Science Spin May 2008

Tales of the unexpected

By Sean Duke

It's a bit like saving for your kids to go to college. Being good to the environment is something that we all agree is good, but, perhaps we don't feel any sense of urgency, or personal responsibility for that matter. What I do as an individual, does it really matter? Climate change is happening, yes, we can agree, but it's happening slowly, and it's probably not going to affect me directly and probably not my children either.

If this thought process sounds familiar then maybe it's time to think again, because researchers at UCD believe there is a possibility of sudden climate shifts happening. If such a shift did happen, it could make many of the world's major cities, towns and small villages uninhabitable, due to extreme heat, or cold, or due to the threat from new dangerous weather phenomenon, or simply because they are buried under water.

Forecasting

The prediction of weather patterns far into the future is the goal of Peter Lynch, the Met Éireann Professor of Meteorology at UCD. Peter was formerly a deputy director at Met Éireann before taking up the newly created position of Chair of Meteorology at UCD - following an agreement between UCD and Met Éireann - in September '04.

Prof Lynch is involved in using mathematical models to try and make predictions about future weather. He works in the long term, trying to predict what climate will be like years, and decades into the future, but, he said, that, despite this, he accepts that it is impossible today for forecasters to precisely predict weather beyond two weeks. Given this fact, how can he be confident of making very long term predictions?

"The idea is that you don't aim to predict the details of the day-by-day weather, we aim to predict the climate, the average situation - not just the average, but also the extremes, the variability," said Prof Lynch.

The idea, he said, is to focus on certain parameters that are known to 'force' change in the global climate, such as the ongoing increase in the carbon dioxide levels in the atmosphere. Then, various levels of, for example, CO₂ change are put into a climate model, to see what the consequences will be. The answer in this case itthat increased CO₂ will lead to an increase in the frequency and intensity of extreme events, such as flooding and storminess. It is important for decision makers toknow that if CO₂ levels go up by a certain amount, then it will increase flooding and storms, he said, as the planners can then make plans well in advance to cope with extreme weather.

The indications from the climate models, said Prof Lynch, are that extreme weather events, such as the flooding that is regularly seen now in certain 'hot spots' around Ireland, will become more frequent into the future. The models also indicate that summer rainfall will decrease by up to 15 per cent and that winter rainfall will increase by a corresponding figure. That means flooding will be more frequent in winter, and that water shortages will become a growing issue with drier summers.

Change

More intense and frequent flooding is very damaging in many ways for a lot of people, but this is nothing compared to what could happen, if, as the models predict, we could 'force' the global climate into rapid, major climate shifts at some point in the future. If this happens, then all bets are off, and predictions become very difficult indeed. In this scenario, anything could happen: Ireland could find itself, frozen over, turned into a desert, or battered by huge storms, or even drowned by rising sea levels.

Trying to predict what might happen, should the climate rapidly shift is the area of research that Peter Lynch is most fascinated and intrigued by. He explained that people tend to think of climate change in linear terms, so that if CO₂ is increased a little bit, then temperature increases by a corresponding little bit. The problem is that climate systems are not linear, but in fact are highly non-linear and highly chaotic.

Evidence

Scientists can use ice cores taken through the Greenland Ice Sheet to re-construct the ancient temperature at various times in the Earth's history. It is possible to track back about half a million years, and these studies have revealed very sudden changes. It has been shown that there have been changes in temperature of up to five degrees Celsius in the space of a decade. This shows that rapid climate change has happened before.

If a change in temperature of that magnitude, and within that timeframe were to happen today, it would be catastrophic for humanity, said Prof Lynch. It would lead to major movements of people from hugh densely populated areas that have simply become uninhabitable. There are no grounds to think this will happen anytime soon, he said, but he also said he can't rule it out.

There are indications that rapid change could be happening right now, from studies of the global ice. The area of ice has been decreasing, and in the past few years there have been huge break ups of ice around the world. For instance, in the southern hemisphere, in Antarctica, a major portion of the huge Larsen Ice Shelf has recently split off into the sea. In one melting season alone - in early 2002 - 3,250 km² of ice, 220 metre thick disintegrated and fell off the Larsen Ice Shelf into the sea. This is an area equivalent to the US state of Rhode Island. This was melting on a level that hadn't been seen before, and is a sign that the climate might indeed be changing rapidly.

"Sceptics will say that this has always been happening, and they are right in the sense that there is a natural shedding of ice from the edge of the ice shelf, but this was quite big and something unprecedented," said Peter. "It freshens the water locally and increases sea levels, it depends where the ice comes from - the Arctic ice is already floating on the water so there is not much change in volume, because the water that is produced just displaces the ice, but if it is coming from Greenland or Antarctic down through the glaciers then, of course, it is adding volume to the water - the sea level will rise."

Unexpected

The really scary, but interesting - from an academic point of view - part about climate change is the whole area of 'unanticipated emergent phenomena'. These are weather events, or features that emerge, unexpectedly, having never been seen before. Trying to predict the unpredictable is difficult. It requires the researcher to think beyond what we now understand in terms of weather, and try and think, what could happen?

Who could predict a rainbow, if we did not know that rainbows exist? A rainbow can be explained in terms of the reflection and refraction of light in water droplets, but it would be very difficult to produce a rainbow effect in a computer model if scientists didn't already know that they exist. The point is that unexpected weather phenomenon could appear in the future - things that we have never seen before - and these could be totally unexpected and beyond the power of current computer models to predict. Prof Lynch wants to develop a much better understanding of what might emerge, and a better means of predicting where, when and how these phenomena might arise.

Another example would be to think of hurricanes. This is weather feature, which we are all aware of, but they will not happen unless the sea surface temperature is above 26º Celsius. If the waters everywhere were below 26º then hurricanes would not exist on Earth. Nobody would know what a hurricane was. The point again is that if climate changes, new weather patterns could emerge that are totally unknown today. These patterns could be beneficial, but are more likely to be devastating, said Prof Lynch.

Ensemble

One way to improve the measure of confidence when predicting future weather, is to use the method of 'ensemble prediction', said Prof Lynch. The way this works is that the researcher runs a large number of weather simulations, looking at a large number of possible future climate scenarios. The larger the number of scenarios that are run, the greater the level of confidence the researcher will have in his future predications.

The goal, said Prof Lynch, is to more precisely determine what is the probability, or the likelihood, of certain things happening with regard to the weather into the future. It could be the case, for instance, that in an ensemble, most of the predictions cluster in a certain region. That means that whatever this cluster represents, has a good chance of happening. The ensemble provides a measure of confidence. There may also be "one or two wildcats" within an ensemble. These represent the possibility of unanticipated events - and this is the area where Prof Lynch is focussing his research.

Action

The problem of climate change is a major global problem, said Prof Lynch. He believes that it requires the same sense of urgency, and effort that was put by the USA into getting a man on the Moon in 1969, or developing the atomic bomb during World War II. In this case, though, the research effort should be on an international scale.

He said that climate change is not treated as an immediate, urgent threat in the same way, for instance, that an asteroid heading for a collision with Earth would be treated. The changing climate must, however, be regarded as an urgent major threat to us all, and as an enemy that all of mankind must get together and beat, said Prof Lynch.

He believes that climate change is underway - though up to recently he would have been sceptical of this - having been convinced by the ongoing increase in mean temperatures, breaking records year after year, by the increase in extreme weather events, by the melting of Arctic Ice, and the movements of animals, such as birds. The problem requires an international agreement, he said, and the UN is the best framework for this, and "sticks and carrots" must be built into this agreement.

One problem is that if climate change mainly affects people in the poorer countries, and is perceived as a problem of deprivation affecting others, then it won't engender the same push for action in the developed nations that generate most of the pollutants.

"I wouldn't wish it to happen, but if a major hurricane hit New York it might ultimately be beneficial for humanity," said Peter. "Let's hope it doesn't happen, let's hope that we get there by a less devastating route."