Scientists developed a forecasting method for predicting the size of the annual summer melt of Arctic sea ice, which could forecast three months before it hits its September peak, according to a research published in the journal Nature Climate Change.

The British research team, led by Reading University researchers, found a strong correlation between the fraction of the floes covered by pond water in May and the eventual sea-ice extent seen in September.

Ponds are darker than ice so they absorb more energy from the sun, promoting further melting. This team developed a model to forecast the evolution of melt ponds in the Arctic and has incorporated this into more general climate sea-ice models.

Using the new method, they predicted that the September 2014 sea ice minimum will be 5.4 million square km, close to the level in 2013.

While this would be higher than the all-time low in 2012 of 3.4 million square km, it would still be 18 percent below the 1981-2010 average of 6.5 million square km, indicating a continuation of the long-term decline, according to experts.

"Melt ponds are crucial to the speed of the annual ice melt. But until now, there has not been a physically-based melt pond model." Professor Daniel Feltham, who led the study, said that: "This system could be used by the shipping, tourism and oil industries, which are all looking to exploit opportunities that new passages through the Arctic could provide."

He added that, their work could also help global climate models become more accurate, further strengthening researchers' confidence in how the climate will change in the years and decades ahead.(Xinhua)