NAU Study Says Climate TRACE Undercounts City Vehicle CO2 by 70%
Updated
Updated · SciTechDaily · May 24
NAU Study Says Climate TRACE Undercounts City Vehicle CO2 by 70%
1 articles · Updated · SciTechDaily · May 24
Climate TRACE’s city vehicle emissions estimates ran 70% below a benchmark U.S. database on average across 260 cities, according to a Northern Arizona University study in Environmental Research Letters.
The researchers compared Climate TRACE’s AI-driven car and truck calculations with Vulcan, an emissions database calibrated to official traffic and fuel-use data; Vulcan’s uncertainty is about 14%, far smaller than the gap found.
Indianapolis and Nashville were each underestimated by more than 90%, and the authors said the discrepancies, alongside earlier issues they found in power-plant estimates, may affect over half of U.S. fossil-fuel CO2 emissions in cities.
The study warns that governments increasingly using high-resolution emissions data for climate policy could be misled unless AI-based monitoring is paired with stronger scientific transparency, expert review and methodological safeguards.
When rival climate databases disagree by 90%, how can cities trust the data?
Is AI's massive energy footprint undermining its own mission to fight climate change?
70% Gap in City Vehicle CO2 Emissions: NAU Study Challenges Climate TRACE Data and Sparks Debate Over Urban Climate Policy
Overview
A major study from Northern Arizona University, led by Professor Kevin Gurney, revealed that Climate TRACE’s reported vehicle CO2 emissions for U.S. cities are on average 70% lower than those from the well-established Vulcan Project. The Vulcan Project, developed over two decades and validated against direct atmospheric monitoring, is known for its reliability and use of government data. This significant discrepancy raises concerns about the accuracy of AI-driven, satellite-based emissions inventories like Climate TRACE, highlighting the need for rigorous validation and transparency to ensure cities have trustworthy data for climate action and policy decisions.