Press Release
Johns Hopkins APL Supports Climate TRACE’s Expanded Database of Greenhouse Gas Emissions
In continued support of the Climate TRACE coalition, APL used a combination of machine learning, satellite imagery and localized emissions data to advance its accurate, scalable and easily configurable greenhouse gas monitoring framework for road transportation.
Credit: Johns Hopkins APL
Using a combination of machine learning, satellite imagery and localized emissions data, APL advances its accurate, scalable and easily configurable greenhouse gas monitoring framework for road transportation.
With support from researchers at the Johns Hopkins Applied Physics Laboratory (APL) in Laurel, Maryland, Climate TRACE published an expanded database that now tracks greenhouse gas (GHG) emissions from more than 352 million assets, up from 80,000 in 2022. Announced at the annual United Nations’ Climate Change Conference in December, this expanded inventory is unprecedented in granularity and pinpoints nearly every major source of GHG emissions around the world.
Derived from satellite data, public and commercial information, and sector-specific models, Climate TRACE provides data at extraordinary resolution. As part of the global coalition, APL leads efforts to monitor road transportation emissions.
“Climate TRACE is unique in that we’re mapping emissions down to the asset level,” said Marisa Hughes, the assistant program manager of Human and Machine Intelligence and APL’s climate intelligence lead. “We want to know how much an individual power plant, or an acre of pastureland, or even transportation on a single road is emitting over time — and how that changes as technologies and policies are implemented and adapted. With that knowledge, we can take more informed climate action and invest in tools that have proven to be effective in lowering emissions.”
The Environmental Protection Agency estimates the transportation sector accounts for the largest portion of GHG emissions in the United States, and 81% of those emissions stem from road transportation — including passenger cars, trucks and utility vehicles. For years, researchers have tried to measure road transportation emissions more closely, but existing inventories were often outdated, incomplete and limited.
Elizabeth Reilly led a team of APL researchers that used artificial intelligence, satellite data and public and commercial information to produce road transportation emissions data for 10,000 urban areas around the world.
Credit: Johns Hopkins APL
APL developed a hybrid approach to accurately estimate road transportation emissions by combining machine learning, road network data and satellite imagery with localized emissions factor data such as information about vehicle types. The team’s methodology was recently peer reviewed and published by Cambridge University Press, establishing an accurate, globally scalable and easily configurable GHG monitoring framework.
“After producing road transportation emissions data for 500 major urban areas globally in 2022, we scaled our methods to produce this data for 10,000 urban areas this year,” said Elizabeth Reilly, a senior artificial intelligence (AI) researcher and member of APL’s Climate TRACE team. “This is a huge effort from both a technical and a computational perspective. Successfully tackling this scaling has resulted in unparalleled coverage of road transportation emissions for major cities around the world.”
APL’s collaboration with the Climate TRACE coalition is one of several projects at the Laboratory that are leveraging artificial intelligence to tackle climate intelligence challenges, such as forecasting climate tipping points to ensure climate security.