Review of “Mapping of sea ice concentration using the NASA NIMBUS 5 ESMR microwave radiometer data 1972-1977” by Kolbe et al.

Summary

This paper presents a new sea ice concentration product from the Nimbus-5 ESMR sensor that operated from 1972 to 1977. Earlier sea ice products have been created from ESMR, but the sensor is under-utilized because of limited data quality and substantial amounts of missing data, plus the fact that it is a single-channel sensor while subsequent sensors are multi-channel and there is no overlap between them. The method here uses several filters to remove bad data, implements dynamical tie-points and uses a radiative transfer model with NWP data for atmospheric correction. Comparisons show good consistency with the NSIDC ESMR data product, but with extent values that are generally higher due to the different methods.

General Comment

This is a valuable new dataset that makes the ESMR data more useful for long-term timeseries analysis and extends the passive microwave sea ice record another 7 years. The filtering methods are well thought out and remove much of the bad data and the RTM with NWP help correct weather effects. It is clear that great care went into the product and that the output is as good as can be expected given the nature of the data. The comparisons with the NASA/NSIDC product are useful and highlights key differences in the methodologies. The paper is well-written and thorough. After revision in response to the very minor comments below, this is acceptable for publication.

Specific comments (by line number): I think of science as a process of gaining knowledge through the accumulation of evidence to continually get closer to knowing the truth.

95-113: The filter discussion is a bit jumbled in my view. In Line 97, “The first filter” is mentioned, but then in Line 101, it says “The following set of filters are applied…”, which is then described as “the second filter” even though it reads to some degree that it is the first filter. I guess I see that the first one is a bit different, but I would just say something like, “The following filters are used…” and then go one-by-one. Or if you want to separate out the analog filter, maybe say something like, “An initial analog filter is used…” and then say something like, “Next, several other filters are employed….”

103-113: Another issue is that the equations for some of the filters are used to describe included data and some are to described excluded. For example, Eq. 1 shows the range of valid values, while Eq. 2 and Eq. 3 describe thresholds to remove invalid values. It would be better to be consistent and have each equation describe valid values or each describes invalid values. Eq. 4 I’m not quite sure about- does “not equal to zero” mean remove or keep?

121: Section 3 is labeled as “The radiative transfer model”, but the sub-sections seem to go into other areas. I guess the RTM affects the subsections, but it seems like only the first part of Section 3 is specifically RTM and then the subsequent subsections are actually about the process of deriving SIC. I would suggest putting 3.1 to 3.3 into a Section 4 “Derivation of sea ice concentration”, or something along those lines.

182: What is the source of the SIC in ERA5? I think this is worth noting since it is an important element in deriving the tiepoints.

184-185: This isn’t clear to me. What do you mean by the 15-day averaging period is maintained even at the beginning and end of the data-set and when there are data gaps. I assume this means that wherever there is a gap, valid data would not start until 7 days after the end of that gap so that there is a full 15 days for the average period. Is this correct? If so, maybe slightly rephrasing this to be clear.

348-360: I’m not a fan of using bullet points for conclusions in a journal article. Maybe just a personal preference by me, but I think it looks and reads better as paragraphs, especially when the bullet points are complete sentences and not a list.

Please find my comments in the attached pdf.