It's no secret that global biodiversity is not evenly spaced over the planet’s surface. Much of it is concentrated in hotspots, many of which are in places that are relatively remote, making it difficult for scientists to gather detailed information on species’ habitats and distribution.

So researchers are taking advantage of remote sensing – using images and data recorded by satellites – to infer the extent of certain habitats and of the species within them.

Now, scientists from the University at Buffalo and Yale University have shown that there is one especially effective way of calculating the whereabouts of species, including those that may be threatened or endangered: look to the clouds.

Photos: Know Your Cloud Types

On one level, of course, it may seem obvious: tropical forests, for example, are going to be in areas of heavy rainfall and thus cloud cover. But smaller-scale differences in cloud cover can cause variations in factors such as leaf wetness, surface temperature, precipitation and sunlight from one location to the next; and the new research, published today in the journal PLoS Biology, drills down in unprecedented detail, showing how sharp delineations in cloud cover correlate to abrupt changes in the ecosystems below.

The study used 15 years of data from NASA’s Terra and Aqua satellites, which orbit and study the Earth, to build a database containing two images per day of cloud cover for nearly every square kilometer of the planet from 2000 to 2014.

“When we visualized the data, it was remarkable how clearly you could see many different biomes on Earth based on the frequency and timing of cloudy days over the past 15 years,” said lead scientist Adam Wilson, who conducted the majority of the research at Yale University and is now an assistant professor of geography in the University at Buffalo College of Arts and Sciences.

Two-Thirds of the Earth Is Covered by Clouds

“As you cross from one ecosystem into another, those transitions show up very clearly, and the exciting thing is that these data allow you to directly observe those patterns at 1-kilometer resolution.”

This approach proved so precise, in fact, that the researchers used it to map the size and location of habitats for two species - the montane woodcreeper (a South American bird) and king protea (a South African shrub) – in far greater detail than before.

They note that one reason for the great improvement in mapping the distribution of the woodcreeper, in particular, was that to this point scientists have estimated the extent of its habitat by examining precipitation data. Not only is that a more crude method than the fine-resolution cloud cover analysis they were able to perform, but it further loses accuracy in places – such as in the Andes, where the woodcreeper lives – where weather station networks are less dense.

Dwarf Dragons Found in Cloud Forests of the Andes

This new approach will likely be of particular benefit in less developed countries – which often are where the information is most needed for management purposes, because of high biodiversity and increasing human encroachment.

“Understanding the spatial patterns of biodiversity is critical if we want to make informed decisions about how to protect species and manage biodiversity and its many functions into the future,” said study co-author Walter Jetz, associate professor of ecology and evolutionary biology at Yale University. “But for the regions that harbor most biodiversity, there’s a real lack of data on the ground.”

“We now have decades of satellite observations that we can pull together to characterize the global environment,” concluded Wilson.