September 21, 2009 – Original Source: New Scientist, by Michael Le Page
In fact, even if the world does cool over the next few years as some predict, it in no way undermines the certainty about long-term warming due to greenhouse gas emissions.
Let’s start with a thought experiment. Suppose you managed to find some children who knew nothing about the oceans, handed them a long measuring stick and sent them off to the seaside find out whether sea level is rising or falling.
As soon they saw the waves crashing on the shore, the children would realise they had been set a tricky task: how do you measure sea level when it is constantly changing?
If they were smart, the kids would try to find a quiet cove or harbour where there were no waves and start measuring. After an hour or so they would come running up to you. “The sea is going down! The sea is going down!” they’d shout in excitement.
“Not so fast,” you’d reply. “Keep going.” After a day or two, the children would realise the sea rises and falls at least once a day. If they were really dedicated and kept going for several days, they’d soon come running back to you.
“The high water mark is getting lower,” they’d declare triumphantly. “That means sea level is falling.” And once again you’d have to tell them to keep at it.
The moral of this story is that it is very difficult to detect underlying trends that are small compared with short-term changes.
Short and long term
We know, thanks to measurements taken all over the world over many years, that in the last decade of the 20th century sea level rose about 3 millimetres per year. That’s the equivalent of 30 centimetres over a century, which would be more than enough to cause serious problems for many low-lying areas.
But a rise in sea level of less than 0.01 millimetres per day is very hard to detect when the sea rises and falls several metres every day, influenced by the moon, sun and winds. A bunch of children equipped with a ruler would have no chance of detecting the “real” change in sea level even if they kept going for weeks.
People who claim we can stop worrying about global warming on the basis of a cooler year or a cooler decade – or just predictions of cooling – are as naive as a child mistaking a falling tide, or a spring low tide, for a real long-term fall in sea level. Just as the underlying change in sea level is swamped by the daily and monthly changes, so the annual variation in global temperature masks any underlying trends.
The up and down of waves can be compared with the day-night difference in temperature, the coming and going of the tides to summer and winter. Only when one stands back and looks at the changes over decades does the long-term trend become clear.
(This, by the way, is why model predictions for 2050 or 2100 are more reliable than those for 2015 or 2020 – the longer the timescale, the more the “signal” stands out from the noise.)
The confounding sea
Of course, while short-term changes in sea level can be predicted fairly accurately based on the motions of the moon and sun, it is a lot harder predicting the ups and downs of the average global surface temperature – there is a lot of noise, or natural variation, in the system. It’s becoming clear that changes in the oceans underlie much of this variation.
Predictions of global cooling in the short term are partly based on the idea that sea surface temperatures will fall in the northern Atlantic, due to slow, irregular swings in conditions known as the Atlantic Multidecadal Oscillation.
The bottom line is that, just as a few hot years do not prove global warming is real, neither do a few cool years prove it is not. Models suggest that it is perfectly possible for a decade or two of cooling to occur even when there is a long-term warming trend.
The strongest evidence for global warming comes from physics and chemistry, not from records of past temperatures, which is why scientists were predicting warming long before the rise in temperature over the 20th century was obvious.
Read more: Climate change: A guide for the perplexed