Even if the atmospheric temperature near the earth's surface has become cooler recently, that doesn't mean the planet as a whole isn't heating up
Imagine two people standing at the South Pole, one dressed in full Antarctic gear and the other wearing not much at all. Now imagine that you're looking through one of those infrared thermal imagers that show how hot things are. Which person will look warmest - and which will be frozen solid after a few hours?
The answer, of course, is that the near-naked person will appear hotter: but because they are losing heat fast, they will freeze long before the person dressed more appropriately for the weather.
The point is that you have to look beyond the surface to understand how a body's temperature will change over time - and that's as true of planets as it is of warm-blooded bipeds.
Now take a look at the two main compilations (see figures, right) of global surface temperatures, based on monthly records from weather stations around the world.
But according to the dataset of NASA's Goddard Institute for Space Studies (see figure), 2005 was the warmest since records began, with 1998 and 2007 tied in second place.
Tracking the heat
Why the difference? The main reason is that there are no permanent weather stations in the Arctic Ocean, the place on Earth that has been warming fastest. The Hadley record simply excludes this area, whereas the NASA version assumes its surface temperature is the same as that of the nearest land-based stations.
It is possible that the NASA approach underestimates the rate of warming in the Arctic Ocean, but for the sake of argument let's assume that the Hadley record is the most accurate reflection of changes in global surface temperatures. Doesn't it show that the world has cooled since the record warmth of 1998, as many claim?
Not necessarily. The Hadley record is based only on surface temperatures, so it reflects only what's happening to the very thin layer where air meets the land and sea.
In the long term, what matters is how much heat is gained or lost by the entire planet - what climate scientists call the "top of the atmosphere" radiation budget - and falling surface temperatures do not prove that the entire planet is losing heat.
Think again about that scantily clad person at the South Pole. If they put on some clothing, they'll appear cooler to a thermal imager, but what's really happening is that they are losing less heat.
Similarly, if you could look at Earth through a thermal imager, it would appear slightly cooler than it did a few decades ago. The reason is that the outer atmosphere, the stratosphere, is cooler because we've added more "clothing" to the lower atmosphere in the form of greenhouse gases like carbon dioxide.
As a result, the planet is gaining as much heat from the sun as usual but losing less heat every year as greenhouse gas levels rise (apart from the exceptional periods after major volcanic eruptions, such as El Chichon in 1982 and Pinatubo in 1991).
How do we know? Because the oceans are getting warmer.
Water stores an immense amount of heat compared with air. It takes more than 1000 times as much energy to heat a cubic metre of water by 1 degree Centigrade as it does the same volume of air. Since the 1960s, over 90% of the excess heat due to higher greenhouse gas levels has gone into the oceans, and just 3% into warming the atmosphere (see figure 5.4 in the IPCC report (PDF)).
Globally, this means that if the oceans soak up a bit more heat energy than normal, surface air temperatures can fall even though the total heat content of the planet is rising. Conversely, if the oceans soak up less heat than usual, surface temperatures will rise rapidly.
This is why surface temperatures do not necessarily rise steadily year after year, even though the planet as a whole is heating up a bit more every year. Most of the year-to-year variability in surface temperatures is due to heat sloshing back and forth between the oceans and atmosphere, rather than to the planet as a whole gaining or losing heat.
The record warmth of 1998 was not due to a sudden spurt in global warming but to a very strong El Nino (see figure, right). In normal years, trade winds keep hot water piled up on the western side of the tropical Pacific.
During an El Nino, the winds weaken and the hot water spreads out across the Pacific in a shallow layer. Its heat is transferred to the atmosphere. (During a La Nina, by contrast, as occurred during the early part of 2008, the process is reversed and upwelling cold water in the eastern Pacific soaks up heat from the atmosphere.)
A temporary fall in the heat content of the oceans at this time may have been due to the extra strong El Nino.
Since 1999, however, the heat content of the oceans has steadily increased again (despite claims to the contrary). Global warming has certainly not stopped, even if average surface temperatures really have fallen slightly as the Hadley figures suggest.
In the long term, some of the heat being soaked up by the oceans will inevitably spill back into the atmosphere, raising surface temperatures. Warmer oceans also mean rising sea levels, due to both thermal expansion and the melting of the floating ice shelves that slow down glaciers sliding off land into the sea. The West Antarctic Ice Sheet, which rests on the seabed rather than on land, is also highly vulnerable to rising sea temperatures.
Some climate scientists are predicting that surface temperatures will remain static or even fall slightly over the next few years, before warming resumes. Their predictions are based largely on the idea that changes in long-term fluctuation in ocean surface temperatures known as the Atlantic Multidecadal Oscillation and the Pacific Decadal Oscillation will bring cooler sea surface temperatures.
If these predictions are right - and not all climate scientists think they are - you can expect to hear more claims from climate-change deniers about how global warming has stopped. But unless we see a simultaneous fall in both surface temperatures and ocean-heat content, claims that the "entire planet" is cooling are nonsense.
And while a big volcanic eruption could indeed trigger genuine cooling for a few years, global warming will resume again once the dust has settled.
Climate Change: Want to know more about global warming: the science, impacts and political debate? Visit our continually updated special report.