Gravity waves help predict invisible turbulence

10 October 2008

It's always an alarming moment when the plane you're travelling in plunges what seems like several hundred feet in the middle of a clear sky.

Usually there's no warning, and passengers are left wondering if their number is up until the pilot provides reassurance a few seconds later. Every year this kind of clear-air turbulence hurts travellers and does millions of pounds worth of damage to planes.

In one recent case, 11 passengers on a China Airlines flight suffered injuries as serious as broken necks after hitting a patch of turbulence near Bali. Until now the methods airlines use to forecast this danger have enjoyed only limited success.

But scientists at the University of Reading hope to change that with a new way of forecasting turbulence in clear skies. This is no easy task because clear air turbulence happens far from any obvious weather systems. But the researchers say their model can predict many areas of turbulence based on standard meteorological data such as wind speeds and pressure readings.

'Because clear air turbulence is invisible, it's very hard for pilots to avoid,' said Dr Paul Williams of Reading University's Walker Institute for climate research, who co-authored the paper. 'We hope that our model will allow them to be aware of likely areas of turbulence as they plan their flightpaths before takeoff, and that this will help avoid serious injuries to passengers and damage to aeroplanes.'

At present, the US Aviation Weather Center is one of the leading forecasters of turbulence. But its model contains many rules of thumb and approximations. And it doesn't account for one of the main causes of turbulence.

Published last week in Journal of the Atmospheric Sciences, the new research takes a fresh approach to the subject. It uses a mathematical model to predict the occurrence of gravity waves - upper-atmospheric phenomena that are responsible for much turbulence. Occasionally these waves can be spotted as a striped pattern in clouds, but they can't generally be seen.

'Most clear air turbulence is thought to be caused by gravity waves, and our method lets us predict where these are likely to appear,' explains Williams. 'We hope the method can be added to the existing prediction techniques, which date from the 1960s when the last major theoretical advance was made. This won't be the end of clear air turbulence, since even our new method only predicts it around 80% of the time - but it should lead to a marked improvement.'

As luck would have it the US government's target, hitherto unattained, is to predict turbulence 80% of the time, so the new method's performance, achieved by testing it against five months of historical data, could see it adopted quickly.

The model can be run for the whole world in real time and needs only widely-available weather data as its inputs. Although in the short term Williams envisages pilots being warned of potential areas of turbulence ahead of takeoff, the model could ultimately be used to provide real-time updates so that pilots could alter course to avoid risky areas.

The areas the model highlights are small enough for this to be practicable - around the size of a US state, or smaller, according to Williams.

The research was part-funded by the Natural Environment Research Council (NERC).