On Jan. 15, 2012, Greg Asner took off from Panama City, Panama in a twin-turboprop Dornier 228 with two pilots, three technicians and some highly specialized equipment. Traveling west at 150 knots per hour, the plane rose above the skyscrapers and suburban developments of the capital city and swooped over Gatún Lake’s Barro Colorado Nature Monument, with its towering espave and ceiba trees. It flew over cleared grasslands and cattle ranches and skirted the steep, wet forests trailing beaches down to the Caribbean. It would soon head east to the wild, largely unpopulated Darien region, with its network of mangroves, rainforests and swamps.
Asner, a Stanford professor and staff ecologist at the Carnegie Institute for Science in California, was on a unique quest. For 15 years, he had been developing the equipment, the team and the mathematical algorithms necessary to measure the carbon stored in forests. Now, at the invitation of the Smithsonian Tropical Research Institute (STRI) and the Panamanian government, he would be undertaking a feat never before accomplished: mapping the carbon stocks of an entire country.
The information would not only help Panama’s government and indigenous tribes manage their forests, grasslands and mangrove stands. It also would allow the country to establish a baseline for its carbon stocks, as well as a monitoring system to track the rate at which carbon is being released into the atmosphere on account of deforestation and other land-use changes. That information could be crucial to Panama’s participation in any future effort to reduce the role of tropical-woodland destruction in climate change.
Aboard Asner’s plane, dubbed the Carnegie Airborne Observatory, two lasers from the Light Detection and Ranging (Lidar) system beamed invisible light to the ground at 500,000 pulses per second, penetrating the forest canopy and generating high-resolution three-dimensional images on a computer of the trees and plants below. Prior to the flight, Asner’s Lidar had been calibrated with field plots measuring biomass in Panama’s different ecosystems. Now, by combining that data with the 3-D images, the team calculated carbon in the land below with amazing speed. “We mapped nearly 390,000 hectares in just 17 days around the country and came up with carbon estimates for every type of high- and low-diversity forest, including primary, secondary, degraded, and logged areas,” Asner says.
Over the next five months, Asner’s team would complete its mission. It would use data from North American Aeronautics and Space Administration (Nasa) Landsat satellites to scale up sampled areas to the nation’s total 7.5 million hectares, providing a contiguous carbon map of the country. And it would do so for 10 cents per hectare, an immense savings compared to hiring laborers to work field plots at a per-hectare cost that Asner estimates at $2,000 to $7,000. A valuable new tool had been added to the older systems of estimating carbon by using field plots and satellite data.
“This is a revolutionary system,” says Ken Andrasko, director of ecosystems services for Winrock International, a U.S.-based nonprofit focused on economic and environmental issues. “Monitoring of forests had previously been based on Landsat imagery, which was pretty crude with a resolution of around 30 meters. With Asner’s system you’re down to a meter, you can see individual trees and habitats, and you’re beginning to be able to assess biomass remotely, including not only deforestation as before but also selective logging or degradation.”
For years, efforts to incentivize tropical forest conservation as part of a global greenhouse-gas-reduction regimen have languished amid continuing uncertainty over the direction of international climate policy. And yet the once-daunting technological and methodological challenges inherent in those efforts are quietly being overcome. The ironic result is that while the push to place woodland preservation in the service of climate protection has been blunted thus far by political and economic questions, arguably the most fundamental conundrum—how to reliably gauge the carbon storage and emissions impacts of forest-conservation projects—is well on the way to being resolved.
To the extent it encourages the use of forest conservation to help fight climate change, technical progress of the type made by Asner could be enormously important. Deforestation is estimated to account for over 15% of global greenhouse-gas emissions—more than the emissions of all of the world’s cars and trucks. It looms particularly large in the greenhouse-gas output of Latin American countries such as Bolivia, Brazil, Panama and Peru. In Brazil, for example, greenhouse-gas emissions from deforestation and other land-use changes are greater than those from the other main activities measured—energy (which includes transportation); farming and ranching; industrial processing; and waste treatment.
Latin countries such as these are well positioned to benefit from programs designed to provide compensation for reduction of emissions from deforestation and forest degradation, an approach known in climate-policy circles as REDD+. Whether to attract investors in such initiatives or simply to improve their own forest-management, countries must not only know how much carbon is stored in their forests’ biomass, but must also show on an ongoing basis that the woodlands are not being destroyed or degraded. Further, they must determine that conservation efforts are not causing “leakage”—displacement of deforestation from one area to another.
For initial mapping of baseline forest-carbon stocks, Lidar is “the gold standard,” says Scott Goetz, deputy director and senior scientist at the Woods Hole Research Center in Massachusetts. The system can create three-dimensional images—using latitude, longitude and canopy height—to provide the most accurate biomass measurements.
Asner says that in addition to surveying Panama’s carbon stocks, he has conducted Lidar mapping of Peru and a large portion of the Colombian Amazon. He says he has funds from the MacArthur Foundation to finish the mapping of Colombia and to conduct a complete carbon survey of Ecuador.
Lidar alone is not a panacea. For one thing, the expense of Lidar airplane flights—while significantly lower than the cost of ground observations—can be justified for initial baseline surveys of forest-carbon stocks, but perhaps not as easily for regular monitoring of subsequent changes in forest cover.
Recognizing this, Asner and others are correlating Lidar carbon-stock maps with satellite images so ongoing monitoring of forest cover can be conducted using free Landsat pictures. Unlike Lidar maps, the Landsat images are two dimensional and thus don’t reveal canopy height, though they do show vegetation.
Goetz believes the ideal both for baseline mapping and ongoing observation would be a space-based Lidar system designed to map vegetation, with the images made available to all countries. Research organizations are lobbying Nasa for such a service, according to scientists interviewed for this article. However, some experts believe that even if a Lidar satellite were in operation, it still would need to be supplemented by some degree of ground observation.
Edward Mitchard of the University of Edinburgh, the lead author of a recent study of carbon storage in the Amazon region, says that while Lidar’s three-dimensional images are useful, they do not show how many close-grained, slow-growing species are in each section of forest. In their study, published this month in the journal Global Ecology and Biogeography, Mitchard and his research team documented through ground observation that forests in the northern Amazon—the basin’s oldest geographic region, known as the Guyana shield—contain the largest amounts of carbon. Faster-growing woodlands to the southwest, in the Peruvian Amazon, hold less, he says.
“Lidar only gives you the height of trees; it doesn’t give you the wood density,” he says. “You’d still need a combination of Lidar and ground plots and good species identification of the trees.”
As a possible alternative to Lidar, Mitchard is experimenting with longer-wavelength satellite radar that can see through leaves and twigs, to calculate carbon storage in tree trunks. Meanwhile, a satellite dubbed BIOMASS, due to be launched by the European Union in 2020, will circle the Earth for five years and allow more precise measurement of carbon stocks from space. Though it won’t carry Lidar, the satellite will be equipped with a radar designed to penetrate the forest canopy and rebound off the substantial, woody parts of trees, thereby detecting their trunks and large branches.
Mitchard agrees that once baseline carbon stocks are measured, space-based radar would be the most cost-effective way to track ongoing changes in forest cover, although it can have a hard time detecting degradation from fires, logging or other activity. So-called citizen science that relies on community members armed with handheld devices can help fill that gap, he says. Besides being less expensive, community monitoring helps people understand their forests better, engendering a sense of ownership of woodland resources, he adds.
In the area of baseline carbon calculation, one important hurdle remains, Mitchard asserts: Lidar and systems such as the EU satellite can only measure aboveground biomass reliably. Says Mitchard: “The belowground and soil carbon stock is really totally unknown.”
As scientists make strides on the technology of forest-carbon measurement and monitoring, they also are devising ways to make it easier for government officials, community groups and environmental organizations to keep tabs on their countries’ forest-conservation status.
Several software packages are now available for analyzing Landsat images, for instance.
Brazil’s PRODES [Projeto de Monitoramento do Desmatamento na Amazônia Legal por Satélite] system, operated by the National Space Research Institute (INPE), has the longest track record in the region, providing data on deforestation in the Brazilian Amazon.
Asner’s research group at the Carnegie Institution for Science has developed its own forest-monitoring software, called CLASlite, offering the package for free, as well as English- and Spanish-language training in the use of it, to governments, organizations and academics. Peru’s Environment Ministry is authorized to license the software to groups in that country.
Last year, researchers led by Matthew Hansen of the University of Maryland released a system powered by Google’s Earth Engine that can process large amounts of satellite data. The system can show land-use changes anywhere in the world since 2000, and will eventually allow real-time monitoring. Most recently, the Washington-based World Resources Institute launched its free, on-line forest-monitoring platform, Global Forest Watch.
Such systems, to be sure, have limitations. While satellite-based radar and optical systems can tell if an area is forested or not, for instance, they may not distinguish between primary and regenerated forest, between forest and tree plantations or—as Mitchard noted—between healthy and degraded stands. Such variations in forest characteristics and health can affect estimates of forest-carbon storage and emissions.
Software developers have been hard at work shaping remote-monitoring systems to gauge degrees of forest degradation. Algorithms are being applied to Landsat data for the purpose, for instance. But to get an accurate picture, users of global systems such as Hansen’s or WRI’s Global Forest Watch still must gather information in the field for purposes of comparison, says Daniel Nepstad, senior scientist and executive director at the Earth Innovation Institute in San Francisco.
Indeed, natural-resource managers already have begun using a combination of monitoring tools. Brazil’s Acre state, which many consider a leader in REDD+ preparation in South America, uses the INPE’s PRODES system. But because the resolution provided by that system is not fine enough to show the very small areas typically cleared by subsistence farmers, Acre developed its own software for more detailed analysis of the satellite data, according to Mónica Julissa de los Ríos de Leal, head of the state’s environmental-services regulation department. So officials use PRODES, which is a nationwide standard, to compare changes in Acre’s forest cover with those of other states, and depend on the more detailed system for local decisions about land-use policy, she says.
Ultimately, countries need a combination of technologies, such as Lidar and satellite images, that have been “ground-truthed”—verified by people who head into the forest to compare the remote-sensing data with what they see there, experts say.
Goetz advocates continued refinement of remote-sensing technology and development of means to make it available to all countries, so data comparisons are consistent. But precision must also be balanced against cost, speed of use and the ability to employ the technology on a large scale, says Charles Ehrhart, technical director and senior partner of the Floresta Group, a private company that backs green-economy projects, including some in Acre.
Perhaps most importantly, it must be “institutionalized,” he says, meaning governments must include the technology in their budgets and policies and use it for monitoring and policing, as Brazil has done with PRODES. “You can have all the data you want, and it can be beautiful data, but unless it is being used to inform intervention on the ground—policy intervention and, in many cases, legal intervention—it means nothing,” Ehrhart says.
Whether the steady technical advances in forest-conservation monitoring will indeed inform interventions on the ground will ultimately depend on the course of climate-protection efforts. If climate negotiators succeed in producing a successor agreement to the Kyoto Protocol that includes mandatory greenhouse-gas reductions, currently flagging hopes for meaningful forest-conservation incentives—and for Redd+—might revive.
But with Redd+ already years in the making and uncertainty clouding the global cap-and-trade system many had assumed would help support it, former champions of the system seem to be losing their enthusiasm. Some stress improved forest management, rather than Redd+, as a way to reduce deforestation. And they’re heartened by events unrelated to cap-and-trade programs, such as recent no-deforestation pledges by global food companies including Anglo-Dutch Unilever, U.S.-based Kraft Foods and Switzerland’s Nestle regarding oil palm plantations and other food sources.
“I’m one of the bigger advocates of Redd and getting it into a cap-and-trade system,” says Tony Brunello, a partner and expert on energy and climate change at California Strategies, a California public affairs firm. “But I’ve been at this long enough to know that it is only one in a portfolio of things being done to reduce deforestation, one of the most relevant of which is this new commitment by major companies.”
Brunello says advances such as Lidar could be critical in helping such companies oversee where deforestation is occurring in the supply chain. He also said it would greatly help countries improve their forestry enforcement. “You need a good inventory system that tells you where forests and cropland are and what the dynamic changes in them are,” he said.
For his part, the Earth Innovation Institute’s Nepstad says he is optimistic that the U.S. state of California, as part of its cap-and-trade market, will eventually allow the purchase of carbon offsets based on forest-conservation projects in other countries. He forecasts that Acre could be ready to participate in such a program by 2016. Meanwhile, meetings this year, including December’s international climate summit in Lima, Peru, will be crucial. Says Nepstad: “We’ll know this year if Redd survives.”
- Steven Ambrus and Barbara Fraser