Good morning Sander,
I am currently trying to smooth EAC4, MOPITT, and IASI CO datasets with TROPOMI in order to make their total columns directly comparable, particularly over specific regions rather than just at individual points. At first, I thought HARP could handle the entire process, but after going through several forum discussions it seems that applying the TROPOMI averaging kernel and a priori to external datasets at regional scales is not straightforward. From the literature, including the paper “1.5 years of TROPOMI CO measurements: comparisons to MOPITT and ATom”, I see that it has been successfully done, but the practical steps remain somewhat unclear. My understanding is that one needs to regrid the external dataset (such as EAC4) onto the TM5 pressure grid used by TROPOMI, then apply the averaging kernel formula x^=xa+A(xmodel−xa)\hat{x} = x_a + A(x_{\text{model}} - x_a)x^=xa+A(xmodel−xa) for each pixel, and finally integrate over pressure levels to obtain smoothed columns that are consistent with TROPOMI retrievals. I have managed to regrid EAC4 to TM5 levels, but I am still uncertain about the correct way to apply the TROPOMI column averaging kernel and how to scale this process up from point-based examples to regional comparisons. Since I am not very familiar with handling L2 products in practice, could you suggest a possible workflow or best-practice approach to achieve this smoothing for regional studies, ideally ensuring consistency with how the TROPOMI columns are officially defined?
Kind regards,
Gezahegn