Every cloud has an invisible halo
Unseen particles may confuse climate models.
Clouds are bigger than they look, according to new measurements by atmospheric scientists in Israel and the United States. They say that clouds are surrounded by a 'twilight zone' of diffuse particles, invisible to the naked eye, extending for tens of kilometres around the cloud's visible portion.
These vast, sparse haloes of droplets may have been overlooked in atmospheric studies, the researchers say. And they think that this could have skewed attempts to understand how clouds influence climate.
Clouds are one of the biggest sources of uncertainty in efforts to measure and predict global warming. They have two opposite effects: increasing warming by absorbing heat radiated from the planet's surface (which is why cloudy nights are warmer), while offsetting this by reflecting sunlight back into space from cloud tops.
Most atmospheric scientists now think that clouds have an overall global cooling effect. Measurements of warming trends therefore have to take into account whether the skies are cloudy or not, and model forecasts of future warming may hinge on whether they predict more or less cloudiness.
Clouds are formed when floating solid particles called aerosols — dust, for example — act as 'seeds' on which water droplets grow. Aerosols reflect light, and do so more strongly as they grow by accumulating water. The large droplets in clouds reflect most visible light, which is what makes clouds look white and opaque.
Koren and his colleagues first demonstrated that it is relatively easy to see from digital photographs that clouds are surrounded by an invisible haze, made up of these water-coated, or humidified, aerosols. If the parts of the photo containing visible white stuff are masked out, the surrounding haze comes into view.
This haze extends far further than anyone has ever imagined. "People may have seen these extended haloes anecdotally," says Koren's colleague Lorraine Remer of the NASA Goddard Space Flight Center in Greenbelt, Maryland. "But thanks to a new generation of instruments, the satellite observations have got much better, and we can look on larger scales, with more sensitivity and at finer resolution."
Satellite images of clouds over the Atlantic Ocean show that the sky's reflectance — a measure of how much humidified aerosol it contains — falls very gradually with increasing distance from the edge of a cloud, and is still declining at least 20-30 kilometres away, Koren's team says.
Into the twilight zone
To study these twilight zones further, the researchers studied several years' worth of images collected by a global network of ground-based lightmeters called AERONET, usually used to ðmonitor the brightness of the Sun.
Sudden dips in the light detected by these instruments are automatically logged as indicating the passage of a cloud. Koren and colleagues discovered that it can take well over an hour for light levels to recover fully after a cloud has passed, indicating that their haloes are very broad.
Not all clouds will have a big twilight zone, the researchers say. For example, the halo might be tightly reined in around the sharp-edged white cumulus clouds that form when moist, warm air rises and cools. But they estimate that for typical global cloud coverage, the halo could encompass as much as two-thirds of the sky usually classed as cloud-free.
Remer says that some climate models might already include these extended cloud haloes — they should 'grow' them automatically if they do a good job of capturing the humidity variations of the air. But other, simpler, models might neglect the effect.
As a result, Remer suspects that the overall cooling effect of aerosols may have been underestimated. But she admits that it is too early to say whether that is really the case, or how significant an impact it might have on climate predictions.
"Right now there is a discrepancy between what global models predict for aerosol effects and what satellites measure," she says. "This might be part of the reason for that."
- Koren I., et al. Geophys. Res. Lett., 34. L08805 (2007).
If you need help or have a question please use the links below to help resolve your problem.