Last September, I travelled to Peru to meet a fascinating scientist who is mapping the Amazon by plane. The piece was published in Wired UK earlier this year, and I’m reprinting it here now. This was one of the most enjoyably things I got to write last year. I hope you enjoy it too
A small, twin-propeller plane flies over the Amazon rainforest in eastern Peru. The scale of the vegetation is extraordinary. The tree canopy stretches as far as the eye can see — an endless array of broccoli florets bounded only by haze and horizon. Greg Asner, 43, has seen the rainforest from this vantage point many times before, but he still stares out of the window in rapt fascination.
This patch of forest in the Tambopata National Reserve is rich with life, even by the Amazon’s standards. A 50-hectare patch of forest — the size of as many rugby pitches — contains more plant species than the whole of North America. “We might as well be exploring Mars,” says Asner. “These are areas where no human has ever been. There’s no access.”
Access isn’t a problem for Asner. Behind him are three state-of-the-art sensors of his own devising which, as the plane flies along, take the forest’s measure. “We’re trying to do something really new,” He says. “This world is changing and it requires science that isn’t incremental.” Using the technology he’s developed, Asner is mapping the shape and size of the trees, down to individual branches, from two kilometres above. He can measure the carbon stored in trunks, leaves and soil. He can even identify individual plant species based on the chemicals they contain. With wings and lasers, Asner is conducting one of the most ambitious ecology studies ever staged. He accumulates more data in a single hour than most ecologists glean in a lifetime. With this data, he aims to influence governments, steer the course of climate-change treaties and save the forests over which he soars.
Asner’s high-flying science has roots on the ground. In 1994, he was working in Hawaii for Nature Conservancy, an environmental nonprofit organisation. He was frustrated. His seemingly simple task of eliminating invasive plants was thwarted by an equally simple problem — he could not find them. “We were stumbling around in the dark. That’s how I started getting into this, thinking how we could get maps of this stuff.” Satellite data was too coarse and aerial photos uninformative. Eventually, during a PhD position at the University of Colorado, Asner found his answer — airborne sensors. Fifteen years and much head-scratching later, his team has developed the Airborne Taxonomic Mapping System (AToMS), a suite of three plane-mounted sensors that he describes as “probably the most advanced Earth-mapping system in the world”.
Asner’s smooth, all-American features make him look young for his age. He spends as much time flying over forests as he does in his lab at the Carnegie Institution for Science in Washington, DC. Although he has travelled widely, the Amazon holds a special fascination. It is where he cut his scientific teeth, where he returns to again and again, and where he flies today. His plane — the Carnegie Airborne Observatory (CAO) — is a Dornier 228. Passengers stoop inside and squeeze through the narrow aisle. The sensors at the back get more room. A grey box known as Light Detection And Ranging (LiDAR) is the oldest. Its two beams scissor back and forth across the canopy at 400,000 pulses per second, and the echoes reveal the geometry of every tree, branch, and sometimes leaf. “It’s not like dropping a cloth on to the canopy,” says technician James Jacobson, who controls the device. “You get deep three-dimensional models of the entire forest.”
The other two sensors — Visible and Near-Infrared (VNIR), another grey box, and Visible Short-Wave Infra-Red (VSWIR), a golden cylinder — are today on their virgin expedition. They emit no beams; instead, they scan the sunlight that reflects off the trees cross a range of wavelengths, from the visible to the infrared. When sunlight hits the canopy, it incites the molecules within the leaves to bend, stretch, vibrate and rotate. This atomic tango scatters and absorbs the various wavelengths of light in different ways, creating a chemical signature that the sensors can detect from afar. VNIR is a zoom lens with better spatial resolution; VSWIR is a wide-angle lens that scans the forests’ reflections at 440 different wavelengths and picks up more chemical detail. Together, they can measure the concentrations of the nitrogen and phosphorus that the trees need to grow, the cellulose and lignins that keep them sturdy, and the tannins and phenols that defend them from enemies. While LiDAR maps the Amazon’s shape, VNIR and VSWIR watch its life.
An opening in the fuselage gives the sensors a view of the world below. Travelling at more than 100kph, they sweep a strip of land 1.5 kilometres wide, at a spacial resolution of 0.5 metres to two metres. In one pass, they can scan a single tree 900 times. None of the sensors is bigger than a bedside table but between them they cost $9.8 million (£6.1 million) and took three years to build. They are so advanced that Nasa has shown interest in mimicking the designs and film-maker James Cameron paid a special visit to Asner’s lab to see them. “He’s really geeky and nerdy,” says Asner. “He loves to create technology and he has asked to be a spokesperson for the programme.” Not everyone has been so immediately supportive. “People thought I was crazy for years. They said it’s too cloudy to map the rainforest. It’s not — you just need to know how to work the system.”
That know-how is at work today. Robin Martin, Asner’s wife and second-in-command, is monitoring the plane via satellite from their base in Cusco, Peru. She steers it around cloud banks that would block the sunlight on which VNIR and VSWIR depend. Pilots Colin Wright and Brian Czirjak keep the plane on a straight and slow course, so the sensors can grab as much data as possible. At speeds of just 55 knots, the CAO almost seems to hover. Its nose tilts dramatically upwards and the pilots have a hard time seeing out of the cockpit. Jacobson and Asner play back-seat drivers, feeding them with instructions, while controlling the sensors and liaising with Martin on the ground.
That’s today’s configuration. Tomorrow, different team members will rotate into the various roles. With such a synchronised partnership, the team easily adapts around inclement weather conditions.
To Asner’s trained eyes, the rainforest is a crowd of familiar friends. From above, he identifies pollen-dusted lollipops as yellow lapacho trees, blooming with golden flowers. Clumps of matchsticks are groves of Moriche palms, thriving in swampland. Brown bundles are giant legumes that have dropped their leaves for the dry season. Identifying trees by sight from kilometres away is impressive enough, but Asner’s sensors far outstrip their creator. They are metal taxonomists, built to classify plants from the sky.
Asner tested this approach in 2004, having returned to Hawaii to tackle the invasive species that had so frustrated him in his early career. Using an older generation of sensors, he mapped strawberry guava trees that were stealing the sunshine from native canopy plants, and an alien ginger that took nutrients away from ground-based locals. Asner has since upgraded his tech. VSWIR is his crowning achievement. It is housed inside a vacuum chamber, chilled to -137°C and isolated from vibrations by a custom-made frame. “With the old sensor, we were getting 80 per cent accuracy. VSWIR will take us up a leap.”
This is a far cry from traditional ecology, which still uses the methods of 20 years ago. “You literally tape off an area and work through it in a grid, counting species and mapping them on paper,” Asner says. Fieldworkers record the size of the trees with tape measures and clinometers, and professional climbers bring down leaves, flowers and fruit for dissection. “Field plots are looking at a microscopic fraction of the Amazon, a few hundred hectares at most. Our plane covers several hectares every second.” To illustrate the point, Asner opens his laptop, zooms into a tiny region of Peru and points to some yellow dots. “These are almost all the field plots in the Amazon.” He zooms back out to a map of the whole country, covered in rectangles and transepts. “And that’s what we’ve covered.”
Fieldwork is still vital — VSWIR’s power is useless without a way of interpreting its readings. That is where Martin comes in. With blue eyes and tremendous energy, she is every bit the “field jock” that he is, and the two are a well-matched team. Since 2005, they have hacked their way into rainforests around the world, collecting samples that help to calibrate VSWIR’s data. Armed with shotguns, slingshots and cranes, they have collected around 10,000 leaf samples from more than 7,000 species of plant. They measure the levels of 21 chemicals within each sample, revealing the chemical fingerprints that make each unique. “Robin and I measured every leaf personally in the woods,” says Asner. In a test run, the duo used these fingerprints to classify around 600 species with 90 per cent accuracy.
The knowledge they have amassed is not just for academic interest. “A lot of money is being put into biodiversity conservation, and it’s worth knowing what we have, so we can make better decisions about what to conserve,” Asner says. “It’s hard to fight against the value of gold, oil or timber. We have to focus our decisions on the areas that are most important as storehouses for the future.”
His words are tinged with sadness and urgency, for his flights remind him daily that he is working against a rapidly draining hourglass. Within barely an hour, the view from the CAO’s window changes from pristine forest to the scars of human activity. Great clearings contain piles of logs slashed down by cattle farmers. Incongruously grey patches are freshly burned areas, their vibrant Technicolor converted into the monochrome of ash. “I’ve lost count of how many special places I have seen fall apart in just 15 years,” he says. “Piece by piece, it’s being taken apart.”
Illegal goldmines pose the biggest threat. This corner of the Amazon is rich with gold. Record prices and a new transoceanic highway between Brazil and Peru have prompted a gold rush, bringing hopeful immigrants by their tens of thousands. It took miners a few years to turn 155 square kilometers of forest into Guacamayo mine, one of the largest around. From the air, it looks like an enormous pus-filled wound that seeps into the forest, feeding the rivers with silt and run-offs that stain their grey-green waters a sickly yellow.
Guacamayo is Spanish for macaw, but no birds live in the permanent scars created by the mines. “It’s not just cutting the forest down. They dig so deep that it becomes uninhabitable,” Asner says. His brow furrows. “You’re talking to someone who has been blistered by 15 years of working in the tropics, but even I can’t believe it sometimes,” he says. “I just think it’s a shame to wipe away four billion years of evolution, and this is nothing compared to what I’ve seen in Brazil or Indonesia or Malaysia.” But with untouched forests a quick plane ride away from despoiled ones, Asner sees cause for optimism as well as despair. “We’re still ahead of the curve here. That’s why it’s worth working in this region. There’s incredible threat but a lot of hope too. In the air, you can see it all in an hour.”
In January 2011, the CAO played host to Colombia’s president Juan Manuel Santos, and flew with two fighter jets and a Black Hawk helicopter. It was the highlight of a very strange month. Asner’s team had been flying over stretches of the Colombian Amazon controlled by Farc rebels armed with high-calibre weapons. That meant military briefings and, with the exception of Santos’s visit, night-flying. “We’d go completely lights-out,” says Asner. “We’d be up high, they’d be below doing their operations and we could see when ordnance was being dropped. But the lasers saw right through, no problem.” These stealth missions were, ironically, all about transparency. Asner was working with Colombia’s Ministry of Environment to measure the carbon locked in the country’s forests. Damage to the world’s forests releases around a fifth of all greenhouse-gas emissions, more than the entire transport sector. The world’s governments are negotiating a treaty called Reducing Emissions from Deforestation and Forest Degradation (REDD), which will allow developed countries to offset their carbon emissions by paying developing countries to protect their forests.
Initially the treaty was held back by scepticism about whether countries could accurately tally the carbon stored in their forests. “It’s like saying you should invest in my company’s stocks, but we won’t report to you,” Asner says. Since then, he and other ecologists have convinced decision-makers that airborne sensors can do the job cheaply and accurately. In 2009, he used LiDAR to map carbon stocks in more than 4.3 million hectares of Peruvian Amazon at a cost of just eight US cents per hectare. He did the same with 14 million hectares of Colombia’s forests, reducing the cost even further. The team converts the raw data into maps, colour-coded by species or chemicals. Asner’s laptop contains dozens of movies showing virtual planes soaring over kaleidoscopic forests. “The fly-throughs are important for reaching people, from presidents to my son,” says Asner.
Meanwhile, the local ministries get all the raw data and Asner trains them, and local NGOs, to use it. “We work hard to accommodate requests from ministries and prioritise areas that are important to them,” says Martin. “We don’t want to be the imperialist outsiders coming in and taking all their data.” Soon, countries will be able to do it themselves. With Asner’s help, Ecuador already has its own version of the CAO. Peru, Colombia and Brazil are not far behind. “The endgame is to pass this on to other people,” says Asner. “I’m picturing myself in 25 years, in my home in Hawaii, reading about all these airborne observatories finding new species, helping decision-makers and doing science that I can’t imagine.” With such advanced technology, the temptation to cash in is always present. Since 2002, venture capitalists have been making increasingly spectacular efforts to convince Asner to turn his technology into commercial ventures. He has turned them all down, although he claims he is not altruistic. His actions say otherwise.
There is no better example of this than CLASlite, a piece of software designed to spot deforestation using satellite images. It is based on an algorithm called Carnegie Landsat Analysis System (CLAS), which Asner started refining in 1999 to map the effects of logging in Brazilian rainforests. CLAS recognises patterns in satellite images and marks out areas of logging and other disturbances. It turns uniform green carpets into detailed colour maps. When Asner turned CLAS into the desktop-friendly CLASlite, he offered training to anyone. Demand was overwhelming. To date, his team have trained 400 active users across 160 organisations in nine Amazonian countries. “I’ve had thousands of emails,” he says. “NGOs want it. The minister of environment in Nepal wants it. I had an email from a 70-year-old guy who has a few hectares of forest in Mexico and really wants to know if it’s doing well. This is great, but we’re already sleepless.”
There was no way Asner’s team of 25 could cope. “Then in 2008, someone from Google.org came to me and asked if there was anything they could do to scale it up,” he says. Now Asner is about to launch CLASlite Online, a web-based version intuitive enough for anyone to use without training. Google will host it on its cloud and run it off the Google Earth Engine, incorporating satellite images from Nasa and other worldwide sources. When CLASlite Online launches, hopefully in a year’s time, users will be able to monitor deforestation on their home computers. If a tree falls in the forest, anyone with an internet connection will be able to hear it.
On a cloudless day, the CAO takes off again, this time with Martin on board. Asner, looking tired from several sleepless nights, guides the flight from the ground, watching as a small, plane-shaped cursor rises and falls on his laptop screen. His team is worn out. After two months in Peru, many are itching to go home. Their efforts have been fruitful but their schedule punishing. They start each day at 5am and finish late at night. They work weekends. They have commandeered the top floor of a local hotel, turning it into a makeshift lab where they rigorously check their accumulated data after every flight. If the weather has been clear, there can be a terabyte to get through.
Martin’s presence is a consolation to Asner (“I don’t see any other way to have a relationship,” he says). Still, he is looking forward to getting home to Hawaii, diving and spending time with his 12-year-old son from his previous marriage. But he is also planning his next expeditions. “I want to go around the planet in the CAO,” he says. Papua New Guinea is on 2012’s itinerary, and his wish list includes South-East Asia, the Pacific Islands, the Congo, northern Australia and a return trip to Madagascar with VSWIR in tow. He exudes urgency. The world’s forests are changing rapidly, in line with rising global temperatures, even before scientists get to study them properly. “If we don’t get a baseline on what we have now, then all we’re doing is observing ecosystems in a state of non-linear transition. How do we interpret that? We need to get in there and take measurements. You can’t do that from the ground.”
But before he embarks on his next international campaign, Asner must finish in Peru. He will exchange the mountainous cool of Cusco to the close heat of Iquitos, in the north-western half of the country. A week in this hot soup will be taxing for both the technical team and the equipment, but Asner is boyishly excited. “We’re going to explore this massive swamp area,” he says, pointing to a map. “We’re also mapping the Napo River, going from Iquitos to the Ecuador border. I’ve wanted to do this for years — we have no idea what’s there.”