Dear Authors,

Thank you for the detailed response to the reviewers' requests.

All issues raised have been fully addressed and the paper can be published upon the reviewers' recommendations. Before proceeding to the publication stage, I would like to ask you considering the following minor technical and technical issues (line and page numbers follow manuscript version 4):

p 23, lines 14+. The reasoning starting with "Interestingly" is misleading. The fact that the two field values based on different foreign continuum hypotheses are significantly different does not depend on the MT_CKD_3.2 self continuum. This inconsistency is foremost an indication of biases related to the two hypotheses. The conclusion that the assumption of a given foreign continuum is important for the actual analysis therefore is thus logically independent of the comparison with MT_CKD_3.2 and the disagreement with MT_CKD_3.2 should not be invoked to arrive at this conclusion.

Some of the following arguments in the paragraph might also be a little bit overstretched. Eg, the CAVIAR-lab 297 K data point seems to agree (within 1.2 sigma) with the Richard et al data and the CAVIAR lab temperature dependence. Consequently, there does not seem to be any anomaly here. Then, "the CAVIAR field estimate being closer to this temperature dependence when using the CAVIAR-lab foreign continuum", is not an argument, because it is based on the assumption that the Richard et al values must be used as a reference. If the Baranov & Lafferty data had been used as a reference then the CAVIAR-field w MT_CKD foreign continuum would be closer to that temperature dependence.

The whole paragraph (lines 8-22) needs to be rewritten as to avoid logical fallacies and circular reasoning.

p 24, line 2 : set "k" in italics.

Supplement,
last phrase p1 : what is "gradient in which"
3rd equation p2 : LHS (Delta tau ?) and equation number are missing
It seems that this equation should come after "Then subtracting tau' from tau:"

Dear Authors,

Thank you for your corrections.

I am happy to announce that the paper can now be published taking into account a technical correction in line 8 on page 23, where "Also shown on Figure 13 is two estimates" should read "Also shown on Figure 13 are two estimates".

With kind regards

Christof Janssen