Predicting Storms Before They Strike

Nothing has struck Europe this summer with the force of Hurricane Katrina. But there have been a number of pretty ferocious storms causing extensive flooding and structural damage.

British and German scientists have embarked on a major project to explore and explain the phenomena.

Unable to See Storms Soon Enough

The Convective Storm Initiation Project, a collaboration between academic climatologists and weather forecasters, involves six British universities as well as the German Aerospace Center and the British weather forecasting service.

Peter Clarke of the United Kingdom Meteorological Office says the project is designed to study thunderstorms more closely than ever before by improving instrumentation so that activity can be detected before storms take shape.

"At the moment we can see the thunderstorm once it is formed and perhaps we can see the clouds from satellites, but we can't see what's happening before the clouds form," Clarke explained.

Detecting Storms from Pressure Pockets

The project team has developed a massive new radar system which allows the scientists to see storms forming literally out of thin air. Thunderstorms and tornadoes happen when warm, moist air gets trapped beneath a layer of cooler air -- like a lid on a pressure cooker. So if this layer ruptures, it's like lifting the lid and warm, wet air rushes up into the cooler atmosphere above. Instantly water vapor condenses as thunder clouds and heavy rain falls directly beneath.

Prof. Keith Browning, Director of the Convective Storm Project, explained that the "piece de resistance," a very large radar, allows them to see more than the average weather radar that sees rain. Indeed, his radar sees clear air features, including the temperature and humidity gradients that reveal the presence of a lid, and can see air rising before it even produces a cloud.

Sudden, Hard to Predict Storms

With instruments deployed across the south of England, the project tracked the origins of a storm that rolled northwards and eventually caused a freak tornado in Birmingham. Professor Browning's team think they've identified the triggers -- air moving over high ground or cold streams colliding that lift the lid and cause a storm. Instruments like this have allowed scientists to study the summer thunderstorms and tornadoes of 2005 in unprecedented detail.

The 130 mile-per-hour whirlwind that ripped through Birmingham in 2005 was relatively small, but it still caused fifty million dollars worth of damage. There was even more flooding damage in the English seaside town of Boscastle in 2004 as well as extensive destruction in north Yorkshire in 2005. John Curtin of the British Environment Agency said the radar detected these storms too late -- after they had already started doing damage --because they developed very quickly from small, low-lying clouds. So, they greatly rely on the Meteorological Office for early warning signs, he said.

Increasing the Radar's Accuracy

The crux of the problem, however, is that the Meteorological Office can't forecast storms on as small a scale as Birmingham, Boscastle or north Yorkshire. Peter Clarke is working on computer models which could become the forecasting tools of the future. On a computer he's been able to recreate the Boscastle floods almost exactly as they occurred. But, according to Clarke, the current models are only good for around 12 kilometers. These models couldn't have predicted the storms in Boscastle because such small clouds (two to three kilometers across) aren't properly represented in the model, he said.

Clarke hopes for better models in ten years time "which will be able to give us perhaps a few hours warning of severe thunderstorms occurring in a particular place." He also hopes to increase the models' accuracy to about five kilometers.

Closely monitored, the events triggering summer storms in Britain will provide the first reality check for those future forecasting models. The goal is to predict precisely where the next destructive tornado will touch down and say not just whether it will rain but exactly where and how heavily it will fall.