Nature Conservation Foundation

Author
Madhu Mysore, scientist, Nature Conservation Foundation
Organization
Nature Conservation Foundation
Tools Used
Google Earth Engine, Google My Maps, Google Maps

The challenge and the organization

The Nature Conservation Foundation conducts research to help conserve India’s wildlife, and designs conservation strategies in collaboration with local communities and governments. Its work touches on the survival needs of endangered species such as snow leopards and elephants, as well as other forms of wildlife such as coral. It also helps conserve the habitats in which they live.

The Foundation must make the most of its limited resources, so it strives to use the latest affordable technologies for mapping and visualizing data. That’s why it chose Google Earth Engine to help power its research, says MD Madhusudan, a scientist with the Foundation.

“From understanding the extent of fire damage in Bandipur National Park, a tiger reserve in the south Indian state of Karnataka, to tracking migrations of the pied cuckoo, the so-called rainbird and beyond, Google Earth Engine has been the ideal platform for us,” he says. “That it is a vast data archive with stunning analytical capabilities is very much at the heart of what we are able to do with it.”

How they did it

Among the foundation’s projects aided by Google Earth Engine are two at Bandipur National Park. The first measured the severity of damage caused by fires in the summer of 2017. Every summer, swaths of the park burn in fires caused by people. Sometimes, the burns are so severe that the wildlife are in danger, and the forest itself can take years to recover. “Knowing when and where the fires occur and the extent of their damage is very important for planning how to prevent them,” Madhusudan says.

To assess the damage, he started with satellite images of the park in Google Earth Engine taken before the fires, and compared them with images taken after. The post-fire images were thick with cloud cover that he had to remove. To do so, he compared dozens of images on which he found individual pixels that were clear on some days, and covered by cloud on others. Using the computational power of Google Earth Engine, he was able to stitch the cloud-free images together to build a highly accurate picture of the fire damage throughout the park.

In this way, Madhusudan was able to estimate burn extent and intensity within days of the fire occurrences. “Without Google Earth Engine,” he says, “it would have taken months and a substantial amount of money to buy the images. The delays would have made it harder to get timely and reliable estimates of fire extent and severity.”

Madhusudan also leveraged the sizeable archive of time-series data in the Google Earth Engine catalog to assess the impact of a conservation effort to reduce forest loss in Bandipur. In 2003, an organisation called Namma Sangha began an experiment to reduce human-caused deforestation at the park. They provided people living nearby with affordable gas for cooking, so they did not have to cut down trees in the park for domestic cooking fuel. NCF was keen to know if Namma Sangha’s efforts had paid off.

Madhusudan had already plotted the loss of forest between 1973 and 2001 and found that most of it occurred near park borders, where people harvested firewood and grazed cattle. Now, using Google Earth Engine’s vast time-series archive of satellite imagery, he was able to pick the “greenest” pixel across a large stack of satellite images each year, and to check if their greenness had steadily increased. He found that starting a year after the cooking gas was provided, the levels of vegetation along the park boundary had been slowly recovering along the fringes, whereas there wasn’t a similar change noticeable in the park interiors. This suggested that Namma Sangha’s efforts were indeed helping.

Madhusudan has also used Google Earth Engine for many other projects, including tracking the migration of the pied cuckoo (Clamator jacobinus), a bird which, in mythology, is said to drink straight from the clouds. Using publicly available bird observation datasets, he traced the bird’s migration patterns and found they followed the progress of the seasonal Indian monsoon. In another instance, he has also used Google My Maps to help a distributed team track each other’s effort in the building of a bird atlas for the city of Mysore, which was India’s first citizen-led bird-atlas project. And he is experimenting using a combination of Google Earth Engine, a Google News feed and Google Maps to track the distribution of India’s endangered Asian elephant population, especially outside protected areas.

Impact

“The key impact of these projects,” says Madhusudan, “is the ability to bring compelling evidence in an easily accessible form into management practice and policy making. Government officials and conservation groups are also able to assess the impact of efforts to protect forests and save endangered species.” Based on Nature Conservation Foundation’s analysis from Bandipur National Park, for example, there is now persuasive evidence that a trajectory of chronic forest decline was actually being reversed at scale with a creative model of making low-cost cooking gas more accessible to firewood dependent villagers. “Knowing that,” he says, “increases our confidence that this is a measure that can help conservation.”

“Google Earth Engine combines a vast archive of data with unimaginable computing power,” he says. “To have an archive that contains every single LandSat image ever taken is, by itself, impressive. But, to realise that you also have access to the staggering computing power needed to process such an archive, and that you can do that just within a browser window, is astonishing. Of course, that can help us answer many worthy questions that need answers. But, what is truly game-changing and still boggles my mind is that we can now ask questions that were not even possible to ask before.”