An aeroplane-mounted 3D imaging system has been used to show that turbulence causes water droplets in clouds to cluster together. The long-predicted effect has been confirmed by scientists in the US and Germany who found that droplets group together in clouds in ways that would not be expected if they were randomly distributed. The clustering may have an impact on rainfall, particularly in highly turbulent clouds, but the researchers say more data is needed to confirm this.

Rain falls when gaseous water vapour in clouds condenses, forming tiny water droplets about 15 µm in diameter that must grow to about 100 µm before they fall as rain. Once the droplets reach around 40 µm in diameter they start to move down through the cloud and collide and merge with other droplets, increasing in size. But, neither condensation nor gravitational collisions effectively explain how the size of droplets increases from 15–40 μm. This is known as the “size-gap problem”.

Scientists believe that turbulent flows within clouds help droplets cross the size-gap: the turbulent flows cause the mid-size droplets to cluster together, increasing the likelihood of them colliding and merging. “The prevailing theory is that naturally turbulent air motion within the cloud induces cloud particles to have some tendency to cluster,” explains Michael Larsen, a physicist at the College of Charleston in the US.

Challenging measurements

While theoretical models and laboratory experiments have confirmed this idea, most clouds are clearly much larger than anything that can be simulated on the ground, and in-cloud measurements have proved challenging.

To explore whether turbulent flows in clouds help water droplets cross the size-gap, Larsen and his colleagues, flew a 3D holographic imaging system through clouds and took thousands of images of water droplets. This enabled them to analyse the 3D distribution of droplets in the clouds for the first time.

“We used holographic images taken by a cloud probe – HOLODEC [Holographic Detector for Clouds] – mounted on the bottom of an aircraft wing,” Larsen told Physics World. “By carefully analysing the data from this instrument, we were able to statistically demonstrate that there was a greater likelihood for cloud droplets to be found 1-5 mm apart than you would expect if the cloud droplets were distributed perfectly randomly in space.”

Three key predictions

Their results are also consistent with three key predictions: droplets tend to cluster in common regions of the turbulent flow; droplets cluster more densely as the space between them decreases,; and droplets start to cluster when their diameter is roughly within the size-gap.

Wojciech Grabowski, from the National Center for Atmospheric Research, in Colorado in the US, says that while the study is not ground-breaking “it does confirm without doubt what we knew and expected”, and provides “a solid observational foundation” for the effects seen in model simulations.

Larsen says that it is hard to say what impact droplet clustering has on rainfall until we have a better understanding of the amount of clustering in different turbulent conditions. He adds, however, “It is possible that ambient drop clustering – if strong enough in highly turbulent clouds – could make the process of cloud droplet growth through drop–drop collisions more efficient than currently expected and ultimately result in more rapid cloud drop growth to rain drop sizes than current theories typically suggest.”

Grabowski says, “It is now well appreciated that turbulence does affect rain formation in clouds and that the effect strongly depends on the droplet sizes and levels of turbulence. That said, I do not think the effects of turbulence on rain formation have any significant impact on weather forecasting: there are simply more uncertain aspects of weather prediction.”

The research is described in Physical Review Letters.