Met Office scientist Rosie Eade has received international recognition and an award for her landmark research on the predictability of Atlantic climate seasons for years ahead.
Much of the weather in the North Atlantic is dominated by the strength of the North Atlantic Oscillation – a pressure gradient between high pressure around the Azores, in the southern North Atlantic, and low pressure in the northern Atlantic, generally south of Iceland. The greater the pressure difference between the two areas, the greater the strength of the NAO and therefore the greater the impacts for Europe and beyond. A strong NAO tends to bring more warm and moist air from the west, bringing wetter and windier weather to northern Europe.
Rosie Eade and her fellow collaborators realise that the potential for accurate forecasts (‘predictability’) of regional events, such as the NAO, is likely to have been previously underestimated.
The significance of Rosie Eade’s work has been recognised by the World Meteorological Organization, which has awarded her the prestigious WMO Professor Mariolopoulos Trust Fund Award for 2016 for her paper: “Do seasonal-to-decadal climate predictions underestimate the predictability of the real world?”
Adam Scaife, the Met Office head of long-term forecasting said: “This highly original paper demonstrates that predictability of climate out to years ahead is likely to have been previously underestimated, particularly for the Euro-Atlantic region in winter.”
The research shows that climate predictability is underestimated in many studies because the strength of predictable signals in Atlantic winter climate is underestimated in climate models, and individual climate predictions out to a decade ahead match each other less than they match the real world.
Adam Scaife added: “This radical discovery means that apparent uncertainty is overestimated in Atlantic winter climate predictions, with the potential to underestimate the predictable influence of the North Atlantic Oscillation on our weather in winter.”
Rosie said: “I am excited about our area of research as we are developing predictions that we can personally verify, but further ahead than standard weather forecasts. In fact, far enough in advance that we have the potential to help communities prepare for weather and climate-related impacts around the world.
“This can mean, for example, providing guidance to city councils to aid planning for the coming UK winter, or to the UK renewable energy industry to aid planning for the coming years in terms of changes in supply related to wind, or demand related to temperature.”
The team of scientists find that current climate models can agree well with the real world, but that the chaotic element within the model itself can make it hard to see the year-to-year variation of the North Atlantic Oscillation.
If a climate model is run enough times resulting in multiple predictions for a single event – for example, for next winter’s NAO index – then averaging over all the runs can give a useful ‘ensemble’ forecast, and these agree well with the real world even though the strength of the ‘signal’ may initially appear low.
Rosie added: “This is a little like trying to capture the melody from a crowd of singers who are each slightly varying the song in their own way – the signal (song) is there but you need plenty of singers to be sure to average out the differences and so pick out the intended tune.”
This result holds much promise for the future as it shows that there is scope for further improvements in predictions from months to a few years ahead.
The study has stimulated great interest at scientific conferences and is now inspiring others in the field.
The paper “Do seasonal-to-decadal climate predictions underestimate the predictability of the real world?” by Rosie Eade, Doug Smith, Adam Scaife, Emily Wallace, Nick Dunstone, Leon Hermanson and Niall Robinson is available online.