Skip to main content

D4.4 Benchmarking of plume detection and quantification methods

Authors
Janne Hakkarainen, Johanna; Tamminen, Janne Nurmela, Hannakaisa Lindqvist; Diego Santaren, Gregoire Broquet, Frederic Chevallier, Erik Koene, Gerrit Kuhlmann, Dominik Brunner
Abstract
This report synthetizes the benchmarking and the evaluation of the main current plume inversion methods that has been conducted in Task 4.2 of the CoCO2 project. The methods have been identified based on their ability to quantify the carbon dioxide (CO2) emissions from satellite images of plumes downwind of sources with relatively low computation cost. The methods are either based on the analysis of individual plume images or on averaging the images over long periods of time. Five computationally light methods for emission detection and quantification have been tested: the Cross-sectional flux method (CS), Gaussian Plume model fitting (GP), the Integrated Mass Enhancement method (IME), Light Cross Sectional flux methods (LCS) and the Divergence method (Div). The methods have been described in detail in Deliverable 4.3, and here we focus on their implementation and performance.