The following components of the methodology described in Sadavarte, et al., 2021 were not implemented in the present software:
Adjustment of TROPOMI retrievals for surface elevation effects.
De-striping of TROPOMI imagery to mitigate across track artefacts.
Subtraction of emissions attributed to known oil and gas sources.
Removal of plume contributions from neighbouring coal mines.
In Sadavarte, et al., 2021 , the TROPOMI retrievals were modified by applying an additive correction of 7 () per kilometre of surface elevation. Because this adjustment introduces an artificial spatially varying offset, it affects all subsequent stages of the analysis, including the correlation based filtering, background determination, definition of the downwind box, and ultimately the enhancement and emission rate calculations. The publicly available TROPOMI product documentation does not provide a clear justification for this elevation dependent correction. It is possible that this procedure was required to compensate for biases specific to the retrieval processor version used in the original study.
In the same study, Sadavarte, et al., 2021 , corrected striping artefacts in the TROPOMI data using a fixed mask. This preprocessing step affects all subsequent stages of the CSF algorithm, including plume identification and background estimation. The TROPOMI product documentation acknowledges striping as a known instrument artefact and notes that a dedicated de-striping algorithm has since been developed. However, based on the metadata of the datasets used in this work, that algorithm does not appear to have been applied to the publicly available products analysed here.
It is the position of this report that pre processing of TROPOMI data should remain the responsibility of KNMI, the organisation charged with producing and maintaining the publicly released products. Any additional data modifications beyond the standard processing chain would require clear scientific justification and validation, and addressing such issues constitutes a separate research effort outside the scope of this work.
Failure to subtract emissions from oil and gas sources may lead to an artificial expansion of the inferred plume, which in turn can alter the estimated plume geometry and affect both transect placement and emission rate calculations.
Subtracting plume contributions from other mines implicitly assumes that all plumes within a given TROPOMI scene are either estimated simultaneously or are known a priori. Implementing such a procedure introduces substantial methodological complexity that is beyond the scope and resources of this study. Moreover, it is not clear whether plume subtraction would systematically decrease or increase the estimated emission rates, as it would affect both the inferred plume geometry and the background concentration against which enhancements are calculated.