Skip to content

The basics of climate prediction

Updated Monday, 13th February 2006

How can we say with any degree of certainty what the climate will be like in the future? Here's an introduction to the basics.

This page was published over five years ago. Please be aware that due to the passage of time, the information provided on this page may be out of date or otherwise inaccurate, and any views or opinions expressed may no longer be relevant. Some technical elements such as audio-visual and interactive media may no longer work. For more detail, see our Archive and Deletion Policy

We like to talk about the weather, to complain about its variability and to blame the weather forecasters for getting it wrong. But what is 'climate', and how does the weather we experience on a day-to-day basis relate to climate change, a subject which is increasingly dominating our newspapers and television screens? Why is it that we can't make a perfect weather forecast? And how can we hope to predict the climate of the 21st century, when we can't say what the weather will be doing in a week's time?

To start with, what affects the climate of the Earth? Why has the temperature of the Earth stayed approximately the same over very long periods of time, without varying by hundreds of degrees?

The temperature of the Earth can be represented by the level of water in a bucket which has a hole in the bottom and water flowing in from a tap at the top...

Click on the arrow to find out how it works.

The amount of water flowing out of the hole in the bucket is determined by the depth of water in the bucket - the higher the water level, the more water is forced out of the hole, so if you open the tap, the water level rises until the amount of water leaving the bucket is again equal to the amount coming in from the tap.

But how does this relate to the climate?



Our tap pouring water into a bucket with a hole is a simple way of looking at the temperature of the Earth.

The water flowing from the tap represents the energy from the Sun, the water escaping from the hole in the bucket represents the energy the Earth loses to space, and the water level in the bucket represents the temperature of the Earth.

The hotter the Earth, the more energy the Earth loses to space. If the Earth were to get more energy from the Sun the temperature of the Earth would rise, just like the water level in the bucket, until the amount of energy it's losing to space is again equal to the amount of energy it's getting.

At its simplest level, whether or not the Earth is heating up or cooling down is determined by the difference between the amount of energy the Earth is getting from the Sun and the amount of energy it is losing to space.

Since the Industrial Revolution, the increasing amount of greenhouse gas in the atmosphere has been restricting the amount of energy the Earth is losing to space - as if the hole in the bucket were partially blocked. As a result, the Earth is heating up. If we were to stop increasing the amount of greenhouse gas in the atmosphere, the Earth would eventually (in a few decades) reach a new, constant, warmer, temperature.

But how does climate relate to the weather?


So, how does the climate relate to the weather we experience on a day to day basis? We know from experience that the weather can be very different from one day to the next, let alone from one year to the next, without any change in the climate.

Suprisingly, dice are a good way to think about the difference between weather and climate...

The animation below allows you to choose how many times to role a dice and then see how often you get each of the six sides. Try a low number of rolls, then try some larger number of rolls and see what happens:

Throw the die a few hundred times. What is the average (mean) of the scores? The more throws, the closer the average gets to 3.5. If you were to throw the die one more time, you would not be able to predict the number that the die would land on, as the probability of throwing each number is the same. However, you could be very confident that the mean would still be 3.5.

But what has this got to do with weather and climate?


What if we associate weather types (for example, cloud cover) with each number on the die?

Try rolling the dice in the animation, again explore what happens as the number of rolls increases.

As when there were numbers on the sides of the die, you can't predict what the weather will be on the next throw. Climate is defined as being the average of the weather over a long (typically 30 years) period of time. The 'climate' of this die is 50% cloud cover. A single throw of 0% or 100% cloud cover won't affect the climate very much if you are taking the average of 100s of throws. In the same way we can have a very hot summer one year, and a very wet one the next, without the climate, the weather we expect to happen, necessarily changing.

"Climate is what we expect, weather is what we get"

So why do the weather forecasters never get it totally right? Mostly because the weather is a 'chaotic' system.


So, what's a chaotic system? The animation below uses the analogy of rabbits breeding to illustrate the difference between a chaotic system and one that isn't. Let's start out with a non-chaotic system - simply adjust the slider on the left to pick a starting number of rabbits in the warren and watch how the population changes with time. Try a few different starting numbers and see what the impact of changing this is on the population size.

Whatever number you choose, the population should grow until it reaches 32 and then stabilise - this is the maximum number of rabbits this warren can support.

But, the weather isn't this simple - it's chaotic. What does this mean? Let's find out by having another go at the animation. By clicking on the 'chaotic model' tab we can then see what the impact of changing the starting condition (the initial rabbit population) is.

As you should have seen, this time, the number of rabbits in the warren varies haphazardly. How many rabbits are there at the end? Now vary the starting number by just 1 and start the animation again. How many rabbits are there at the end this time? Even though the change in the initial population size was very small, the final population size can be very different.

The weather is also chaotic - very small changes to the starting conditions can lead to completely different weather patterns developing. This observation led Ed Lorenz to suggest that the flap of a butterfly's wings in the Amazon rainforest could lead to a tornado in Texas. It is very unlikely, but it could.

This means that, to make a perfect weather or climate forecast, we need to know what the atmosphere is doing currently, down to the scale of individual butterflies flapping their wings, which is obviously impossible!


So, since tiny changes in the starting conditions of a weather system can make significant differences to the outcome, when making a forecast we have to try to take into account what might be happening now, as well as what might happen in the future to affect the atmosphere. The best we can do is to produce a range of forecasts, with some indication of what is most likely to happen.

To help illustrate this, consider throwing two dice instead of one:

With two dice, the probability of throwing a combined score of a number between 2 and 12 is not the same. There is only one combination of number that would give you a 2 or a 12 (two 1s or two 6s respectively) but, for example, for a combined score of 4 you could throw a 3 and a 1, two 2s or a 1 and a 3 - so you are 3 times as likely to throw 4 as 2 or 12. There are most possible ways of throwing a combined score of 7, and no way at all of throwing a 1 or 13 or more.

Move the slider to pick a number and throw the dice a large number of times. Notice the shape of the graph that is produced - the middle numbers are rolled more often than the smallest or largest numbers.

This sort of shape of graph is very common. For example, temperature measurements will often show a similar distribution, although temperature can of course take any value, not just the numbers one to twelve.


The animation below explains how we can say some temperatures are likely, some possible, and some temperatures are impossible, by looking at what we've experienced in the past. You can skip backwards and forwards in the animation by using the slider:

In this way, the results of many climate forecasts can be combined to show what is most likely to happen, what is unlikely to happen and what almost definitely won't happen.


What do recent climate prediction experiments tell us about our likely climate in the future? The map below shows predictions for how the climate might vary over the next century in various regions of the world, given two particular scenarios for how the world might develop, and how much fossil fuel it might burn.

You can click on the regions highlighted in blue to show the prediction for the years 2050 and 2100 in that region. You can also see how the predictions are affected by basing their forecasts on a future with lower emissions.

The results are different for each region. In some areas, we can be more certain about what will happen - the horizontal spread of the 90% region is narrower. Also notice that, while we can be more confident about what will happen by 2050, the most likely temperature rise by 2100 is greater than in 2050. By 2100, the temperature will probably have increased dramatically.

But what about extreme events? How will the likelihood of an extreme event change as the climate warms? It is never possible to attribute one particular event to a particular cause. To go back to the dice example, you could load a die so that sixes occur twice as often as normal. But if you were to throw a six using this die, you could not blame it specifically on the fact that the dice had been loaded. Half of the sixes would have occurred anyway, even with a normal die. Loading the die just doubles the odds of throwing a 6.

Lake on surface of the ice Copyrighted  image Icon Copyright: Production team

In general, if the climate warms, the whole bell shaped curve of temperature for a particular place shifts to warmer temperatures:

Figure 4.1 from the SYnthesis report of climate change 2001 (
Taken from the SYnthesis report on Climate Change, 2001,


Record hot events are more likely in a warmer world, and record cold events are less likely.

So, for example, we can say that the hot summer of 2003, which killed 22,000 - 35,000 people in central Europe, is twice as likely because of the global warming that has resulted from the man-made emissions of greenhouse gases. By 2050, we can expect summers as hot as that every other year.

Image by Reto Stckli, Robert Simmon and David Herring, NASA Earth Observatory, based on data from the MODIS land team.
Summer 2003 temperatures relative to the average of 2000 - 2004 summer temperatures

Similarly, in the U.K., we can expect the number of extremely rainy days, with associated flooding, to increase. Already, the kind of rainfall that you could have expected once every 30 years in the 19th century is happening once every 12 years now. By the end of the century, it could be expected every 4 years.

Earl Steet flood [Image: Sylvia Knight]

There has been a lot of debate recently about whether the number of hurricanes and typhoons has increased because of the effects of global climate change. In general, in a warmer world, you can expect the sea surface temperature - one of the key factors in hurricane development - to be warmer.

There is already evidence to link an increase in the power of hurricanes that have happened in the recent past to increased global temperatures - a recent study has suggested that the power of hurricanes has almost doubled over the last 30 years. Hurricanes will probably become more destructive in the future.

So, to summarise:

  • Even with perfect forecasting techniques, we could never say exactly what the climate will do over the next century. This is because:
    • weather is chaotic
    • we don't know how the world will develop and how much greenhouse gas will be emitted
    • We don't know what other, natural, factors may affect the climate in the future - volcanic eruptions, changes in solar activity etc.
  • We can, at best, say what the climate is most likely to do, and what it probably won't do.
  • The longer into the future a forecast is made, the less certain you can be about what will happen.
  • We can expect extreme events - such as abnormally hot seasons and storms, to become more frequent in a warmer world.

All the above animations were developed by and The University of Oxford Department for Continuing Education (Technology Assisted Life-long Learning Unit) as part of their climate prediction resources, and are used with permission.

The BBC and the Open University are not responsible for the content of external websites







Related content (tags)

Copyright information

For further information, take a look at our frequently asked questions which may give you the support you need.

Have a question?