Climate TRACE tracks 660M pollution sources with AI

For decades, industrial polluters could hide behind fuzzy reporting and delayed audits while neighboring communities breathed in the consequences.

Now, satellites and machine learning are teaming up to eliminate that plausible deniability.

Climate TRACE, the non-profit coalition co-founded by former U.S. Vice President Al Gore, just launched an AI-powered tool tracking fine particulate pollution (PM2.5) across more than 660 million individual sources worldwide.

Built by a global network of over 100 universities, researchers, and AI scientists, the platform taps real-time satellite imagery and remote sensing to monitor emissions from almost every corner of the Earth.

As Gore noted, tracking 662 million sites simultaneously simply was not conceivable before artificial intelligence entered the picture.

Exposing the gap between reports and reality

Fine particulate matter is a quiet killer. Decades of health research have linked PM2.5 exposure to low birth weight, kidney failure, Parkinson’s, dementia, and type 2 diabetes. Even under legal thresholds, it still leads to tens of thousands of premature deaths in the U.S. every year.

Climate TRACE’s monitoring engine zeroes in on 137,095 pinpointed facilities, including roughly 4,000 designated "super emitters." The project builds on work pioneered by Joel Schwartz, the scientist whose findings helped ban leaded gasoline decades ago.

The real breakthrough lies in busting open the data black box. More than 100 countries worldwide lack any formal emissions inventory, while others rely on outdated self-reported spreadsheets.

Climate TRACE data from 10 months ago showed that actual oil and gas production and transport emissions were three times higher than officially acknowledged. Meanwhile, emissions in the first half of 2025 ticked up 0.13% compared to the first half of 2024, heavily driven by fossil fuels.

Instead of trusting press releases, the tool publishes raw data alongside visual plume maps, letting users watch fine particulate clouds drift toward nearby neighborhoods in real time.

Why it matters

For everyday residents and local organizers, hyperlocal visibility completely flips the power dynamic. When health issues surge in a specific zip code, affected communities no longer have to guess where the bad air originated. Having street-by-street plume tracking provides actionable leverage to contest permits and demand regulatory enforcement backed by hard data.

For heavy industry, this effectively kills the standard ESG playbook. Self-reported sustainability claims become useless when third-party algorithms monitor every flare, smokestack, and pipeline leak from space. Competitors and regulators alike can instantly cross-reference what a company claims against what sensors actually measure.

We think Climate TRACE is showing one of the best applied use cases for AI: turning unmeasurable, distributed harms into an indisputable public record. But turning satellite detections into real cleanup is the hard part. The data is here, but whether local authorities have the political appetite to act on it is still an open question.