An analysis found that Climate TRACE may substantially underestimate city vehicle CO2 emissions, raising concerns about data accuracy in climate policy.
Some of the world’s most widely used climate emissions estimates could be missing far more pollution than anyone realized.
A new study from Northern Arizona University reports that the global greenhouse gas emissions database created by the Climate TRACE consortium, co-founded by former Vice President Al Gore, may be undercounting vehicle carbon dioxide emissions in cities by an average of 70%.
The findings, published in Environmental Research Letters, come as governments and cities increasingly rely on high-resolution emissions data to shape climate policy and track progress toward emissions goals.
Led by Kevin Gurney, a professor in NAU’s School of Informatics, Computing, and Cyber Systems (SICCS), the study examined how Climate TRACE estimated emissions from cars and trucks and compared those figures against established transportation and fuel-use data. According to Gurney, the discrepancies — combined with similar issues his team previously identified in power plant emissions estimates — raise broader concerns about the reliability of rapidly emerging AI-driven climate monitoring systems.
“Given the importance of vehicle CO2 emissions in cities, we carefully examined the Climate TRACE data which relied on promising new artificial intelligence-based approaches,” Gurney said. “When combined with our previous study on Climate TRACE power plant CO2 emissions, our results suggest that the Climate TRACE data significantly underestimate over half of U.S. fossil fuel-based CO2 emissions in cities.”
A city-by-city comparison found gaps
Gurney and his colleagues compared Climate TRACE estimates for urban vehicle CO2 emissions in the United States with Vulcan, a similar “onroad” emissions database developed in Gurney’s lab. Vulcan is calibrated using official traffic and energy use data.
“While the Vulcan onroad data is not perfect, with uncertainty of about 14%, this is far lower than the differences found when we compared 260 city vehicle CO2 emissions in the U.S. to the Climate TRACE database,” said Bilal Aslam, a SICCS postdoc and co-investigator on the study. “The Climate TRACE CO2 emissions were, on average, 70% lower than those same emissions in the Vulcan onroad CO2 emissions database.”
“Individual cities such as Indianapolis and Nashville were lower by more than 90%,” added Pawlok Dass, a research associate in SICCS and contributor to the study. The study’s authors suspect that the undercounting may also occur outside the United States, and they raise broader concerns about other parts of the Climate TRACE database.
AI still needs scientific guardrails
The authors say artificial intelligence could become a powerful tool for measuring many environmental indicators, but they stress that accuracy still depends on scientific rigor, transparency, and review by experts. Reliable greenhouse gas estimates remain a foundation for effective climate policy.
The publication also offers several recommendations for improving Climate TRACE’s work so that it can better support policy and budget decisions aimed at cutting emissions.
“We will never estimate emissions with perfect accuracy, but we must ensure that the data shared with policymakers and the public is unbiased and meets best practices and the most rigorous scientific standards available,” Gurney said. “Without this, we mislead decision-makers and potentially lose public trust in our ability to tackle climate change.”
Reference: “Assessing the accuracy of the Climate Trace global vehicular CO2 emissions” by Kevin R Gurney, Bilal Aslam and Pawlok Dass, 5 May 2026, Environmental Research Letters.
DOI: 10.1088/1748-9326/ae6355
Disclosure: Gurney, who specializes in atmospheric science, ecology, and public policy, has spent the past two decades developing a standardized system quantifying greenhouse gas emissions in the U.S. His Vulcan and Hestia projects, funded by multiple federal agencies, quantify and visualize greenhouse gases emitted across the entire country down to individual power plants, neighborhoods, and roadways, identifying “hotspots” and enabling better decisions about where to cut emissions most effectively. His estimates have shown excellent performance when compared to direct atmospheric monitoring.
Never miss a breakthrough: Join the SciTechDaily newsletter.
Follow us on Google and Google News.
Full article can be found at: https://scitechdaily.com/scientists-discover-major-errors-in-al-gore-founded-climate-pollution-database/
