Identifying Methane Super-Emitters with Machine Learning for Climate Action
Methane, the second-largest contributor to climate warming after carbon dioxide, is a potent greenhouse gas that retains 30 times more heat than carbon dioxide over a century. However, unlike carbon dioxide, methane has a shorter lifespan in the atmosphere of about 10 years, providing an opportunity for swift climate action. By reducing methane emissions, we can witness a tangible reduction in global methane levels within a decade, helping to mitigate the enhanced greenhouse effect.
Methane super-emitters, which release a disproportionately large amount of methane compared to other emitters, are often found in industrial facilities such as oil and gas operations, coal mines, and landfills. These super-emitters can be easily fixed through relatively simple repairs, offering significant climate gains. However, the challenge lies in identifying these super-emitters to target efforts effectively.
To address this issue, researchers from the SRON Netherlands Institute for Space Research have developed an algorithm that utilizes machine learning to automatically discover methane super-emitter plumes in data collected by the Copernicus Sentinel-5P satellite. This satellite, equipped with the Tropomi instrument, produces a global map of methane concentrations every day. By analyzing the shortwave infrared bands observed by Sentinel-5P, the algorithm can detect methane with remarkable precision.
The wealth of data provided by Sentinel-5P plays a critical role in comprehending the consequences of methane emissions and combating climate change. The algorithm automatically calculates the emissions associated with the detected methane concentrations and concurrent wind speeds. The identified methane super-emitters are then communicated to other satellites with higher resolution for precise source identification.
The detection of methane emissions also relies on combining data from multiple satellites. While Sentinel-5P offers daily global coverage and high-precision methane measurements, it cannot pinpoint the exact source of methane leaks due to its relatively coarse spatial resolution. On the other hand, Sentinel-2 satellites can identify precise locations of major methane leaks with a resolution of 20 meters, but lack daily global coverage. The Sentinel-3 satellites, equipped with multi-band radiometers sensitive to methane concentrations, can retrieve methane enhancements and detect large methane leaks every day with global coverage and a ground pixel resolution of 500 meters.
Through the combination of data from Copernicus Sentinel-5P, Sentinel-2, and Sentinel-3, researchers are able to accurately detect, localize, and quantify methane emissions on a global scale. This three-tiered approach provides valuable insights into methane super-emitters, allowing for targeted efforts to reduce emissions and mitigate climate change.
Source: Not provided