JT Jeeraphun - stock.adobe.com

Green IT: How Google’s cloud and AI tools are supporting a methane-tracking satellite project

Cloud and AI will help to analyse data from MethaneSAT, due to launch on a SpaceX Falcon 9 in March 2024

A satellite due for launch will track how much methane is being emitted by oil and gas facilities, aiming to encourage moves to reduce the amount of harmful gas emissions in the atmosphere, with some help from artificial intelligence (AI) and cloud computing.

The Environmental Defense Fund’s (EDF) satellite, MethaneSAT, aims to map, measure and track methane emissions. The satellite is due to launch in early March on a SpaceX Falcon 9 rocket, and will orbit the Earth 15 times a day at an altitude of over 350 miles.

Earlier this month, the MethaneSAT team finished assembling its satellite in Colorado and shipped it to California’s Vandenberg Space Force, from where it will be launched.

The MethaneSAT satellite carries two passive infrared spectrometers. The project said that current satellites focus on either scale or precision, by tracking methane emissions over large geographic areas or targeting point sources. MethaneSAT will fill the gap in between, using a 200km field of view to measure emissions on a regional scale as well as using high-resolution sensors that are capable of narrowing emissions sources to a single facility.

Greenhouse gas pollution continues to warm the planet: 2023 was the hottest year on record, according to the US National Oceanic and Atmospheric Administration, and the past 10 years have been the hottest years since 1850. As such, reducing this warming is key to reducing the risk of extreme events such as wildfires and droughts as well as improving air quality.

Methane from human sources is responsible for about 30% of global warming today, and a big contributor of methane in the atmosphere comes from extracting oil and gas. But countries around the world have under-estimated just how much methane is being pumped out by infrastructure.

Google said that it is working with the EDF on a project to combine their science and technology to reduce methane emissions. “This is one of the most powerful, short-term actions we can take to reduce warming,” said Yael Maguire, vice-president of Geo Developer and Sustainability at Google.

Google Cloud will provide the MethaneSAT team with the computing capabilities required to process MethaneSAT’s massive data stream quickly and securely.

“Our goal is to measure, process and post emissions data as quickly as possible, and Google’s cloud resources will be an invaluable asset. The company will also help us improve our database of global oil and gas infrastructure so emission data for specific regions can be accurately attributed to verified facilities,” the EDF said.

Google said by powering the methane detection algorithms and applying AI to satellite imagery to identify oil and gas infrastructure around the world, it will help EDF quantify and trace methane emissions to their source.

“With this information, energy companies, researchers and the public sector can take action to reduce emissions from oil and gas infrastructure faster and more effectively,” said Maguire.

To calculate the amount of methane emitted in specific places and track those emissions, Maguire said that EDF had developed algorithms powered by Google Cloud in collaboration with scientists at Harvard University’s School of Engineering and Applied Science and its Center for Astrophysics, and scientists at the Smithsonian Astrophysical Observatory.

Google said it was also creating a global map of oil and gas infrastructure, with the goal of understanding which components contribute most to emissions. Google will use AI to identify oil and gas infrastructure, such as oil storage containers, in its imagery, and combine this with EDF’s information about oil and gas infrastructure to locate where the emissions are coming from.

Once the infrastructure map is built, it is then possible to overlay the MethaneSAT data on top to see how emissions correspond to specific infrastructure, which makes it easier to understand the types of sources that generally contribute most to methane leaks.

“This information is incredibly valuable to anticipate and mitigate emissions in oil and gas infrastructure that is generally most susceptible to leaks,” said Maguire.

The insights will be available later this year on MethaneSAT’s website and accessible through Google Earth Engine, which Google said has more than 100,000 monthly active users.

Maguire said that making MethaneSAT datasets available on Earth Engine would make it easier for users to detect trends and understand correlations between human activities and environmental impact, such as by combining the methane data with other datasets like land cover, forests, water and regional borders.

The oil and gas industry emits 80 million tonnes of methane every year, making it the second-largest source of human-caused methane after agriculture.

The satellite will be able to monitor regions accounting for 80% of global oil and gas production, and will be able to measure how much of the gas is escaping – and how fast.

“Our goal is to measure, process and post emissions data as quickly as possible, and Google’s cloud resources will be an invaluable asset. The company will also help us improve our database of global oil and gas infrastructure so emission data for specific regions can be accurately attributed to verified facilities,” the EDF said.

Having easier access to the data should make it easier to find out where the worst emissions are coming from and then work towards getting them reduced. While attempts to reduce emissions will inevitably vary by country, being able to show directly that the gas is coming from a particular installation is likely to concentrate the minds of environmentalists and infrastructure owners.

According to the International Energy Authority (IEA), the methane emissions from oil and gas operations can vary greatly, with the best performing countries delivering an “emissions intensity” more than 100 times lower than the worst performers.

But there is some limited good news. The IEA said that the ways to prevent these methane emissions well-known and have been deployed in multiple locations around the world. “Many measures can also save money because the outlays required to deploy them are less than the market value of the methane that is captured and can be sold,” it said.

Depending on the price of natural gas, somewhere between 50% and 80% of the ways to reduce emissions from oil and gas operations worldwide could be implemented at no net cost; implementing these would cut oil and gas methane emissions by more than 60%, the IEA said in a report last year. As such, cutting oil and gas methane emissions could be one of the best short-term options for climate action.

Read more about greenhouse gas emissions and technology

Read more on Artificial intelligence, automation and robotics

Data Center
Data Management