---

title: "Review for manuscript egusphere-2024-2666"

---

## Overall impression

The manuscript proposes and describes a new processing of GOME2 data that improves the SIF retrieval. The results do show convincing improvements and the description is clear and detailed. The resulting dataset will be useful for the community and this manuscript will serve as a good reference for those who need to go in the details. 

I am not an expert in the actual SIF retrieval nor the GOME instruments. I must admit that this manuscript is more technical than I initially thought, and that it thus fall beyond my comfort zone in terms of technical details. Therefore I cannot pronounce myself too much on the very technical satellite retrieval details and hope that this is covered by other reviewers.

## Specific points

L61: Maybe state that this is FLUXCOM X-BASE products

L88: Not too sure (for me) how the information on the throughput tests is actually useful for the average reader. Maybe some more context (if needed) could help.

L116: What seems also very clear is a downward trend after the jump. This would be good to point out (and state the reasons behind)

L140: is this assumption correct given noted trend in global greening?

Fig 2: To be clear, the +0.1 should also be mentioned in the legend

L192: for completion, please state what E0 is in Eq. 3.

Fig 11: I feel this visualization does not show well the actual improvements. Consider additional/complementary plots showing residuals with respect to the mean seasonal cycle, or differences with respect to one product.

Fig 12: why these who dates, which are showing very similar information? Would it not be more approapriate to show a date after the sensor jump of 2013 to see if things hold there too?

Fig 13: it is a pity that the spatial variability is not well showcased. Could you consider adding another figure showing differences in spatial patterns over these regions (i.e. showing the actual spatial variability with maps rather than time series)?

L398: My understanding is that FluxSAT does use some GOME2 data at some point in their processing, while FLUXCOM XBase does not at all. Please investigate/confirm is this is the case and discuss the possible repercussions (and circularity) that may come out from this comparison with SIFTER.

L453: It says here the SIFTER V3 data "will become publicly available", but this does not say when. It should be made available along with this manuscript.

What about the code? It would be good practice to provide the code used to do this processing and the analyses done within this study.