About the Author

Avatar photo

Professor Michael Kelly FRS FREng

The Challenge of Climate Prediction

Print Friendly and PDF
Posted on

You may have heard in recent weeks that 2015 is shaping up to be the warmest ever based on the hottest recorded July.  In fact we have the same positive El Nino weather pattern in the Eastern Pacific that we had in 1988, the hottest whole year on record to date.  This in spite of the record cold winter in some parts of the northern hemisphere, and the winter in New Zealand has not been as mild as in recent years.   If 2015 does turn out to be the hottest, it will only be by one-to-three tenths of one degree Celsius or less, a small fraction of the actual error associated with estimating the globally averaged temperature which is greater than half of one degree.

Those concerned with rising temperature will be quick to quote President Obama saying that the last 15 years are the hottest on record.  But I assert that the global temperature has not changed significantly in 15 years.  How can we both be right?  This is a point I have had to explain more than once, and a simple analogy will suffice.

Imagine you setting out to walk from the southern edge of Lake Taupo to the Chateau using the state highway.  At first you climb, up to the level of the Desert Road.   As you proceed you are then on the flat so you are not climbing.  You can still say that you are on the highest point of your trip, but it is approximately the same for several miles.  That’s why President Obama and I are correct so far.  The crunch comes in the future, as to whether you turn off and go on up to the Chateau, or you find your way to a local river and go down to the level of Lake Taupo.   The recent behaviour of the weather of itself does not tell us what will happen next.  We rely on our understanding of the weather and our ability to forecast ahead.   This is where President Obama and I part company in terms of the certainty about what happens next.

The atmosphere is a complex, tightly coupled, highly non-linear multi-component system which is rapidly changing and it is also chaotic.  Predicting its future properties in 35 years is like trying to predict the shape of a particular rain cloud in 35 minutes as it scuds over the sky on a windy day.  No one can have any great confidence in specific predictions.  In 2004 the Met Office in the UK was >90% confident that the temperature in the UK would rise by 0.3C by 2014: i.e. nearly half of the total rise of 0.8C in the 20th century*.  In the end there was no rise at all over that period.  The confidence in the quantitative predictions of the models misplaced then, as it still is now.

President Obama said that 2014 was the warmest on record.  That is disputed in many quarters.    It was as warm as many other recent years, but not significantly more so.  When you can only measure the temperature to an accuracy of 0.1C, you are not allowed to use a difference of 0.02C from several averages to quote a positive increase – it is well within the error margin.

These are very important points.  It is 25 years now since climate model predictions have been made, and the vast majority have been overestimates, some gross overestimates, of temperature change.  Every freak weather event is now, quite erroneously, attributed to man-made climate change. I retain faith in the self-correcting nature of science, and if the temperature does not rise for another five years, climate scientists will be back to the drawing board, as the probability of such a temperature stasis over 25 years in the models is vanishingly small.  Watch this space.

Where I still have some hope, but am less confident, is that the present madness of politicians subverting the world economy by putting a severe cap on the carbon emissions that brought the western countries to their present level of prosperity, but denying it to the Chinese, India and African people who aspire to the same prosperity, and the only way they will achieve it is with more fossil fuel use in the next 20 years.

Why might we return to the level of Lake Taupo, as opposed to climb to the Chateau?  As I wrote in the Times (of London) on 14 October 2010: “One possible scenario for 2050, no less possible than any projected on the basis of climate models, is that we are in the middle of a deep solar minimum, and it is only the CO2 pumped into the atmosphere over the previous 100 years that is staving off cold climates that would lead to crop failure and mass starvation. The uncertainty is not only in the science and in the scenarios, but in what is a reasonable response today.”    Nearly five years later the sunspot cycle is very weak.  In the last solar minimum we had the little ice age, and many of the one billion people died.  This time we will have nearly nine billion humans to feed.

But shouldn’t we take future climate change seriously, just in case, and pay an insurance premium?  The short answer is that we should not pay a premium that would bankrupt the world economy: any more than we take prudent measures about earthquakes, so we should take only prudent measures to protect against future climate.  The real problem is that there is a common belief that cutting the carbon emissions will actually solve the problem: it won’t necessarily, and there is no proof.