"These results suggest improved seasonal forecasts in winters where a sudden stratospheric warming is likely to occur," David Fereday of the Met Office told environmentalresearchweb. "This would allow warning of an increased chance of cold conditions with a lead time of up to a few months, allowing better preparations to be made."

But what causes a sudden stratospheric warming event? Each winter a vortex forms in the stratosphere over the Arctic, Fereday explained, with cold polar air at the centre surrounded by strong westerly winds. In about half of all years, this vortex breaks down in a sudden stratospheric warming event; much warmer air moves in from lower latitudes and the vortex is disrupted.

"The breakdown of the polar vortex causes a reduction in the westerly winds in the stratosphere, generating a signal that can often propagate down to the surface over the course of a few weeks," said Fereday. "This reduces the occurrence of surface westerly winds that bring mild air to northern Europe in winter from the North Atlantic. Instead, northern Europe experiences cold and blocked conditions."

In order to predict such events better, the Met Office has begun using a so-called "high-top" version of its GloSea4 seasonal-forecasting model. This version calculates physical quantities such as winds, humidity and temperature to higher levels in the atmosphere and also at more levels, "so it is able to simulate conditions in the stratosphere – and hence sudden stratospheric warmings – more realistically".

The high-top model improves the accuracy of forecasts in winters where a sudden stratospheric warming is likely to occur, the team found. "If the high-top model had been in use for winter 2009/10, where sudden stratospheric warming occurred, it could have more accurately predicted the increased chances of a cold end to the winter a few months ahead," said Fereday.

Now the team is continuing to develop the Met Office seasonal-forecasting system. "The 2010/11 winter was also cold, and was also successfully forecast," said Fereday. "Different forcing factors were involved in this winter compared to 2009/10, and are the subject of a forthcoming paper by members of the monthly-to-decadal forecasting group. Research is also ongoing into other factors that contribute to potential predictability on a seasonal timescale, such as sea ice and soil moisture."

Fereday and colleagues reported their work in Environmental Research Letters (ERL).