Monitoring methane emissions at Hassi Messaoud from 2021 to 2024 using CH4 Sentinel and integrated methane inversion: assessing Algeria’s progress toward its 2030 GHG reduction target

Methane (CH4) is a potent greenhouse gas (GHG) with an outsized impact on near-term warming, and the oil and gas sector is the largest man-made contributor to post-industrial revolution anthropogenic CH4 emissions. Accurate measurement and localisation of CH4 emissions is crucial for mitigating the near-term effects of climate change, and for improving the efficiency of petrochemical facilities. A novel, open source, software tool - CH4 Sentinel - was developed to process Sentinel-2 data and quantify CH4 super emitting point sources using a machine learning approach. CH4 Sentinel was applied to the Hassi Messaoud oil field in Algeria for the years from 2021 to 2024, to determine if progress was being made to Algeria’s 2030 GHG reduction goal. CH4 Sentinel processed Short Wave Infrared (SWIR) data from Sentinel-2’s Multi Spectral Instrument (MSI) using a Multi-Band Multi-Pass method (MBMP) to visualise the plumes. Then a novel processor-light machine learning approach with an XGBoost regressor was employed to quantify superemitting plumes in tonnes per hour (t/h). The regressor model showed an R2 value of 0.81 and a RMSE of 1.57 t/h on unseen evaluation data. To provide context for plume measurements Sentinel-5P Tropospheric Monitoring Instrument (TROPOMI) data was processed using the Integrated Methane Inversion tool (IMI) for the years 2020 to 2024 to estimate total yearly emissions for the field. The IMI analysis revealed that while overall emissions at Hassi Messaoud are significantly underestimated by the Global Fuel Exploitation Inventory (GFEI), there has been a decrease since 2021 of 29,352 t/y. The Sentinel-2 analysis indicates that CH4 super emissions have significantly reduced at Hassi Messaoud between 2021 and 2024 by an estimated as 25,220 tonnes per annum, approximately 86% of the IMI reduction for that period.