The heart of the matter

FAAM aircraft flying

1 March 2013

Helping to improve our predictions of flash flooding means unpicking everything that's happening inside a convective cloud - from the inside. Alan Blyth and colleagues describe a project that will use the best-equipped aircraft - and the strongest-stomached scientists - to do just that.

Convective clouds are designed to produce heavy rain, and we have all experienced how well they do it. When you're driving, in particular, it can seem like the heavens suddenly open then just as suddenly the deluge stops. But if it keeps raining in one place, or is particularly heavy (or both), there's a good chance of a flash flood. These can develop quickly over a small area, as we saw in 2012 in Devon when Clovelly's main street turned into a fast-flowing river.

Our ability to forecast where heavy showers are going to form, and to a lesser extent when, has greatly improved over the last few years. But we've made less progress on predicting how much rain is going to fall. To do this we need to understand in much more detail the physical processes in a cloud's complete life cycle.

We know that clouds can form along lines where warm and moist air converges, when the atmosphere is unstable and will allow warm air to rise. We know that cloud drops are formed by water vapour condensing around minute particles in the air (aerosol) and collisions between these cloud drops can create raindrops. We also know that ice particles are produced in the cloud and cloud drops freeze onto those particles which then melt to form raindrops. And we know that strong and persistent updraughts of warm air can cause heavy rain and new clouds building up behind older ones can make heavy rain last longer.

Imagine... plummeting at eight metres per second then being swept straight back up at twice the speed.

But these are just the basics. We still don't know exactly how ice particles are produced in the cloud, or the detailed physical processes responsible for turning them into rain, particularly heavy rain. Strange as it seems, the tiny cloud drops in the heart of a growing cumulus cloud do not freeze at 0oC as water does on the ground. The cloud drops remain liquid - albeit supercooled - until they encounter an ice nucleus to freeze around; but these nuclei are few and far between at altitudes of five or six kilometers. We also need to know how fast these crystals grow as more supercooled cloud drops bump into them and freeze - a process known as riming.

One reason it has proved so difficult to answer these questions is that a convective cloud is a hostile place and techniques like radar, which probe the clouds remotely, can only tell us so much. To get to the heart of the matter you have to fly right into the cloud.

Imagine being suddenly caught up in the air movements inside the cloud, plummeting at eight metres per second then being swept straight back up at twice the speed. Imagine trying to study particles the size of a pin-tip from an aircraft travelling at 200 mph. Imagine that aircraft being battered by raindrops at this speed. And as if that wasn't challenging enough, you have to take your measurements at just the right time in the life of the cloud to understand how ice particles form and grow.

This is what scientists in a new project called COPE (COnvective Precipitation Experiment) will be doing over south-west England this summer. For the first time, we will study the formation and growth of the particles that lead to rain while also learning about the larger-scale air motions in and around the clouds themselves. By better understanding the processes that control rainfall intensity we can improve the way these processes are represented in our forecast models - and improve the forecasts.

COPE will call on the NERC/Met Office Facility for Airborne Atmospheric Measurements (FAAM), whose aircraft is equipped with some of the most sophisticated research instruments in the world - instruments that can distinguish between liquid and solid particles at 200 mph. One goal of the project is to find these needles in the haystack - the first few ice crystals that form in amongst the hundreds of cloud droplets per every cubic centimetre of cloud.

The radar dish

COPE team members Lindsay Bennett and James Groves with the radar.

Another research aircraft, from the USA, will use unique on-board Doppler cloud radar and lidar (which uses light waves rather than radio waves). The radar can measure the properties of the small precipitation particles just as they are forming, as well as air movements inside the cloud. The lidar will tell us about air being brought into the cloud from outside, which can kill off the cloud but under some circumstances can actually set the whole rain process going.

When the size of the precipitation particles makes the cloud too dangerous for the aircraft, the new sophisticated NCAS radar on the ground will take over. Met Office radars and other instruments run by NERC at a Cornish base will tell us about, among other things, the larger-scale development of precipitation, properties of the aerosols in the clouds and the movement, relative humidity and temperature of the air below them.

That is only part of the story of course: all this information will only help our predictions if our models are up to the job. COPE will use several different models to help with interpretation of the data from the field campaign. The Met office will use the information to test and improve one of their models, which can warn of potentially dangerous weather down to areas of just 1.5km2.

COPE is a good example of collaborative science, with different organisations pooling their resources and expertise to improve weather forecasts and give the best advice to policy-makers and the insurance industry.

The impacts of this research will be significant, not least because the estimated bill for the 2007 floods in the UK was £3 billion. But the benefit to society of improved flood forecasts goes beyond money; if they can give people and businesses enough time to act before the flood arrives it will prevent misery and heartache too.


Professor Alan Blyth is director of weather science (part of NERC's National Centre for Atmospheric Science, NCAS) at the University of Leeds.

Phil Brown is cloud physics research manager at the UK Met Office; Professor Tom Choularton is head of the Centre for Atmospheric Science, University of Manchester; Professor Chris Collier is head of strategic partnerships at NCAS; and Humphrey Lean manages mesoscale modelling research at the Met Office.

COPE is a collaboration between NERC, the Met Office, UK universities and international partners.