A New View
on CO 2
A climate initiative uses satellite imagery and AI algorithms
to quantify carbon emissions from power plants.
BY CHRIS HAYHURST
35
Melting ice caps. Rising sea levels. Wildfires in increasing numbers.
ILLUSTRATIONS BY KEITH NEGLEY
These are some of the most common
photographable effects of global warming. But
how would the face of climate change shift if we
could better visualize—and quantify—destructive
carbon emissions? That’s where the Global
Emissions Monitoring System (GEMS), an initiative
funded by Google.org and led by San Franciscobased
nonprofit WattTime, enters the picture: The
project will leverage satellite imagery and artificial
intelligence (AI) algorithms to monitor the world’s
biggest atmospheric polluters.
The idea, says WattTime executive director
Gavin McCormick, is to hold coal-, gas-, and
oil-fired power plants accountable for the impact
their emissions have on human and environmental
health. While the primary target of the project is
carbon dioxide (CO 2 )—the heat-trapping gas that
contributes to global warming—the initiative also
will track other pollutants, including mercury and
sulfur dioxide.
“It’s amazing how hard it is to actually find out
where all the pollution is coming from,” McCormick
explains. A small number of countries provide
comprehensive reporting around which power
plants pollute how much, but, in most parts of the
world, he notes, “these plants are not tracked,
and there’s no way to prove or even know who is
causing the problem.”
WattTime is partnering with two other
nonprofits on the project: the London-based
Carbon Tracker Initiative and the World Resources
Institute in Washington, D.C. Using a monitoring
system developed by Carbon Tracker, the three
organizations will deploy machine learning