AS NAILS in the coffin go, they don鈥檛 get much bigger: three independent studies have shown that climate sceptics who claim that Earth is not warming have been using faulty data to make their point.
The debate on climate change has often centred on the temperature of the lower troposphere. Common sense and computer models suggest that as the Earth鈥檚 surface warms, so should this layer of the atmosphere. But measurements from satellites and balloons did not always support this.
In 1992, John Christy of the University of Alabama in Huntsville analysed the satellite measurements and concluded that the lower troposphere had cooled over the decades, relative to Earth鈥檚 surface over the tropics. For those arguing against global warming, this analysis was pure gold. 鈥淭he data from the satellites have taken on almost iconic status,鈥� says Ben Santer of the Lawrence Livermore National Laboratory in California.
Advertisement
But the findings remained puzzling. 鈥淚t is very difficult to understand physically how the lower troposphere could be cooling while the Earth鈥檚 surface and the middle and upper troposphere were warming, as this study found,鈥� Santer says.
Now Carl Mears and Frank Wentz of Remote Sensing Systems in Santa Rosa, California, have an answer. They reanalysed Christy鈥檚 data and corrected for errors caused by satellite drift. 鈥淭he satellite is supposed to go over the equator and take measurements at the same time every day,鈥� says Mears. Initially this was at around 2 pm local time, but after a few years it was crossing the equator at 5 pm, he says. 鈥淐ommon sense tells you that it鈥檚 cooler at 5 pm than at 2 pm, and that was biasing the results.鈥� Once they factored this in, the data showed that the troposphere is warming (Science, DOI: 10.1126/science.1114772).
Mears and Wentz have strong support from Santer and his colleagues, who used 19 climate models to simulate the changes that would have occurred during the course of the 20th century. 鈥淒espite the fact that these models are all different in their physics, they all yield similar results in the tropics,鈥� he says. They all predict that warming at the Earth鈥檚 surface should be amplified in the troposphere. 鈥淭his makes sense physically,鈥� Santer says (Science, DOI: 10.1126/science.1114867).
鈥淎s researchers improved on ways to shield the balloon sensors from the sun, they recorded a trend of declining temperatures鈥�
Mears agrees. 鈥淭he only thing left for sceptics to point at would be the weather balloon data that also showed discrepancy with the models.鈥� And now Steven Sherwood of Yale University in New Haven, Connecticut, and his colleagues have shown that the balloon measurements are unreliable too.
The problem has to do with protecting temperature sensors from direct sunlight. 鈥淚t is just a little thing dangling from the balloon and there鈥檚 no way to shield it consistently,鈥� Sherwood says. Over the years, researchers found new ways to shield the instruments from the sun, but rarely bothered to make a note of this shielding or calculate its effect on the raw data. Improved shielding led to a drop in the temperatures recorded by the sensors, and this can explain the trend of declining temperature in the troposphere as recorded by weather balloons, Sherwood says (Science, DOI: 10.1126/science.1115640).
Christy welcomes Mears and Wentz鈥檚 analysis of the flaws in the satellite data he used. 鈥淭heir suggestions helped me fix my error pretty quickly,鈥� he says. His reanalysis now shows the Earth is warming by about 1.23 掳C per century. Mears and Wentz calculate the trend to be about 1.9 掳C per century.
The world is not warming as fast as the 1.5 掳C to 6 掳C per century that models suggest, Christy says. 鈥淲e all agree that warming is related to human effects, but it鈥檚 not as dramatic as models say.鈥�
Sherwood agrees that the debate will linger. 鈥淚 don鈥檛 think we have resolved the controversy over global warming,鈥� he says. 鈥淏ut there is no longer any data contradicting the predictions of global warming models.鈥�
