The term “surveillance capitalism” is used to describe the practices of companies like Google and Meta, which actively use consumers’ private (digital) experiences to create and commodify behavioral data. The philosopher Shoshana Zuboff, who is widely credited with popularizing the term, wrote in a 2019 TIME article, “It is no exaggeration to say that the Internet is owned and operated by private surveillance capital.”

Yes, your browse history is a window into your psyche, and much of that information has been boiled down into datapoints that strongly indicate what you will buy and who you will vote for. And yes, Orwell was onto something. But what if the vast surveillance machine that we’ve unleashed – now with the help of AI and some 3,000 satellites currently orbiting Earth – could be leveraged for good? As a means not of pernicious opportunism, but of transparency and meaningful action. Could there be such a thing as surveillance environmentalism?

Tracking Emissions from Space

A global coalition known as Climate TRACE, founded in 2020, uses satellite remote sensing data and machine learning algorithms to identify and estimate sources of greenhouse gas emissions around the globe. With the aid of approximately 300 satellites, owned and operated by NASA, the European Space Agency, the Japanese Aerospace Agency, and Planet, among others, Climate TRACE is now tracking emissions and other metadata from over 352 million assets, including manufacturing facilities, oil and gas fields, industrial farms, shipping ports, and more. (The forerunner to this came in 2019, when a group of nonprofits, including UK-based Carbon Tracker and U.S.-based WattTime, received a $1.7 million grant from Google.org to employ AI to monitor power plant emissions from space; WattTime’s executive director, Gavin McCormick, is a Climate TRACE co-founder.)

The coalition’s emissions map represents the world’s largest inventory of GHG emissions, spanning continents, sovereign borders, and industry sectors. Further, the data are primarily based on direct observations, with minimal reliance on disclosures from governments or municipalities. And buildings are next. Later this year, Climate TRACE plans to unveil a more robust database focused directly on tracking emissions from residential and commercial buildings.

Currently, Climate TRACE’s data are derived from EDGAR (Electronic Data Gathering, Analysis, and Retrieval), an independent and public global database of anthropogenic emissions and air pollution on Earth. A downside to this approach is that EDGAR provides coarse resolution data, meaning emissions figures are on a municipal scale.

According to Kyle Bradbury, director of the Energy Data Analytics Lab at Duke University, the coalition member devoted to modeling emissions from buildings, “we’re trying to make this higher resolution than any of the existing datasets.” Bradbury says Climate TRACE’s modeling can pinpoint emissions figures from building all the way down to a one square kilometer grid. “We’re looking to account for emissions from direct consumption … we use estimates of the footprints of those buildings.”

This process begins with Sentinel imagery, a publicly available satellite dataset that is a key input for the planet’s human settlement layer. Bradbury’s work is focused on “ingesting the data derived from that imagery for building estimates and adding to it by including other emissions factors, energy use intensity values” and more. This quantum leap is largely made possible with the aid of Sentinel 2. When asked if such an undertaking could even be possible without the benefit of machine learning, he offers an emphatic “No!”

Translating Global Data into Local Action

The value of such a tool cannot be overstated. But at some point, the question of how to leverage all that metadata and incentivize action on the part of municipalities and other large emitting bodies must come to bear. “Our goal is certainly to have transparency around emissions, which is a global problem” that affects everyone, says a Climate TRACE spokesperson. Some governments and organizations can be disingenuous in their self-reporting, but Climate TRACE’s purpose isn’t to play big brother. “There are a lot of countries, particularly developing countries, that don’t have the resources to do this type of analysis,” the spokesperson adds. In that regard, the coalition’s work can inform decarbonization efforts across all major emitting sectors, from forestry to agriculture to manufacturing. “We hope that it’s a public service, in addition to being a support mechanism for transparency and accountability.”

Drew Shula, CEO of Verdical Group, a California-based sustainability consultancy, doesn’t shy from the idea of calling out big GHG emitters. “This is the opposite of greenwashing,” he says. “I think this can be a powerful tool to hold them accountable.” Having reliable data to guide decisions on a multitude of scales, from governmental to hyperlocal, has largely been missing from the equation until now. To Shula’s point, tracking specific emissions sources across sectors with such granularity, down to the neighborhood, could well be used to identify bad actors that are failing to comply with commitments laid out in the Paris Agreement, the 2030 Challenge, or similar.

“Every industry, every company, every country is going through the process of drawing down emissions,” Shula says. “They’re not required to reduce emissions in most cases, but there will be more pressure for them to change [once this data] come out. You can’t manage what you don’t measure.”

The eyes in the sky may be watching, but unlike so-called surveillance capitalism, the metadata being gathered by the Climate TRACE coalition is not being applied on a micro, behavioral scale. In other words, your wood-burning stove and backyard barbecues are not getting factored into the metadata.

When the coalition’s building emissions datasets for both residential and non-residential categories come out, it will indeed be a watershed moment. How such a tool will be received by governments and industries alike is likely a question that will evolve over time, much like the machine learning systems that enable such an ambitious endeavor to be carried out in the first place. According to Bradbury, the hope is that this “can be used for pathways towards sustainable development. That is very much an aspiration for these data.”