How UNEP & IMEO Use AI to Track Methane Emissions

Nowadays, AI is being utilised across a wide range of applications, from streamlining administrative tasks to monitoring emissions and ecosystems.

According to the United Nations Environment Programme (UNEP), reducing methane emissions offers one of the fastest and most cost-effective ways to slow global warming. 

This is a critical focus, given that the United Nations (UN) attributes approximately one-third of current global warming to methane.

To tackle this challenge, the International Methane Emissions Observatory (IMEO) is leveraging AI to process vast volumes of satellite data, enabling governments and industries to identify and mitigate methane emissions far more efficiently. 

By blending AI, advanced technology and sustainable practices, IMEO is transforming environmental monitoring and driving global climate action.

Enhancing detection through human-AI collaboration

IMEO has integrated AI into its Methane Alert and Response System (MARS) to significantly improve detection and monitoring capabilities. 

The system ingests data from more than 30 satellite instruments, enabling it to analyse more than 1.3 million satellite measurements, a volume that would be impossible to review manually.

By filtering these immense datasets, AI highlights potential methane emission events that require expert review.

Human analysts then verify every single AI-generated detection, ensuring scientific accuracy and reliability throughout the entire process. 

This human-in-the-loop approach allows IMEO to process 12 to 15 times more data while maintaining rigorous standards.

“The climate crisis is accelerating and we are now on course to overshoot the 1.5°C limit in the coming years,” António Guterres, UN Secretary-General, told delegates at the super-pollutants reception during London Climate Action Week (LCAW). 

“Reducing methane is a fight we can win and benefit from in our own time.

“Our task is to keep that overshoot as small, short and safe as possible and to bring temperatures back down. 

“That can’t happen without drastically reducing emissions, starting now and accelerating the shift away from fossil fuels, starting now.”

By combining AI with satellite technology and expert validation, MARS provides governments and companies with timely information to support methane mitigation efforts.

Designing energy-efficient AI models

While AI offers immense potential, UNEP recognises that these systems require energy, water, critical minerals and computing resources, making sustainability a vital consideration during development. 

Consequently, IMEO prioritises AI tools that deliver measurable environmental benefits while minimising resource consumption.

The AI models embedded in MARS are explicitly lightweight: they require only a single graphics processing unit (GPU) for training and operate on standard processors that consume significantly less energy. 

This design reduces power requirements while continuing to deliver accurate methane monitoring on a global scale. 

By balancing the environmental costs of AI against the climate benefits of faster methane detection and mitigation, IMEO demonstrates that advanced technology can support climate action without creating unnecessary environmental impacts.

Driving sustainability through AI innovation

Looking ahead, IMEO continues to expand its use of AI to enhance methane monitoring, data integration and climate decision-making. 

By combining satellite observations, scientific research, and industry reporting, the organisation creates high-quality datasets that strengthen AI performance and improve overall transparency. 

Furthermore, AI supports the verification of methane mitigation by confirming whether emissions have ceased after corrective action is taken.

In the future, IMEO plans to develop further AI tools that make environmental data more accessible and actionable for governments, industries and other stakeholders.

At LCAW, António also added: “Multilateral development banks, climate funds, donors, insurers and development partners must join forces to put pre-arranged finance within reach of developing countries.

“Together, let’s ensure that resilience becomes the foundation for a safer, more secure and more sustainable future and that resources are made available for that to be possible.”

Additionally, the organisation promotes equitable access to AI technologies through open datasets, strategic partnerships and training programmes designed to build technical capacity. 

Together, these efforts demonstrate how responsible AI and technological innovation can contribute directly to long-term sustainability and global methane reduction goals.