Reduction of Methane Emissions Using AI and Hyperspectral Sensors: Oxford University Research

Jakarta

The zero emissions or Net Zero target apparently does not only focus on reducing CO2 carbon, but also methane emissions which play an important role in the process of increasing temperatures.

Methane turns out to be 80 times more effective at trapping heat than CO2. Therefore, reducing methane emissions from anthropogenic sources will have an impact on slowing the global warming process.

It is estimated that the reduction in methane emissions that can be achieved will be able to prevent global warming by almost 0.3 degrees Celsius in the next two decades. However, until now there are only a few methods that can be used to map the amount of methane in air images because this element is a transparent gas that can only be captured in the spectral range of large satellite sensors.

That’s why Oxford researchers are trying to detect methane plumes by narrowing multispectral bands and then using AI to read larger amounts of data.

“What makes this research exciting is the fact that more hyperspectral satellites will be deployed in the coming years, including from ESA, NASA and the private sector,” Vit Ruzicka, a computer science student at Oxford University.

If these sensors are combined, they will provide global hyperspectral coverage, enabling automatic methane detection.

More Accurate Methane Detection Tools

Reporting from the University of Oxford website, the researchers trained the artificial intelligence model using 167,825 hyperspectral tiles, each of which represents an area of ​​1.64 square kilometers. This data was captured by NASA’s AVIRIS air sensor in the Four Corners region, United States.

The algorithm is then applied to data from other hyperspectral sensors in orbit, such as data collected from NASA sensors installed on the space station. Overall, the model had more than 81 percent accuracy in detecting large methane plumes. This figure is 21.5 percent more accurate than the previous most accurate approach.

This method is also able to reduce the false positive detection rate of methane by up to 41.83 percent compared to the previous approach.

This project is funded by the European Space Agency (ESA) laboratories through the 3CS or Cognitive Cloud Computing in Space campaign and is carried out as part of the Trillium Technologies Networked Intelligence in Space (NIO Space) initiative.

To further detect, the researchers have opened the annotated dataset as well as the methane modeling code on the project page on GitHub. In the face of climate change, such techniques enable more independent global validation of greenhouse gas production and leakage.

“This approach could easily be extended to other important pollutants, our ambition is to apply this approach inside satellites so that instant detection becomes a reality,” said Andrew Markham, Oxford University professor of computer science.

They developed the model to operate directly within the satellite itself, allowing other satellites to provide follow-up observations as part of the NIO Space initiative.

Ruzicka said, “Such on-board processing could mean that initially only priority warnings need to be sent back to Earth, for example a text warning signal with the coordinates of an identified methane source.”

Apart from that, he also assessed that data processing on the satellite itself would encourage independent collaboration from other satellites.

This will enable early detection that serves as a guidance signal for other satellites in the imager constellation of the desired location.

Watch the video “Kominfo will issue regulations for AI”

(nah/nah)

2023-12-09 10:00:46
#Experts #Develop #Methane #Detection #Space

Leave a Comment

This site uses Akismet to reduce spam. Learn how your comment data is processed.