperature anomalies, the flip-flopping of convection, the rise and fall of the cold Labrador water-may all have different typical timescales. Sometimes they could beat in sync and sometimes they could cancel each other out. That would fit with the fact that the NAO's variability varies. In the 19th century, most of the variation was on timescales of between two and three years. Today the variability is measured in decades.
The NAO's oscillations are a hot topic in global warming. McCartney remembers that when The New York Times ran a piece last year reporting that the NAO effects and those from other natural cycles might account for most of the observed warming in the northern hemisphere, he got two calls. One was from a noted critic of ideas about greenhouse warming as a result of human activity. He was "very happy", McCartney recalls. The other, he says, was from a group representing automobile manufacturers, who "wanted to sign me up as a consultant." McCartney admits that natural shifts in the NAO may be responsible for a fair-sized chunk of the warming previously ascribed to greenhouse gases. But he prefers a different view: that the greenhouse effect may be changing the way the NAO and other natural climate variations actually vary. The idea is that if the greenhouse effect is going to manifest itself, it will have to do so via existing climate patterns such as the NAO. Changes in the stratosphere's winds and structures could reflect changes in the NAO's behaviour, and the high indexes in the early 1990s may be an example of just that. The eruption of Mount Pinatubo in the Philippines threw a girdle of dust around the world which cooled down the surface but heated up the tropical stratosphere. As a result, stratospheric winds, driven by the temperature difference between equator and poles, got stronger, especially those around the poles. Chris Folland of the Hadley Centre points out that strong circumpolar winds in the stratosphere mean strong westerlies in the troposphere below, just like those seen in the early 1990s. McCartney observes that global warming and ozone depletion also cause changes in the stratosphere-both are thought to cool it. No clear mechanism for this has been worked out yet-as Folland stresses, no models capture the details of the interaction between troposphere and stratosphere very well. But until we know if global warming is strengthening or changing the NAO in some way, it's worth being cautious. As McCartney says, "No matter what you do to the atmosphere, it will get mapped on the same systems." It's just that, as yet, no one knows exactly how the mapping works. To find out will require rather better modelling of the subtleties of the NAO, the excesses of El Niho, and the intricacies of other naturally varying climate systems-and a lot more messing about in boats.