Going back to his childhood, Stephane Germain, the founder of Montreal-based GHGSat, dreamed of figuring out how to pursue a career that combined space, technology and the profit motive. After completing his graduate work in engineering physics, he found himself drawn increasingly to both the world of commercial satellites and the scientific challenge of using specialized orbiting sensors to measure atmospheric greenhouse gas concentrations.
But when the province of Quebec prepared to join California’s cap-and-trade system, “the light bulb went on,” Germain says. Cap-and-trade forces companies to buy credits for the right to pollute, which incentivises them to reduce emissions. Pondering the logic of such systems, he realized the province’s industrial emitters suddenly would be forced to manage their emission risk and therefore had a financial incentive to seek out better information about the carbon leaving their facilities. “I knew there was market demand,” he explains.
Germain launched his satellite startup in 2011 with seed funding from Sustainable Development Technology Canada. Seven years later, GHGSat, one of only five Canadian companies flying satellites, is launching its second device and has completed two private placements, the first involving an equity infusion by U.S.-French oil services giant Schlumberger. GHGSat’s 15-kilogram satellite, which is the size of a microwave oven and known as Claire, was sent into orbit by the Indian Space Research Organization in June 2016, circling the earth every 98 minutes at a fixed altitude of 500 kilometres. It is fitted out with tools that measure infrared light and telescopic lenses.
The goal, still in the demonstration phase, is to use the on-board sensors to identify and then quantify industrial emissions, especially from very large or remote operations that are difficult to monitor, such as tailings ponds, shale oil pumping sites or pipelines. The satellite’s sensors are designed to calculate emissions within “pixels” that are 50 metres square, and assemble concentration estimates on areas extending over 250 of these units, which amounts to a 12-square-kilometre region. Early customers are not only emitters, but also regulatory bodies as well as competitors curious to find out how their rivals are doing in terms of emissions reduction. “We blazed the trail five years ago,” Germain says. “People didn’t think [this] could be done.”
Early generations of satellites, say experts like Ray Nassar, an atmospheric scientist at Environment and Climate Change Canada, were never intended to track carbon in the atmosphere. But climate scientists and agencies like NASA have developed observational and computational techniques for not only measuring concentrations of carbon but, increasingly, distinguishing between natural emissions and those from human sources.
What’s becoming apparent – especially with the addition of machine learning onboard – is that these satellites can provide emissions data not just to regulatory bodies, but also to asset managers and investors seeking to mitigate climate risk.
For example, Carbon Delta, a Zurich firm that evaluates satellite emissions data using artificial intelligence, has amassed an inventory of 22,000 companies with tens of thousands of hard assets that give off carbon. Its algorithms generate what the company calls a “climate value at risk” metric, essentially a dollar value related to that portion of a company’s assets considered vulnerable to carbon taxes, regulatory penalties or other charges.
The shift for business came with the adoption of the Paris climate accord in 2015, says Carbon Delta co-founder Oliver Marchand.
“Before Paris, the corporate world wasn’t so much included in the solutions to climate change” because the onus was on national governments to implement reduction strategies, he says. With more multinational corporations involved, there’s been growing demand from both firms and large investors for company- and industry-wide emissions data that is both accurate and consistent but hasn’t been sanitized to make the company look good, he says.
The purchasers of this analysis are asset managers and institutional investors, including large sovereign funds and pension plans. Carbon Delta counts a number of Canadian firms among its clients, including Manulife, the Caisse de dépôt et placement du Québec, Desjardins and TD Bank, according to Marchand.
David Lunsford, also a Carbon Delta co-founder and head of development, points out that the firm is now using its algorithms and satellite data to develop other metrics for investors, such as “two-degree compatibility” evaluations. These would essentially offer scenarios predicting whether a firm can survive if global average temperatures exceed two degrees Celsius.
The Bank of England and some of the bond rating agencies have been examining Carbon Delta’s methodologies, according to Lunsford. But the most encouraging signal is coming from the investment world itself. “We’re seeing a lot of asset managers seriously integrating climate risk into their management processes,” he says.
The prospect of using satellites to track carbon goes back to the 1992 adoption of the United Nations Framework Convention on Climate Change and its requirements that advanced economies create and report their emissions. Emissions data at the time mainly came from reporting at facilities and involved cataloguing carbon inputs, such as fuels used in power plants, Nassar, the satellite expert, says. But climate scientists have worked for years to find more precise ways of measuring emissions from urban areas or those related to activities such as the extraction and burning of heavy oils. “CO2 is easy to measure but hard to measure well,” he explains.
Satellites provided an opportunity to fill out the picture from above because they could detect plumes from smoke stacks or methane leaks. Earlier satellites could offer fairly accurate estimates of carbon concentrations high up in the atmosphere, but ground-level carbon has proven more difficult to gauge. The reason is that carbon dioxide accumulates in the atmosphere and lasts for a long time, so it can be difficult to measure incremental changes due to human activity.
In recent years, climate scientists have developed methodologies for making more accurate estimates of low-level carbon using satellite imagery. For example, Nassar led a team that developed ground-breaking techniques for using satellite data to estimate the emissions from U.S. power plants. The group’s findings, published last year in an academic journal, showed for the first time that the satellite images could be used to calculate concentrations that came within one to 17 per cent of the readings generated by plant-level sensors. In other words, satellites, with their broad field of vision, could potentially collect vast amounts of reasonably accurate emissions estimates.
The satellite used in Nassar’s analysis was NASA’s Orbiting Carbon Observatory-2 (OCO-2), one of a handful of relatively new orbiters purpose-built for the detection of atmospheric carbon.
Several scientific studies published in the journal Science in 2017 showed how vast amounts of carbon data gathered by such satellites has shone new light on the emissions cycles generated by complex atmospheric and terrestrial cycles, including forest fires, El Nino and extensive biomass loss due to harvesting. [pullquote]“The next step in our understanding of Earth’s carbon dynamics will be to build sensors, satellites and computer models that can distinguish human activity from natural processes,” the authors of the papers wrote in an essay on The Conversation blog.[/pullquote]
The challenge with these new satellites, Nassar says, is that their fields of vision are relatively narrow and episodic, even though they are able to detect tiny amounts of carbon. OCO-2, launched four years ago, can capture images eight kilometres wide but only returns to any given location twice a month – a frequency that may not be sufficient to develop rigorous measurements of industrial emissions. China and the European Space Agency are developing newer satellites, which can observe areas of 100 to 200 kilometres wide. As Nassar points out, a constellation of such satellites, deployed at different altitudes, could help overcome other observational gaps created by cloud cover.
Other unexpected applications have surfaced elsewhere. In Australia, for example, recent data from satellite tracking showed that promised carbon reductions from re-vegetation projects in Queensland, worth $1.07 billion and funded from carbon credits, were overstated: The plantings likely made little difference.
Montreal’s GHGSat represents one of the first private-sector forays into carbon-tracking satellites and has the first mover’s advantage of knowing customers’ needs.
Oil and gas companies, founder Germain and his team have discovered, aren’t just looking for carbon dioxide data. They’re also looking to spot potentially explosive methane and gas leaks from facilities.
Says Germain: “One of the most interesting things about this business is that almost universally, customers are interested in our science because they know they have challenges on their hands and want to address them.”