1 The climate forecast
How can scientists predict climate change in a hundred years, when they can’t even predict the weather next week? The answer will hopefully become clear in Activity 1.
Activity 1 Heads or tails?
Find a coin and a volunteer.
Ask your volunteer to toss the coin six times in a row. But before they start, ask them to write down a prediction for what the results will be: for example, ‘heads, tails, heads, tails, tails, heads’.
Were they right?
Chances are they were not! In fact, your volunteer had only a 1.6% chance of getting it right (0.5 multiplied by itself six times).
Now do the same again but ask your volunteer to make a different prediction: to write down how many heads there will be out of the six coin tosses.
Do they get this prediction right? If so, can you explain why this might be?
They are much more likely to get this right. This is because they are predicting the average frequency of heads over six coin tosses rather than making six separate predictions for each individual coin toss.
Does each type of prediction become easier or harder if you do the same activity with a larger number of coin tosses?
As the number of coin tosses increases, it becomes harder to predict the sequence of coins and easier to predict the fraction that are heads. The random fluctuations of a coin toss are ironed out the more times you try, so the average fraction of heads becomes closer to 50%.
Which of these is more like predicting weather, and which more like predicting climate, and why?
The first of these is like predicting weather because the sequence of specific coin tosses is analogous to a sequence of day-by-day events. The second is like predicting climate because the frequency of heads is analogous to the frequency of different types of weather.
However, predicting climate is – in one important sense – harder than predicting the statistics of coin tosses. In the coin toss activity, everything about the environment stays the same each time. This is why it becomes easier to predict the average number of heads as the sequence of coins becomes longer. In the real world, both human-caused and natural forcings will continue to change. This makes future climate harder to predict.
But the overall principle is similar – weather is a sequence of days and climate is a distribution of those days. This is an example of how the statistical definition of climate that you saw in Session 1 can contribute to confusion about how scientists predict it.