Little studied but with potentially profound effects on the future of the world's climate, clouds are, according to one expert, the "800-pound gorillas" of climate change.
And now a new research centre in Abu Dhabi will help scientists around the world better understand the impacts of climate change.
The Centre for Prototype Climate Modelling, at New York University Abu Dhabi (NYUAD), will study clouds that form over the tropics.
That study, according to Prof Andrew Majda, the new centre's principal investigator, is badly needed.
Efforts to predict climate change largely revolve around ever more sophisticated computer models that trying to predict the impact of rising levels of atmospheric CO on the factors important for maintaining life - not just temperatures, but sea levels, rainfall, and ocean currents.
These models attempt to boil down the relationships between the various components into a series of equations. That way, they simulate how rising CO might affect the climate over the course of decades or even a century.
But the climate is a hugely complex system, and current models still omit many important variables.
One is soot, or black carbon. The result of incomplete combustion of fuel in diesel engines, cooking stoves or even forest fires, black carbon is a major contributor to global warming. But because these discoveries are relatively recent, most climate models have yet to take its warming effects into account.
Even the effect of something as apparently straightforward as clouds is poorly modelled.
"All computer models get [the effect of clouds] very badly [wrong]," said Prof Majda, who is also a professor at the Courant Institute of Mathematical Sciences at the university's main branch in the US.
"Our centre has the opportunity to make a huge change in the way people do operational modelling. If this succeeds, it will be a very high-profile centre in understanding extreme weather and short-term climate change."
Clouds are difficult to factor into a climate model, explains Olivier Pauluis, associate professor of maths at the Centre for Atmosphere Ocean Science at New York University, because they are small - from a few hundred metres to 10 kilometres.
That kind of fine detail is largely ignored by current models, which tend consider only atmospheric motions of the order of hundreds of kilometres.
"This means climate models must use physical parameterisation, a set of simplified mathematical equations, to account for the combined effects of clouds on various weather systems," says Prof Paulius.
"To make things more difficult, clouds have important effects on the amount of energy received and emitted by the Earth.
"They reflect light and infrared radiation, which makes them particularly important for predicting future climate change."
Particularly significant, he believes, may be the clouds that form in the tropics.
The impact of convection in those regions is felt around the globe. It is an important component of a global planetary circulation known as the Hadley cell: air rises near the equator, moves toward the poles at high altitude and descends in the subtropical regions.
This convection has a big impact on the weather not only in the subtropics, but in mid-latitudes, too.
The air that descends is very dry - and that is a large part of the reason why subtropical areas, such as Abu Dhabi, are so arid.
It is also responsible for maintaining the jet stream - the very strong winds about 12km above the earth's surface at latitudes between 30 and 50 degrees that drive the weather systems in those latitudes.
What happens over the tropics also has an impact on the formation of storms in other latitudes.
So a better understanding of tropical clouds can help predict the likelihood of extreme events elsewhere.
For arid Abu Dhabi, the rain storms may not be a serious concern. But the amount of rainfall is.
"For a region like Abu Dhabi and the subtropics, the issue of what is going to happen to water supply with climate change is very important," says Prof Paulius.
Current models suggest areas like Abu Dhabi, and others already suffering water stress, are likely to get dryer still.
The Abu Dhabi centre is due to start work soon. For Prof Majda, it will a chance to test some complex mathematics he has developed to describe tropical clouds. Besides their small size, they have been difficult for computer climate models because of how they behave. There is an aspect of randomness as to how they form, he said, calling them "chaotic turbulent dynamic systems". Yet, Prof Majda has gleaned some regular patterns about them.
"They produce coherent patterns where they change air circulation," he said. "We have the first theories that agree with observations."
He is using partial differential equations - equations with an element of randomness to them - to describe the tropical clouds' behaviour. Similar maths is used to calculate risk in insurance and finance. Potentially, it could lead to more reliable climate models.
"Our centre has the opportunity to make a huge change in the way people do operational modelling.
"If this succeeds, it will be a very high-profile centre in understanding extreme weather and short-term climate change."