What is the Greenhouse Effect? Carbon Dioxide, Water in the Air, and Climate
[Back to top]
Well, there is widespread concern and worry that man-made emissions of infra-red absorbing gases, like carbon dioxide, methane and CFCs from aerosol cans, may lead to an enhancement of the natural greenhouse effect, and may lead to global warming of the atmosphere, and also lead to changes in other aspects of the climate, like the rainfall, the snow and ice cover, sea level, soil moisture, and so on. Unfortunately, the subject has become very controversial, and I think it's become overhyped and certainly distorted by the media. I think there's quite a lot of confusion about the scientific basis, not only among the general public, but even among professional scientists themselves. And as you know, just last week there finished a big ministerial conference in Berlin where ministers came from 120 countries for a week, to decide what they were going to do, if anything, about carbon dioxide emissions beyond the year 2000. The meeting was somewhat of a fiasco I believe; very little was decided. I'll give my view later, when I've discussed the scientific aspects.
Now of course there's no doubt that the CO2 in the atmosphere is increasing. It's increased steadily since the industrial revolution and it's increased by about 27% since 1860. And it's increasing at the moment at about 1/2% per annum. Obviously, if that continuesand it has been until the very last few years acceleratingand certainly if the world population continues to increase at the present ratethen there will come a time when the carbon dioxide content of the atmosphere will, for example, double, somewhere towards the middle or latter part of the next century. So it's not really a question of whether the greenhouse effect is being increasedit isbut the main question is what will be the magnitude and the timing of the effect as far as the climate is concerned. Are the climate changes likely to be so large and so imminent that governments now have to take some major action, or are they likely to be sufficiently small or sufficiently delayed that we can either live with them or adapt to them?
And I think the best thing I can do this evening is to summarise our latest views about what we understand of the climate, what makes the climate work, how we predict the man-made impacts, and perhaps just as important, the uncertainties involved in the predictions of what the climate may be in 50 or 100 years time
Let me start by saying that the natural greenhouse effect is a terribly important part of the earth's climate. When I talk about the earth's climate system, I mean the earth's atmosphere, the oceans, the land surface, the biospheretrees, vegetationand the land and sea iceglaciers, landbased ice, and the sea ice. All of those together make the climate system, and they all interact in a very complex way. But we have to treat that whole system if we are to understand the climate.
Now this big global system is driven by energy from the sun. The incoming solar radiation to the top of the atmosphere, averaged over day and night, and over all latitudes and over all seasons, is 340 watts on every square metre. So if you think of a square metre, and three 100 watt and one 40 watt bulb shining on that all the time, that's the total amount of energy that's coming in from the sun. That's equivalent to 340 megawatts on every square kilometre. You can think of a square kilometre and a medium-sized power station on it. So I mean the amounts of energy are tremendous. As that incoming radiation from the sun comes through the atmosphere, some is reflected back by clouds and scattered back by gases to outer space; some is absorbed; only about half gets down to the earth's surface. It warms the earth's surface, and then the earth itself radiates heat back to outer space. Of course the earth's temperature is much lower than the sun's; the sun is about 6000 degrees Centigrade, the earth about 15 degrees Centigrade. So the radiation which the earth radiates back is not the visible light, that you see from the sun, but invisible infra-red radiation, that you get from a fire. That radiation, if there were no greenhouse gases in the atmosphereno CO2, no water vapour, no ozonethat would escape to space. And then the earth's surface temperature on average would be -11 degrees centigrade. It is in fact plus 15, over the whole earth. So you see there's a 34 [sicmy notes have a list of small blunders Mason made] degree difference due to the greenhouse effect, because these gases absorb the infra red radiation on its way to spaceonly about 8% of it escapes and gets right awaythe rest is absorbedmainly by carbon dioxide and water vapour (which is even more important that carbon dioxide by the way)and then it radiates a lot of that back to the earth again, and keeps the earth 34 degrees warmer than it otherwise would be. So that's the natural greenhouse effect, and if we didn't have that we'd have a frozen lifeless planet. So we mustn't knock the greenhouse effect!
The concern is whether man by putting additional CO2 and greenhouse gases into the atmosphere enhances that natural greenhouse effectovercooks it slightlyand gives us an additional warming.
As I said the carbon dioxide is increasing. There's no doubt about that. This [slide] is the curve of carbon dioxide in the atmosphere. It's increased 27% since 1860, but it's been measured every day since 1958 very accurately indeed, mainly in the mountains of Hawaii. You can see it's increased ever sincein fact, the curve gets slightly steeper, it's been accelerating, until just the last two or three years, it's levelled off a bit for the first time since 1958. That I think is due to the worldwide recession, the dash for gas, and the collapse of industry in the former Soviet bloc. It's 356 parts per million by volume now, and in 1860 it was 275. So if that continues over the next fifty years or so, we will come to something like double the amount of CO2 in the atmosphere. The second most important greenhouse gas in the atmosphere, apart from water vapour, is methane, and you can see methane has also increased in a rather similar way. (You might wonder why we have this rather nice sinusoidal oscillation. There's an annual cycle. I think there must be a lot of biologists in the audiencethat's because the trees are breathing. In the northern hemisphere, when they're growing they take up carbon dioxide, so it's a minimum in the summer and a maximum in the winter. That shows up rather nicely).
Uncertainties in Predicting Climate
[Back to top]
Of course, if we're going to predict what's going to happen in the future, we need to know what the emissions of CO2 and these other greenhouse gases are likely to be over the next 50 or 100 years, and that's almost anybody's guessit will depend on government policy, on what mix of fuels they're going to have, whether we're going into nuclear in a large scale
[At that time, few people, including Mason, were nuclear power sceptics; the link is to the 2011-2012 nukelies forum-Rae West] and so on. That's a very big uncertainty. It's almost idle to speculate as far as the actual emissions are concerned.
Even if we knew, we need to know what happens to those gases when they get into the atmosphere, because they don't all remain there. The amount of carbon dioxide that man puts into the atmosphere is about 5 1/2 billion tons a year by burning fossil fuel and wood. But only about half of it remains in the atmosphere, just slightly under half. Deforestation and cultivation probably puts in about another billion tons a year. 6.5 6.6 something like that. Only about 3.2 of that remains in the atmospherewe know that because we can measure it very accurately. So that leaves 3.4 billion. Much of that must go into the ocean, we know some is taken up by photosynthesis by plankton. The best estimate of that is about 2 billion. That leaves about 1.4 billion tons that go into the atmosphere that we can't account for. Some think more is taken up by regrowing forests and so on, and others that it's taken up in the soil, but the biological community is very split about this.
The fact that we can't account for 1.4 out of 6.6 is a measure of our ignorance about exactly what happens to the carbon when you put it into the atmosphere. That's another uncertainty.
We overcome that as meteorologists by saying, well, we shall just assume that the carbon dioxide increases, doubles after a certain time, or it increases at some arbitrary rate, and we'll put that into our climate models.
Now I mentioned that water vapour is by far the most important of the greenhouse gases. I said that the outgoing infrared radiation from the earth's surface is absorbed by these greenhouse gases. These gases radiate back to the earth 150 watts on every square metre; water vapour radiates 100 watts, carbon dioxide radiates 50 watts. So water vapour is twice as important as CO2 (it's practically never mentioned in the press). But it has another important consequence. Because if by putting carbon dioxide in the atmosphere you warm it, the warmer atmosphere will evaporate more moisture from the surface of the oceans, so you increase the water vapour. And therefore you get an amplification. So you have to consider not just the greenhouse gases that man puts in, but also the effect on water vapour.
Now, if we want to estimate what man has done since the industrial revolution, we can find out how much carbon dioxide is locked up in the little air bubbles in the ice sheets. You can go back tens of thousands of years, in fact back to the ice age, and see how much carbon dioxide there was in the atmosphere when the ice was laid down. And that's how we know that in 1860 it was only 270 parts per million. And you can do the same for methane. So you can estimate how much carbon dioxide has been added since 1760, rather early, or 1860, rather late, in the industrial revolution. If you do a simple calculation, how much greenhouse gas man has put in the atmosphere, and how much additional heating that would have produced, it's only about 2.4 watts from then until the present day. That would produce a temperature increase of about 1 degree. On that basis, if we look at the meteorological methods, you should find that surface temperature has increased about 1 degree since that time. Now, meteorological records aren't very good when you go that far back, but an awful lot of work has been done to reconstruct the historical record of temperatures since 1860. These [graphs on slide] show the differences from the average. You can see that there has been a gradual increase from then until the present day, 1994. That whole increase is about half a degree warmer, whereas the simple calculation would suggest it ought to be about one degree.
Now, is that evidence that there is greenhouse gas warming, or is it evidence just of natural fluctuations? Until very recently, I would have said that was due to natural fluctuations. You can see the last decade is the warmest on record, we've had about .3 of a degree rise, and there's a little falling off in the last few years, due to the Pinatubo volcanic eruption in the Philippines. You see how sensitive this is. But you'll notice there's an even bigger rise, of about .4 of a degree, from about 1910 until about 1950, when the emissions of greenhouse gases were much less than they are today. So you can't interpret the rise is simply due to greenhouse gases. A lot of that is due to natural variation, and in any case we shall see that you would not expect to see a strong effect yet because of the oceans; the oceans take up nearly half the carbon dioxide, they also take up heat; the oceans have an enormous inertia, so they store this heat and probably won't release it for another twenty or thirty years, so if there has been greenhouse warming due to man-made activities, we shall not see the full signal for some time to come. (Very recently I shall tell you at the endyou may have seen something in the press about thiswe've discovered that the aerosols, pollution particles, in the atmosphere also produce a cooling, and that also partly masks any greenhouse warming there might have been.)
Nature is full of these twists and turnsit's never quite straightforward. There are so many positive and negative feedbacks, some amplifying and some deamplifying the system, and all these various effects have to be looked at very carefully, and you have to look at the sum of these things, and that's why the story is changing very fast at the present.
Now, if we can't see strong evidence of heating due to man's activities in the meteorological records, and we may not be able to for a decade or so, what's all the fuss about?
The Earliest Computer Model of Climate
[Back to top]
The reason is, because the computer models have predicted that there will be a change. It's really the computer models that have brought up the subject, and alarmed people. Starting with the very simple computer models of the early 1970s, and now some of the enormous computer models I'll be describing in a moment, the biggest proper scientific models in existence. And we shall see how their predictions have changed quite markedly in the last five years or so.
If I were been talking to you five years ago, I'd describe the Met Office model, one of the world leaders (there were two in the US as well). We modelled in the 1980s the global atmosphere, coupled to a shallow ocean, just one kilometre deep. A lot of physics goes on in the deep ocean, and we didn't do that then, for computing reasons.
And the first thing you have to do with such a model is to see if you can make it simulate the present-day climate of the earthhow it varies from month to month, from season to season; if it gets the wind systems right, if it gets the rainfall right, if it gets the temperature right, and so on. It's got to do that before you ask the model what will happen if you make a small change, by increasing CO2 for example through man-made activities. Now it is a fairly small change, because there are 750 billion tons of carbon in the atmosphere, and man is putting in 6 or 7 [sic], so you see it's a fairly small change. And if you doubled the amount in one go, the increased heating would be about 4 watts per square metre, out of about 150 watts from the natural greenhouse effect, so you're looking at a few percent. So that model has to be very good at predicting the present climate. If it doesn't get that right, there's no sense in asking what happens if you change things by only a few per cent. So the first thing you have to do is built a model which, out of the basic laws of physics, will produce a climate on the computer which is similar to that we observe. It's the same problem as weather forecastingBracknell [in Britain - RW] for example predicts the weather for the whole globe every day for seven days ahead, and in fact we do the forecasting for New Zealand, and we can do it much better than they can do it. So that the models are similar in principle to weather forecasting models, except that in weather forecasting you're looking say 7 days ahead, but if you're talking about climate you're asking about average changes over decades or hundreds years, so you have to bring in a lot of extra physics. And you certainly have to bring in the oceans and the deep oceans.
I haven't got time to explain how we build these models, but I'll give you a flavour. What we mean by a model is a set of mathematical equations which just express the laws of physics. We try to describe all the physical and dynamical processes, and some chemical processes which go on in the atmosphere. We have Newton's equations of motion, we have the laws of thermodynamics, we have the transfer of heat by radiation, conduction, and convection, and we consider all the forces acting on the airwe calculate the east-west component of the wind, the north-south component, and the vertical component. When the wind blows, it evaporates moisture from the surface of the earth; it carries it up, the air expands and cools, the relative humidity of the air increases, it becomes saturated, and the water vapour condenses and produces cloud, the cloud droplets change and amalgamate to produce raindrops, or they freeze to become ice crystals and the ice crystals join together to make snowflakes, and they fall out as rain or snow. And on the way down you have to calculate how much it evaporates. You have to calculate how much runs off into rivers and lakes, how much flows out to the sea, how much evaporates back again, how much is stored as ice, how much is stored as soil moisturethis is terribly important to agriculture. So, you see, you have an enormous amount of physics. You have to consider how snow accumulates, how the ice cover spreads, how much the snow evaporates and melts; same with the sea ice. You even have to calculate how much the sea ice drifts. And so onto give you just some idea.
We take this on a sphere, and divide the atmosphere into concentric shells, like an onion. There might be ten or twenty of these levels from the ground up to, say, 100,000 feet. And then on each level you have a latitude-longitude grid. And so we have several thousand points at each level. And you solve, calculate, each of these processes every twenty minutes. You might take it up to a hundred years.
Newer Uses of Mathematics
(1978, Penguin) Ed. James Lighthill, a book dating from about the time Mason refers to. (He appears as Dr B J Mason).
Shows half the full stereographic projection of the northern hemisphereapparently this projection means that squares on the earth are represented by squares on the map, though not to exactly the same scale.
Lighthill goes on to explain that the atmospheric layers assumed by the model are in fractions of the weighteach slice for example weighing 1/10 the total pressure, so, because air density reduces with height, the upper slices are deeper. This presumably does
model the actual wind-shear, jet streams, and cloud levels found in practice, but is chosen for convenience.
Here's the old modelno deep ocean. We had 7,000 points at each level100,000 points in all, and these are some of the things we would calculate every twenty minutes: the components of the wind, the vertical motion of the air because that takes up and forms the clouds and the rain, the air temperature, the humidity, the heights of these levels, the incoming radiation from the sun, the long-wave infra-red radiation from the ground, the amount and heights of the clouds, how much water the cloud contains and how much ice, how much falls out as ice and snow, the change in pressure at the earth's surface, the land surface temperature, the soil moisture, the snow cover, the extent of sea ice and its depth, the ice surface temperature and the sea surface temperature. And many other things. Every 24 hours, just doing it for 1 day, was 1 followed by eleven noughts numerical steps; on the computer we had then the Cray MYP, every twenty minutes for one whole year took about ten hours. We can do much better than that now.
I'm just going to show you a few slides with the old model. I said just now that the model had to simulate the present climate properly. It doesn't do it absolutely perfectly! This slide is an average for June, July, and August, how the surface pressure patterns change. That's what we get from daily weather observationsthe Azores anticyclone, the Pacific anticyclone, the string of anticyclones in the southern hemisphere, the low pressure areas over the land. This is what we had with our crude model, at the end of the 1980s, and you can see that there's very good agreement. There are some small systematic errors but by and large it simulates the actual pressure. The temperatures are perhaps a little easier to see: the top is the computer model. The computer model got it slightly too cold over Greenland, for example. But it's pretty good. So that gives us confidence that we can produce the world's climate on a computer. We don't actually need any observations to do that! We can start at the beginning. We start with an atmosphere completely at rest; there's no winds, the temperature is the same everywhere, it's isothermal. It's completely drythere's no water vapour in the atmosphere; there aren't any clouds. We tell the computer the distribution of land and sea. And we say to the computer, switch on the sun now! And let the computer generate its own climate. If you switch on the sun, you create a temperate difference between the pole and the equator. If you do that, you set up a pressure difference between the pole and the equator. That makes the air move, so the winds start to blow. They start to evaporate moisture. About a year on the computer will generate a climate, just like I've been showing, from scratch. For this, you don't actually need any observations! You can start from Genesis, if you like. Isn't that remarkable! You can actually do that. Just from the mathematical equations of physics, you can create the world climate, without putting in any observations! You can't do that with weather forecasting!
Now, when you do that, you can then say to it, well, what will happen if we double the amount of carbon dioxide? In the early experiments, we doubled the carbon dioxide in one go and allowed the computer to adjust the climate itself, to come into equilibrium with double the carbon dioxide. I'll show you now what that model said would happen. This map is for the changes in temperature which you would get if you doubled the amount of CO2 for the months December, January and February, in other words winter in the northern hemisphere. The purple colour here is more than 12 degrees centigrade. And that's pretty alarming. As you go towards the Arctic, you get these very very high temperature increases, ten or twelve degrees, so that was pretty alarming. June/July/August in the southern hemisphere gives a rather similar map, but the other way round, with all the warming in the Antarctic, not much in the tropics. If you average for the whole year, over all latitudes, and over the whole earth, that would be 5.2 degrees centigrade. And the changes in the rainfall: the first thing you see is how patchy it issome areas have increased, some have decreased. Areas of major increase are here in the tropics and subtropics; and there's some drying out in Euro-Asia and parts of the prairies. The next slide shows the same pattern in June, July, and August. There were increases in the Monsoon areas and the tropical areas. And some decreases. Over the whole year, averaged, the increase in rainfall would be 15%. So we have 5.2 degrees rise in temperature, and 15% increase in rainfall. And that was the story five years ago. It looked rather alarming. We understood why there would be this increase in the polar regions. There are very good reasons physically.
Adjusting the Elementary Model, Mostly in a Downward Direction
[Back to top]
But of course we realised this model wasn't perfect. We only had a shallow ocean, so the deep ocean wasn't involved, and the deep ocean must play an important part in the long run. That was the first thing. Secondly, we knew very well that we weren't treating the clouds properly in this model. The clouds have a profound impact, both on radiation coming in from the sunthe clouds scatter back the incoming solar radiation into space. They also absorb some infrared radiation from the ground. So they play an enormously important part in regulating the heat balance of the atmosphere. In this early model, we did not ask the model to predict its own cloud, for technical reasons, including the lack of computer power we had at the time. So we actually told it what the clouds were. We have satellite observations from the clouds across the world every day. We actually imposed the observed clouds on the model. Now, that's a dangerous thing to do, because you don't necessarily get the two things into synchronisation, and the model didn't like that, and we knew that wasn't the right thing to do. So we knew we'd have to put a big effort into making the model predict its own cloud. And as you've heard, I was the first professor of cloud physics in the worldthe first and the last! One was enough, probably! So I had a very great interest in getting the clouds right in this model.
So the first thing we did was make the model predict the cloud in terms of the vertical motion of the air, and the relative humidity, and how much water condensed out. And the first step was to leave it all as liquid water. That reduced the warming from 5.2 degrees to 3.2. The next step: the higher colder clouds gradually transform into ice, and ice crystals affects reflection very differently from water drops. So we then calculated this gradual transformation of water into ice crystals, and the ice joining into snowflakes. And that reduced the cooling from 3.2 to 2.6. We took greater account of the shape of the ice crystals and so on. And that reduced it down to about 2.3 degrees. So just by changing the cloud alone, the global warming had come down by 5.2 to 2.3.
The next step was to do something about building a really complicated model of the whole deep ocean, and then couple that to the atmospheric part. For that we had to wait for the next generation of computers. We had a big effortit takes years to build one of these models with a big team; the centre at Bracknell has 80 scientists working full-time on this climate problem. So we built a complex model of the oceans and coupled that to the atmosphere. And now we could get away from this business of doubling the carbon dioxide in one go. That wasn't a very realistic thing to do. So we decided to add carbon dioxide at 1% per annum compound, which means it doubles after 70 years, and do it very gradually, like it does in nature. You can't ignore the deep ocean over seventy years. Well, that's been done. We have a model with 17 layers in the ocean. We calculate the ocean currents, the changes in salinity, the temperature changes in the ocean, and so on. And then we let the CO2 double gradually after seventy years. This map shows the average changes in surface temperature at the 70th year, when the CO2 has gradually doubled. And now you see there's a big difference. There are no reds and purples on this. The warming is very very much less. You can see the biggest change is 4 degrees in the Arctic. There's practically no warming in the Antarctic or in the southern hemisphere oceans. It's practically all in the northern hemisphere. And that was mainly the effect of having a deep ocean. The average temperature change in the 70th year was now 1.7 degrees. And if you'd like to see how that goes from year to year, you see [on a slide] the average for the globe varies from year to year, and there's 1.7 degrees after 70 years. What's pleasing is you get natural variation from year to yeara model, if it's any good, must get the natural variability. It's not a straight line, but the trend is upwards. You can see most of it's in the northern hemisphere. The reason is because the deep ocean in the southern hemisphere takes a lot of heat down to the deep layers and keeps it there. So what happens to the oceanthis is the surface of the ocean from north to south pole. Here's a vertical slice. It shows the changes in temperature you get. The maximum change is only about 1 degree. So the oceans absorb a good deal of heat. They reduce the heating, and delay it. That heat will come out eventually. So we're down to 1.7 degrees; that's the best estimate at the moment of global warming if you increase CO2 1% compound for 70 years until it doubles. Let me summarise by telling you about models elsewhere. The American model, which now also has a deep ocean in it, says 2.3 degrees. In Hamburg, they've just got their first coupled run, with 1.3. They're the lowest at the moment. We're right in the middle with 1.7. So it's somewhere in that range1.2 to 2.3as the best estimate at the moment for the carbon dioxide rise.
Now what effect would that have on sea level? You've all read these scare stories about melting the Antarctic ice and the sea level rising fifteen feet. It's all a lot of rubbish, of course. If you melt the sea ice, it doesn't make any differenceyou can melt all the sea ice and it doesn't change, if you know about putting ice cubes in your cocktail glass. You can't melt the Antarctic ice. It has to be land-based ice to make any difference and there's so much of it and it's so cold you can't make much impression on the Antarctic ice. Now the best estimates are that over the last hundred years the sea level has risen about 10.5 centimetres. And 4 of that comes from just expansion of the oceansif the sea gets warmer, it just expands. There's a little bit from the melting of the land glaciers, which we all know aboutthe recession of the glaciers in Switzerland and Europe. And the melting of the Greenland ice cap about 2 1/2. Totalling abut 10.5. Now what will happen if you double the carbon dioxide over seventy years according to our model? It says that the oceans expand about 2 cms per decade, melting glaciers 1, Greenland about 1. So you get 4 centimetres per decade. And you get nothing from the Antarctic. In fact, because of increased snowfall over the Antarctic, it probably goes the other way. So the Antarctic, instead of melting and raising the sea level, goes the opposite, and the sea level would drop. So that gives about 4 cms per ten years. So over the seventy years that's about 28 cms. Which is about a footso it's different from fifteen feet! I'm not saying that's negligiblein the lower lying areas like Bangla Desh and low-lying Pacific Islands that would be substantial. But it's nothing like the catastrophe that some people were talking about a few years ago.
Aerosols (in the Scientific Sense)
[Back to top]
Well, now, finally, the last part of the story is that during the El Chichón volcanic eruption in about 1986, and Pinatubo in 1992, we realised we have been able from satellites to measure the amount of radiation that is cut off from the sun by the volcanic eruption. Most of the stuff goes in the lower atmosphere and is washed out by rain in about a week. To have any effect on the climate it has to get to the stratosphere, where it stays for a couple of years or so. It was roughly 92 to 94%; then you have El Chichón, and it was down to about 78%. And this is Pinatubo, not quite as big, down to about 82%. So there's no doubt that dust up in the stratosphere, put there by volcanoes, mainly sulphate aerosols, scatters back the sun's radiation and stops it getting down to warm the atmosphere. So then we have to worry whether we could put into the model the effect of these aerosols, because they would have a cooling effect and partly offset any warming due to greenhouse gases. So that's what's just been done. This work hasn't been published yet, and its all very new, and the first thing that the Met Office did, that I was very pleased about, they started the model back in 1860, taking observations such as they were in those days, and running it without any increase in man-made CO2 or greenhouse gases, or any aerosols, whether man-made pollution or natural. They ran it to 1990; the temperature varies, but you can see there's no trend, no overall increase. Then they went back to 1860 and increased man-made greenhouse gases year by year, the best they could estimate, and ran the model again. And you can see by 1990 there was about 1 degree rise in temperature, which the model said would have been produced cumulatively by all the greenhouse gases introduced since the industrial revolution. I showed you the actual temperature only rose half a degreethe model says one degree, the observations half a degree. So they've run the model again with the aerosols in, and that's closed that gap. The actual rise is half a degree if you put the aerosols innot just due to volcanoes, which don't occur very often, but including man made, output, power stations and so on, that produce acid rainbut that's another story. This computed map of world climate has pollution twice as big as it actually is, to show up the effect. And this map shows areas of actual cooling, here the warming effect has been overwhelmed, these blue areas, you can see where they are, either over or downwind of the big industrial areas of the worldeastern United States, western Europe, China and southern Asia. So probably one of the reasons we haven't seen a global warming effect in the meteorological record is because it's being largely offset by the aerosols man has put in as well.
Now we have a dilemmawe want to get rid of the aerosols because of acid rain and asthma and all the other thingsbut if we do that, the greenhouse warming will be rather more serious. That's a problem that will have to be sorted out.
That was what Happened, Ladies and Gentlemen.
[Back to top]
That's the latest story, ladies and gentlemen. What's happened is that as we've improved the models, over the years, the threat from greenhouse gases has steadily come down, from 5.2 to a bit over 1 degree, or if you take aerosols in, probably a bit less than that. We have to be a bit careful about thatclearly the thing is not as alarming as it appeared five years ago. But the models are not perfect. They can still be improved. We're dealing with an extremely complex system, highly non-linear, so we may have some surprises, there may be some things we haven't taken into account yet. If it was going to be 5 degrees or more, I think governments would be right to think of drastic changes. But if we're down to one degree in fifty years' time we can live with that I suspect, we can adapt to that. In the present state, there's no scientific base for telling governments they've got to do something really radicaldisrupting industry, going nuclear in a big way
, vast expenditure on sea defences, moving populations, and so on. But that doesn't mean we should do nothing at all. Over the next twenty years we must do two things: put a big emphasis on improving the models and our understanding, and collecting observations and see what's going onboth from the atmosphere and the oceans so that we can detect greenhouse warm when it appears above natural variability. We don't have observations from the oceans. Satellites can do marvellous things: they can measure the ocean surface temperature within a tenth of a degree, they can measure the height of the sea from 800 kms up, they can measure the wave height, they can get the wind speed at the surface, but they can't see into the deep ocean. We've got to think about automatic unmanned submarines to get observations from the oceans. That would be very expensive. But we've got to do better science. We've got to think how to adapt to climate changeit will come. Genetic engineering for agriculture, energy transport, water supplyall these things which are climate-sensitive can be made more robust and resilient against climate change. It's not just man-made climate change you need worry about. At the moment you get much more disaster from natural climate changes [sic]the big storms, hurricanes, and all the rest. And we've got to be able to understand and predict those. So even if there was no man-made threat at all, even if this whole thing, this sort of bubble burst, we'd still need to carry on for predicting natural climate disasters.
So that's how I see it. In a way the politicians are in a very difficult situation. Several of the western powers have already decided at Rioit's about the only thing they did decidethat they would try and reduce the CO2 emissions in the year 2000 back to what they were in 1990. Well, that wasn't a very big deal. In the UK and most of western Europe, we'd already made that target, with the dash for gas, increased nuclear
, and some old heavy industries disappearing. That's no great big deal. What they do beyond 2000 they couldn't agree, and the Americans didn't want the imposition. The OPEC countries were dead against itthey just want to go on selling their oil. You'd think they'd have enough sense to put up the price and keep it in the ground! But of course the real problem is the world population problem. If the population is going to double in the next 50 years as is inevitable, there's no way we're going to be able to reduce overall global emissions of carbon dioxide. Most of that's going to come from China, India, the Pacific rim, Mexico. They're going to industrialise, to burn fossil fuel, because they've got nothing else. It's these countries that will decide. At some stageit may be at the end of the next century or beyondwe shall have a problem, a bigger problem than I've been talking about tonight. But perhaps by hat time we'll have a much deeper understanding of what produces these climate changes and a more intelligent attitude to what we can do to mitigate them or adapt to them.
Thank you very much indeed.
QUESTION/ANSWER SESSION BY JOHN MASON
[Back to top]
[NOTE: The transcriptions of these answers are incomplete; I've aimed to give the essence of Sir John Mason's replies]
Amazon nature reserve guides told me the climate had changedit was much more unpredictable. They say they've had climate change just from cutting down the rain forest.
They may well have had. What we've done in the modelling, we have actually cut down the whole of the Amazon rain forest in the computer! It reduces the rainfall and the evaporation by about 10 or 15% over the area itself, and the effects decrease as you go away. As far as local climate is concerned, you have to examine the detailairflow over the mountains and so on. In this climate modelling business, we're talking on a coarse, global scale. But people want to know about regional and subregional areas. Now the different models agree reasonably well on a global scale, but on smaller scalescontinental, sub-continental, regional, they differ much more. That's not surprisingbecause although we have 300,000 points in the model, they're still 250 x 375 km apart. If you want to bring that down say to 100 x 100 km, which is what they're hoping to do, that's 16 times as much computing. The latest version of the model with the deep ocean for one year is 2 followed by 14 noughts arithmetical calculations. The new computer at the Met Office has 16 central processors, each of which will do 1 billion calculations per second. All 16 processors at a time would take three hours to run the model for a year. And in the oceans the model's far too coarseyou have weather systems in the oceansfronts, depressions, and so on, but they're only one-tenth the size of weather systems in the atmosphere, so you really ought to have a computer grid one tenth the resolution. There's almost no limit to the amount of computing power you need. (The next generation of supercomputer always costs the same amount of dollars, as long as I've been in the game, about 15 million pounds; you get ten times the computing power each few years.) So the models at this time cannot really say anything on the local scale. Politicians are always being asked what's going to happen to the UK. Now that's a daft thing to ask and if I'd been Director-General they'd have been told that. That's stretching the models far beyond what they're capable of doing. They've got a long long way to go.
You emphasised uncertainty. But at the end of your talk you seemed to become more certain.
All these things have a cost; you have to make a balance between costs and benefits. It's not easy to do these studies. Over the next ten years, I'm sure the balance lies with putting the money into the science and into the observations. Or we've no basis for any kind of policy. We're hoping the models will converge, so we'll have a lot more confidence in them. There are about ten or twenty other groups starting up round the world, working on climate modelling, so they'll be much more activity. We're talking about maybe 1.5 or 2 degrees towards the middle or end of next century. So we do have time. We shall look foolish if we'd worked on the basis of 5.2 degrees five years ago. That has to be a political decision. I don't think it's absolutely urgent to turn policy upside down. I've heard people say with global warmingI heard one of the leading members of the Royal Society tell Mrs Thatcheryou can't grow potatoes in England any more! I said, for God's sake they grow potatoes in Egypt and Cyprus! You get that sort of statement which doesn't do any good at all. The countries which have one cash crop agriculture could diversify. That's the line the US took. Study by the National Academy of Sciences and the National Academy of Engineering said we don't care: if it rises three degrees we can live with it in the US. Then Al Gore, who as you know has written a book on carbon dioxide, not a very good book I must say, changed the American attitude to Rio to some extent. Of course we should save energy; we could design sea defences on an incremental basis, adding to them later. Adaptation is a very important factor.
Can I ask about Ice Ages on a large scale? And about Methane?
Methane absorbs infra-red about 25 times as effectively, molecule for molecule, as CO2; and CFCs about 10,000:1. But of course they're there in much smaller concentrationsCFCs in parts per trillion, methane hundreds of parts per million. Methane is the second most important. If carbon dioxide is say 60%, methane is about 15%. It's significant. It's increasing. If there was a major climate change, and the permafrost started melting, then a lot of methane would come out of the soil. Most at the moment I think is caused by leakage from gas pipelines in places like the Soviet Union and places like that. You get quite a lot from belching animals. We have to watch the methane.
The big ice ageswhere the temperature fell 5 or 6 degrees centigradethat's a different kettle of fish. They are triggered by the fact of the variations of the orbit of the earth round the sun. There are three components: the eccentricity of the ellipse, the inclination of the axis of the earth, and the precession of the perihelion. One has a 100,000 year cycle, one has 40,000 and one has 20,000. You can detect these cycles from deep ocean cores or deep ice cores. You can actually detect these cycles. I was the first, in 1976, to put these into an elementary computer model, the so-called Milankovitch effect, and satisfied myself that this triggered the big ice ages. They needed a bit of help from changes in the ocean circulation and changes in carbon dioxide, but I think everybody's agreed now that the real ice ages are changed, not by the output from the sun, by the change in the geometry so the amount that falls on the earth's atmosphere varies. That's really not controversial. There's more and more evidence every day from the paleoclimatologists who've done a marvellous job.
If the oceans warm as predicted, have there been any predictions of the effects on surface currents?
The models predict the surface currents. About six layers are near the surface. Of course the El Niño is an extremely good examplethe depth of the top layer suppresses the upwelling layer from below bringing up the nutrients, which is why the fish die. Up till now the models haven't been very brilliant at predicting the El Niño, largely because the spacing of the grid in the ocean is too coarse. It doesn't get the Gulf Stream or Kuroshiothe currents that matter from the point of view of ocean biologyright. They're too narrow. A model is being developed with a 50 km grid fine-mesh model embedded in the big one. I shall be very much happier when they predict El Niño rather more realistically. I think that'll come within a year or so.
What about CFCs and the ozone layer?
The CFCs together contribute, if you add them all together, about 12% compared with 66% by CO2. So they've been small but significant contributors to the greenhouse effect. But ozone is also a greenhouse gas. So if you abolish CFCs, and that's one thing that is coming down, you increase the ozone, and if you work it out, the two things practically cancel out. Of course there's little doubt that the destruction of ozone in the stratosphere is largely due to CFCsthe chemistry is very very complicated, with at least 200 reactions, many taking place on the surface of ice crystals in very high stratospheric clouds which only occur in the dark winter night, which is a devil of a job to study in a laboratory. You don't get these clouds very much in the Arctic. No model will predict the ozone hole without something like that in it.
So you see, the atmosphere, how clever it is! In the end the atmosphere is rather better a solving its equations than we are. It seems to have a very clever way of getting itself, to some extent, out of troubleit doesn't go too far from equilibrium. You do get these things like the ozone hole, which is an anomaly. It's serious, but it's not as serious as it's been made out to be. There's practically no change in measurement of UV-B, which has biological effects, at the ground. In fact, the ozone in the troposphere has increased, due to motor cars and the rest of it. It's the total amount of ozone that matters. So all this stuff about increased cataracts and skin cancer, due to that, is on very weak ground indeed. A lot of it comes from additional sunbathing. People start to build from a weak scientific base. Before you know where you are it gets out of hand. One has to take a very critical view of some of these things. You have to rely on measurements, and there's very little evidence at all of the UV-B getting to the ground increasing.
Does your model take account of water vapour. and how much does it matter?
Oh yes. Water vapour content is absolutely crucial in the models. We calculate the humidity at each point to get the clouds right and so on. That's crucial. Without it you won't get your clouds right, you won't get rain right, you won't get heat radiation right. It's more important than the carbon dioxide. The point is that CO2 is mixed uniformly, but water vapour varies!
You've just said the global scale is remarkably stable, resilient. Yet on the small scale it's chaotic. Where does the boundary lie?
If you're talking about deterministic predictionstarting with an initial observed state, and run your mathematical model to predict what will happen, cause and effect, that's deterministic. We know that somewhere about one or to weeks ahead the random turbulence destroys it. I was the first to say it, about 1957, that I thought two weeks ahead was about the limit we could make a useful forecast starting from an initial state. We're struggling to get beyond a week at the present day. The random motions accumulate and destroy the forecast. Can you say nothing beyond ten days, or two weeks? You bring in a probabilistic element; you don't say black and white. You start from an observed state and run the model forward two or three weeks. Then you change the initial conditions slightly and run it againwhat we call an ensemble forecast. You might have a dozen forecasts. Now, if they all keep very close to one another, you say that the atmosphere is in a reasonably stable mode, and you have high confidence in the forecast, and some determinism has been preserved. If they diverge, you've got very low probability. You have to start talking in terms of odds. That's quite valuable. It's better than saying nothing! That's the medium term.
Now the climate problem is very different. We're averaging out a lot of that shorter timescale chaotic movement.
But what was worrying me was that there might be a chaotic element in the global prediction, as well.
Yes, there'll certainly be some there. The climate model gives you natural variability, natural fluctuations. But if you keep taking averages, you take that variation out. [...] It took a long time to get the variability right-getting just the average isn't very meaningful. It was a lot of sweat, a lot of computer-time to get that.
You haven't said anything about the other gases in the atmosphere.
The variations in permanent gases, nitrogen and oxygen, are negligible. In any case they aren't active gaseswell, there are some weak bands in oxygen, but they aren't strong absorbers like CO2. The concentration of CO2 is uniformwell, there are small variations. But oxygen, nitrogen, argon, don't really come into it. Ozone does. Research into minor trace gases has been revolutionised. Remember 20 years ago we couldn't find the CFCs. Jim Lovelock invented the instrument for measuring things like CFCs, in a few parts per trillion. We didn't know those things were there! Impurities in water to one part per billionwe didn't know it was there, we couldn't measure it, before! We need a much more critical approach to some of these things, and get away from newspaper headlines. Too much science these days is being driven that way. And I have to say that some scientists are not above reproach in the way they capitalise on this, in my view. You remember 15 years ago we were going to have an instant ice age!
NOTES by Rae West
[Back to top]
I'm at present engaged in trying to present new ideas to the British Meteorological Office. Unfortunately, modern science is in a series of rutsand each section would rather waste millions, and billions, of other peoples' money than investigate ideas which might show them to have been wrong. So this may be a long job.
It's important to understand my claim here: I'm simply saying that modern computer models are too crude to predict anything about climate. It doesn't follow that there isn't, or is, warming or other effects; the methods just aren't good enough to tell us.
Mason is listed as writing 'Predictions of climate changes caused by man-made emissions of greenhouse gases: a critical assessment' in
, Vol. 36(5), pp. 299-319. I haven't read this, or seen the journal, but expect the contents to be similar to this talk's.
In case people don't know,
are both colourless and odourless gases, together making up most of our atmosphere, in proportion about 1:4.
is a form of oxygen which is more energetic and reactive than oxygen, and is made from oxygen (and returns to it) depending on physical conditions, usually in small amounts. (To illustrate the tiny amount of ozone, the weight of the atmosphere is equivalent to about three feet of lead. The ozone in it is equivalent to one sheet of the thinnest paper).
is an inert gas (there's more of it than carbon dioxide) which just stays there.
is just a gas formed by burning carbon; it isn't dangerous, except in the sense that it can asphyxiate.
is a simple hydrocarbon, which can burn and presumably therefore is somewhat unstable in the air.
are hydrocarbons with chlorine and/or fluorine added; they don't occur naturally.
is the way water is believed to exist when it evaporates.
The point John Mason was making about the sun's temperature and the earth's is connected with the idea of a 'black body radiator': the hotter it gets, the more the average wavelength moves, from for example red heat to white heat. Hence the sun gives light, but the earth radiates only humble infra-red. Both the atmosphere and the oceans are layered, though the reason isn't yet known: these are observations, not something deduced from (or explained by) theory. An 'aerosol' in the popular sense of advertisers, a squirty thing, is
an aerosol in Mason's sense; in the scientific sense an aerosol is tiny particles distributed in air (or other gas). Milankovitch seems to be credited with the idea that changes in the earth's orbit's eccentricity, the earth's tilt, and precession (slow apparent change of the pole star), affect climate, although some of these things were of course known of millennia earlier. Mixing of the gases in air is less even than implied in this lecture, no doubt to simplify matters. For example there is some separation between air circulation in the northern and southern hemisphere, as is perhaps shown by Hawaii's CO2 peak in the northern hemisphere's winter.
Mason seems to have made a few slips: I suspect UV-B should have been UV-C. And some of the figures seem wrong, e.g. 356 is not 27% more than 275, and 15 is not 34 more than minus 11! He understandably had some of the dates of eruptions wrong, and I've corrected these errors with the aid of reference books. Also he seems to have mixed up the weight of carbon in the atmosphere with the weight of carbon dioxide, which of course is more than three times higher, although nobody else in the audience noticed. I suspect some older temperatures may have been measured in Fahrenheit, causing confusion with Centigrade/Celsius. US billion is used throughout.
To give an idea of the crudeness of the early model: this had 7000 points at each level, making about 14 levels. The surface of the earth is 4 pi radius squared, or about 200 million square miles. This gives on average 30,000 square miles per point! A square of about 170 miles side. When you consider how weather can vary over a distance of several miles, this seems unlikely to be quite enough for a reliable model. In fact it seems unlikely that present techniques can model the situation satisfactorily; Mason doesn't discuss whether this aspect was discussed, or whether the practitioners just went ahead anyway.
It seems inaccurate to say the original model had no observations, and was purely based on the laws of physics, since, according to Mason, the clouds were imposed onto the model. So were the layers of air and of water. They seem to have neglected ionisation of air, which is important in thunderclouds. And neglected Amazonian trees. It's impossible to guess how much the model was juggled, possibly unconsciously, as it was developed to make known features of the world's climate fit. Presumably a theoretically ideal computer model would have a list of physical properties of all the ingredients of the atmosphere and the thermodynamics of their interactions, plus the energy inputs and output, along with a description of the total size and shape, of the surface features. Such a model ought to work as well with Venus, Mars, Jupiter, or for that matter the Sun, as with the earth.
As Mason states, there was a bandwagon effect, with a few dozen groups getting in on computer modelling. The expense is now running at about $4 billion a year. A small book by the British 'Institute of Economic Affairs' (pro-'free market')
Climate Change: Challenging the Conventional Wisdom
(1977), published at £12 but now remaindered for almost nothing, has four essays and two introductions looking at this subject, largely from a non-science viewpoint, and with many amusing quotations, for example on the IPCCIntergovernmental Panel on Climate Changefounded 1988. Although the contributors are aware of the influence of funding (if you were offered lots of money, would you say you weren't sure or didn't know?) for my taste they aren't cynical enough, and don't realise the same pattern appears in other scientific and social topics (as I've tried to show in the my website). They also are rather hopeless on science and aren't aware of the slender basis for much of the material. For example, temperature measurement is extremely difficult; many measurements from the past proved unusable because shielding precautions etc weren't taken. The sceptic will do well to doubt statements on average global temperature ten thousand years ago, or changes in sea level over tens of thousands of years, or changes in the sun's activity.
There's confusion with chaos, prediction, and determinism. I think Mason missed the point of one of the questionsbut, then, perhaps there's something wrong with the question.
On predictions by computer models, I recall that Britain had at least two economic models, which predicted different, probably opposite, things. Someone said in defence of this situation, what's the point of having two models if they say the same thing? Another amusing quotation I found is in a popular book by Paul Ormerod,
The Death of Economics
(1994), stating that the predictions of computer 'econometric' models are usually tweaked in an undisclosed way; and that a simple linear projection from last year gave results as good as the elaborate models. These models have an impressive record in failing to predict recessions, oil price rises, and so on. (This is of course not news: in 1975 the Institute of Actuaries printed an interesting paper on this subject, the author, P T Jenkins, concluding that omission of capital flows helped the models be useless). Ormerod said perhaps more accurately than he realised: '.. economics is perhaps more akin to subjects such as palaeontology, astronomy and climatology. The data are incomplete, subject to error, and there is only one actual history available for study.'
The impending ice age of twenty years ago: I have the November 1976
, a magazine with content curiously resembling Nazi pseudo-science, putting forward this view. Books included
(1976) by L Ponte, and
Climates of Hunger
by Bryson & Murray (1979). In 1981, Fred Hoyle got into the act with
Ice: How the next Ice-Age will come.
And the BBC promoted the same idea, in a series by a game show host from Iceland. Somewhat later the aerosol idea was extended into the idea of nuclear winter, apparently when Carl Sagan published on this, without, contrary to his own recommendation, any peer review.
Recently I was fascinated to find on Internet people who doubt whether ice ages could ever have existed, on the grounds (they say) that the estimated average temperature drop isn't very high, sea-level evidence disproves them, ice hasn't the right properties to move huge rocks, and there are psychological reasons in the US for wanting large drops in sea level: one site is Kurt Johmann's which has the following piece (copied, with permission, onto my site):
Debunking the Ice Age
. If I find more (including whether computer models predict, or retrodict, such things) I'll post it here.
On supercomputers, I think probably Ivor Catt's idea for linked processors could easily handle the situations Mason has in mind. Perhaps another twenty years..?
[Back to top]
Recommended: interesting long site by
John A. Daly
, in Australia. Emphasis on difficulties of measuring temperature, and artefacts caused by city heating. Also interesting political material on, for example, Thatcherism, Al Gore, and NASA. (Some of the graphics are a bit large).
Less recommended is this site,
which accepts the story of global warming and uses it to attack transnationals; not for waste, or damaging peoples' lives, or economic injustice, or corruption, but on account of this dubious story.
Review of a talk
As If There's No Tomorrow
by Mayer Hillman, 28th Jan 2000
[Back to top]
Mayer Hillman is the brother of Harold Hillman (elsewhere on this site) of whom I have a high opinion, so I was interested to hear this talk, delivered to a miscellaneous, rather elderly, local group of people. (Mayer Hillman told me that his views usually leave his audience shell-shocked; I told him he'd be doing well to keep them awake.)
This Hillman was billed as 'Senior Fellow Emeritus of the Policy Studies Institute', which appears to be a quasi-non-governmental organisation (or QUANGO) with the usual semi-secret funding. So this title is a good deal less impressive than it sounds. (I subsequently found the Joseph Rowntree Charitable Trust is one of its sources of money. Also that the global warming part of it seems to have been tacked on to the political part, and appears to be run by someone called Prof Jim Skea, though Jim was not polite enough to reply to a letter of mine). Hillman seems to have been head, or part, of the environmental part for about a decade. His previous work has dealt with such subjects as the uses of walking, and the drawbacks of 'daylight saving' time. He has co-authored many books.
I think I'm right in saying he has, himself, no evidence whatsoever; he is completely reliant on the views of 'climate scientists'. He says things like 'This is the biggest problem we have ever faced' and 'Evidence of its consequences [i.e. fuel consumption] in the form of exceptional weather patterns is accumulating at an alarming rate..' (It's necessary to talk about 'increasing evidence' of course because, for example, despite the industrial revolution many low-lying islands have evidently not been inundated. Hillman also has to fail to notice that, if his claims are true, it's absurd to talk of 'increasing evidence': he was saying there's no doubt he's right).
there is global warming, and if it's linked with CO
, then we must consider the impact on everyday life. (He means in industrialised countries, although there are occasional swipes at China).
His solution is to have some system for adding up peoples' allocation of carbonperhaps swipe cards with a credit limit. So that jet travel, for example, according to him a principal menace, would use up a lot of allocation. If you fly, you'd then freeze through lack of room heating. He seems to like draconian rationinghe was an evacuee to the countryside from London during the Second World War of which he is an admirer. He seems to like the thought of overall control: people in 1939 might not have voted for war with Germany. But the government knew it was necessarythis is an analogy he drew.
Like many non-scientific people, he confuses pollution with emissions, and temperature change with the ozone 'hole', and resource destruction with any use of resources.
An odd aspect of his talk was his attitude that God, or perhaps G-d, was coming to the help of the planet. Hillman looks forward to droughts in the American midwest, as this means the earth is fighting back. 'Global warming' is like the cavalry coming to the rescue of the earth. It's 'of near-Biblical import'.
I was interested to listen to the question and answer session; I counted the questioners showing awareness of science vs those who didn't, and arrived at nil awareness vs nine non-awareness. All the nine assumed that what he said was trueor perhaps they pretended too, overwhelmed by the colourful diagram Hillman had shown them, which showed some lines going up, and then down. I wondered whether, though possibly this impression is far-fetched, in a similar way, the inevitabilities of 'scientific Marxism' were expounded to rather dim-witted audiences, who were unable to examine the roots of the orthodoxy presented to them: what could be more important than global warming (or the oppression of workers?) My impression was strengthened by Hillman's attack on 'capitalism', which he claims will soon die out, because, presumably following Marx in an unthinking way, 'it must always expand'. He confidently predicts that in just a few years rationing will be introduced and 'capitalism' fall.
[Back to top]
Subject terms, keyword list:
aerosols, air, air circulation, carbon dioxide, CFCs, climate, climate change, cloud, CO2, computer simulation, Cray, emissions, forecast, global warming, ice age, ice, infra-red, infrared, John Mason, B J Mason, methane, modeling, modelling, model, nitrogen, oxygen, ozone, ozone hole, pollution, rain, snow, solar system, sun, sunlight, sunshine, supercomputer, UVa, UVb, UVc, volcano, water vapor, water vapour, weather, weather forecast
Brief outline first uploaded 98-01-18 Rae West. Full version uploaded for the first time 99-12-12. But worth the wait.
This revn 2000-12-10. Audio mp3 version link added 2014-08-04. Video version made 2014-10-13.
Mayer Hillman added 2000-02-13. Ice Age link 2000-07-11. Ormerod, Jenkins 2000-08-03. Daly, CompanyWatch 2000-08-19
Tape, transcription, Rae West. Editing, HTML, Notes © Rae West 1999.
This site is www.big-lies.org.