"Previous studies have grossly underestimated methane emissions from the tropics," Paul Palmer of the University of Edinburgh told environmentalresearchweb. Palmer's work suggests that tropical wetlands contribute 52–58% of global emissions, with the rest coming from outside the tropics, including 2% from Arctic latitudes.

"The recent increase in atmospheric methane is due to surface warming of Arctic and mid-latitude wetlands and to a lesser extent increased precipitation over parts of Africa," added Palmer. "The work clearly highlights the dangers of this positive climate feedback."

As high latitudes warm under climate change, they are likely to release more methane, which will in turn boost warming, in what is known as a positive feedback.

The UK–Netherlands team estimates that Arctic wetland emissions increased by 30% from 2003–2007. "Five years of data is not enough to comment on a trend but we show that even with a modest amount of surface warming the methane release from these wetlands has affected global concentrations of methane," he said.

Wetlands emit methane because of anaerobic degradation of organic material by micro-organisms.

The team obtained atmospheric methane concentrations for 2003–2005 from the Scanning Imaging Absorption Spectrometer for Atmospheric Chartography (SCIAMACHY) instrument on the Envisat satellite. There was generally more methane present above continents in the tropics and at mid-latitudes.

The Gravity Recovery and Climate Experiment (GRACE) satellite, a joint project between NASA and the German Aerospace Centre, provided measurements of gravity anomalies, which the team related to groundwater height. Meanwhile, the team employed temperature data from the US National Center for Environmental Prediction/National Center for Atmospheric Research (NCEP/NCAR).

Palmer and colleagues found that methane emissions from tropical wetlands mainly depended on variations in groundwater whereas emissions from higher latitudes were more strongly related to temperature changes. The scientists used their findings from the 2003–2005 data in a simple model to estimate changes in wetland methane emissions from 2003–2007.

Palmer says it was frustration that led him to pursue this approach. "There is nothing in a methane concentration measurement that provides a clue of its origin," he said. "So we had to look at complementary measurements. I'd read about gravity anomalies being used to map changes in Amazon groundwater and something clicked: these data could be used to help to isolate surface emissions from wetlands which are controlled primarily by changes in groundwater and temperature. It worked."

The team estimates that wetlands emitted around 7% more methane in 2007 than in 2003, mainly because of temperature increases at extratropical latitudes (45° to 67° N). Emissions from Arctic wetlands increased by around 30% to roughly 4 Tg of methane a year. Tropical wetlands showed roughly constant emissions from 2003–2006 but increased by 2 Tg/year in 2007, because of higher groundwater volume in the Congo and Sahel.

Next Palmer and colleagues will analyse data from 2008 and 2009 to see whether the warming of the previous five years has increased, and whether they can correctly describe the changes in methane over this period. "In the longer term, we'll be trying to use the newer data from the GOCE [Gravity field and steady-state Ocean Circulation Explorer] satellite from the European Space Agency," he said. GOCE was launched in March 2009.

The researchers reported their work in Science.