Feasibility of Detecting Low-Level Methane Emissions from Abandoned Oil Wells Using a Multi-Band, Multi-Pass Satellite Approach

In this study, we develop a multi-band, multi-path (MBMP) model using shortwave infrared (SWIR) data to quantify point-source methane emissions from abandoned oil wells. Atmospheric methane concentrations reached about 1,942 ppb in 2025, more than doubling since pre-industrial times. As the second most important greenhouse gas after CO2, methane emissions from both natural and fossil fuel–related sources contribute significantly to climate change. The greenhouse effect of atmospheric methane (CH4) is approximately 25–28 times greater than that of carbon dioxide (CO2). Even small emissions from abandoned oil wells, if left unmanaged, can contribute to long-term global warming. In Japan, we found that backfilling (plugging) old wells is highly effective in reducing greenhouse gas emissions. The abandoned oil wells represent a minor but ongoing source of methane, emitting approximately 200 m³/day. Even assuming a cost of 50 million JPY per well, the cost to reduce one ton of CO2 is approximately 3,500 JPY (25 USD). This is a lower cost compared to other greenhouse gas reduction methods. Based on the field measurements and model estimates, methane emissions were estimated at 7,300 t-CO2e per year. This represents a significant proportion compared to the total emissions of 47,400 t-CO2 reported in Akita Prefecture for the 2019 fiscal year. This study presents a comprehensive satellite-based methodology for monitoring methane leakage from abandoned oil wells using time-series observations from the ESA Sentinel-2 satellite and develops a MBMP model based on these observations. The study aims to conduct the first quantitative evaluation of small amount of methane emissions from abandoned oil wells in Japan by integrating high-resolution satellite data with in situ field survey measurements obtained at the former Kurokawa oil field in Akita Prefecture.