The 2015 Paris Climate Agreement has generated a great deal of interest in how we can globally monitor greenhouse gas emissions – particularly those related to fossil fuels. Consequently, several space agencies are developing space-based imagers capable of observing CO2 plumes emanating from large industrial and urban sources.
In order to assess their potential, researchers at the LSCE have developed an algorithm that estimates CO2 emissions from the burning of oil, natural gas or coal, based on satellite measurements of CO2 "columns".
This algorithm was tested on the archives of two NASA Orbiting Carbon Observatories: OCO-2 and OCO-3. These provide (since 2014 for OCO-2 and 2019 for OCO-3) dense data at a high spatial resolution (3 km2) and on a narrower swath than future imagers.
The emissions calculated from the two OCOs were compared to a global emissions inventory based on international statistics, which were placed in their temporal and spatial context using "auxiliary" data.
The respective variabilities of the two datasets are largely in agreement, despite their associated uncertainties. The median values of the emissions vary in a consistent way at various time scales: year, month, day of the week, first hours of the day, etc.
These results suggest that the differences between the satellite inverse model estimates and the inventory estimates are mostly random.
Emission trends can therefore be calculated robustly in areas with favorable satellite observation conditions – especially given that future space-based CO2 imagers will provide ten times more data. This will be the case as of 2025, when Europe will deploy a group of CO2 imagers covering a much wider swath (250 km versus 2-10 km).