It is as mysterious as they come. A huge, 2-kilometre-wide circle spotted during a routine flight over an Antarctic ice shelf in 2014 has had scientists scratching their heads ever since. What could have caused it?

Initially they suspected that a large meteorite had slammed into Antarctica in 2004 and made the ice crater. But now it appears the real answer is blowing in the wind.

When scientists visited the circle on foot for the first time in January 2016, they found a 3-metre-deep depression with raised edges and, in the centre, three moulins – vertical well-like shafts in the ice — draining two meltwater streams.

These features point to the crater having once been a meltwater lake, which subsequently drained into the ice shelf.

It seems strong, downslope winds, which originate from the ice sheet’s interior, are what kicked off the process that led to the crater-like formation.

Sheets to the wind

These winds, which most often blow in a single direction over the ice sheet and shelves at speeds in excess of 35 kilometres per hour, erode the surface snow and expose the blue ice underneath (see graphic, below).

This increases the absorption of sunlight and, because the wind also raises the surface temperature, the ice shelf melts more than usual where these winds are strongest. In Antarctica, the surface is nearly always colder than the atmosphere above but the vertical mixing by the winds brings warmer air in contact with the ice.

When the researchers drilled into the ice and ran radar tests further west of the crater site, they found multiple lakes several metres under the surface. This further support the idea that the surface melting had taken place and formed new lakes in the region.

What this means is that a large meltwater lake could have formed at the crater site, and eventually collapsed and drained away, leaving the telltale circle. “The build-up of pressure on the lake, which will have become so big and full of water, caused it to drain or collapse under the surface,” says co-lead author, , at Utrecht University, in the Netherlands.

The remnants of the disappearing lake gave the appearance of a circle on the surface. Then the winds further eroded snow making the circle even more visible.

“We have definitely proved that this is not a meteor site but that this crater is the remnant of a surface, or subsurface, meltwater lake,” says Lenaerts.

Much more melt

The researchers discovered similar “hot spots”, or sites of surface melting elsewhere around East Antarctica. Previously, scientists had thought that this degree of surface melting, which leads to instability of the ice sheet, only existed on Greenland or West Antarctica.

“We’ve shown that the East Antarctic ice shelves are prone to local surface melting, which is important to take into account when looking at the future response of Antarctica to climate change,” says the other lead author Stef Lhermitte, at Delft University of Technology in the Netherlands.

It’s unlikely that the instability in East Antarctica is going to lead to the type of disintegration seen in the Antarctic Peninsula in the near future, says of the National Snow and Ice Data Center in Boulder, Colorado.

“These researchers have really dived in and shown us how winds can lead to melting and instability,” Scambos says. “But the East Antarctic ice sheet is unlikely to break up anytime soon.”

Read more: Huge circle in Antarctic ice hints at meteorite impact; Mysterious fairy circles now discovered in Australia’s desert; Neanderthals built mystery underground circles 175,000 years ago

Last September’s lunar eclipse was disappointingly dim. Now, it seems a Chilean volcano sucked some of the sparkle out of the spectacle.

During a lunar eclipse, the moon takes on a ruddy glow despite lying in Earth’s shadow. That’s because our atmosphere acts like a lens, refracting sunlight on to the moon.

The September 2015 eclipse, however, was widely felt to be dimmer than usual. Trained amateur astronomers from seven countries sent measurements of its colour to Richard Keen of the University of Colorado in Boulder, who crunched the data and concluded that the eclipse was 33 per cent less bright than expected.

In a poster presented this week at the NOAA Earth System Research Laboratory in Boulder, Keen blamed this gloominess on the Calbuco volcano that erupted in April 2015. The eruption, last year’s largest, spewed ashes, dust and sulphuric acid droplets into the atmosphere.

Volcanic shadow

The volcanic pollution spread to the northern hemisphere, where it lingered for months, absorbing sunlight and dimming the eclipsed moon. “It is like a single volcano on Earth has the ability to cast a shadow on the moon,” Keen says.

The rays that illuminate the moon during eclipses mostly pass through the stratosphere, which stretches between about 10 and 50 kilometres above Earth’s surface. That means they are especially prone to being filtered out by volcanic particles, as these tend to settle higher in the atmosphere. Particles from pollution, fires or deserts rarely reach higher than the troposphere, the next layer down.

Keen has studied more than 30 lunar eclipses since 1960, and uses their brightness to gauge the dustiness of the stratosphere. Last September’s eclipse proved the first time Keen had seen changes to an eclipse from such a small volcano. “The ability of this eclipse to identify such a relatively small eruption gives me confidence in using this technique,” he says.

“We know that the moon’s surface temperature changes during an eclipse, and it’s really interesting to think that a single volcano on Earth can impact the temperature on the moon, even for a small amount of time,” says of NASA’s Lunar Exploration Analysis Group. “This goes to show that volcanoes have a global effect on the atmosphere and what goes into the atmosphere doesn’t just disappear straight away.”

Climate hotspots

Regional differences

Want to save a species? Look to the skies. Patterns in the clouds can be used to decipher the complex patterns of ecosystems on Earth, predicting where species live more accurately than the methods currently used by conservationists.

of the University at Buffalo in New York and his team have developed a technique that uses clouds to map ecosystems in fine resolution. The tool could prove helpful when deciding how best to protect a species or ecosystem, providing a cheaper and easier way to gather habitat data about hard-to-reach locations.

żěè¶ĚĘÓƵs have traditionally predicted all the habitats contained within a region by taking a sample of ground-based measurements, and using these to make a statistical estimate across the whole area. “But imagine how the conditions can vary over mountains and valleys,” says Wilson. “These statistical models aren’t able to capture the fine-grained variability of the environment that can determine where a particular species can survive.”

But clouds are different. Because they affect the amount of sunlight and rainfall that reaches the ground, as well as the temperature, clouds help determine what plants can photosynthesise in an area and the animals that can thus survive there.

“Sunlight drives almost every aspect of ecology,” says Wilson. “So when you put something in between the sun and plants, that is going to have implications on the amount of energy they are receiving, soil moisture, leaf wetness, and humidity – almost everything that is important.”

Predicting species

To map ecosystems using clouds, the researchers took data from NASA’s Terra and Aqua satellites that orbit the Earth. From this information, they built a database of two images a day of cloud cover for nearly every square kilometre of the planet from 2000 to 2014.

When they analysed the frequency and timing of cloudy days over those 15 years, they found a pattern that clearly correlated with the different biomes lying below.

To test this correlation, the researchers then picked two species with habitats that are well-known, and used the data to guess where they lived. They did this for the montane woodcreeper, a South American songbird, and the king protea, a South African shrub.

The model they had built from the data predicted the distribution of these species better than conventional methods, honing in on a range that was 43 per cent smaller for the woodcreeper, and an area 18 per cent smaller for the shrub.

“This is potentially huge – it’s the first time you have global cloud information at the appropriate grain size of 1 kilometre and from actual observations rather than being interpolated data from weather stations,” says of NASA’s Ecological Forecasting programme. “It could be a great contribution to the field.”

Journal reference: PLoS Biology, DOI: