Cloud detection over snow and ice with oxygen A- and B-band observations from the Earth Polychromatic Imaging Camera (EPIC)

Zhou, Yaping; Yang, Yuekui; Gao, Meng; Zhai, Peng-Wang

doi:https://doi.org/10.5194/amt-13-1575-2020

Articles | Volume 13, issue 3

https://doi.org/10.5194/amt-13-1575-2020

© Author(s) 2020. This work is distributed under
the Creative Commons Attribution 4.0 License.

https://doi.org/10.5194/amt-13-1575-2020

© Author(s) 2020. This work is distributed under
the Creative Commons Attribution 4.0 License.

Articles | Volume 13, issue 3

Research article

|

01 Apr 2020

Research article |

| 01 Apr 2020

Cloud detection over snow and ice with oxygen A- and B-band observations from the Earth Polychromatic Imaging Camera (EPIC)

Yaping Zhou, Yuekui Yang, Meng Gao, and Peng-Wang Zhai

Download

Final revised paper (published on 01 Apr 2020)
Preprint (discussion started on 21 Oct 2019)

Interactive discussion

Status: closed

AC: Author comment | RC: Referee comment | SC: Short comment | EC: Editor comment

- Printer-friendly version

- Supplement

RC1: 'Comment on Cloud Detection over Snow and Ice with Oxygen A- and B-band Observations from the Earth Polychromatic Imaging Camera (EPIC)', Anonymous Referee #1, 29 Oct 2019
- AC1: 'Reply-to-Reviewer1', Yaping Zhou, 27 Jan 2020
RC2: 'Referee's comments', Anonymous Referee #2, 22 Nov 2019
- AC2: 'Reply-to-Reviewer2', Yaping Zhou, 27 Jan 2020

Peer-review completion

AR: Author's response | RR: Referee report | ED: Editor decision

AR by Yaping Zhou on behalf of the Authors (27 Jan 2020) Author's response Manuscript

ED: Referee Nomination & Report Request started (05 Feb 2020) by Piet Stammes

RR by Anonymous Referee #1 (18 Feb 2020)

Suggestions for revision or reasons for rejection

I commend the authors for a thorough and professional response, and now that I can follow the data choices and logic I see how this is a sensible approach to improve atmospheric measurement. I fully support publication but have a bunch of minor comments below. Mostly they're very small changes to improve readability and clarify logic, but I have one slightly larger comment about Figure 7.

Figure 7 is fine as it is, but could be more informative as a 4x4 annotated heatmap or confusion matrix-like figure. The 4x4 elements would contain the % of each sample for each pairwise condition for (at least) LEO/GEO vs v01 and LEO/GEO vs new and (optionally) v01 versus new. The diagonal elements would be agreement between categories, and the off-diagonals the disagreement (e.g. LEO/GEO cloud fraction <5 % but cloud flag > 1). This would contain the same information but be a lot easier for readers to get figures from. I leave this up to the authors though, but if you want to keep the current version then please match the scales on the y axes in each row (see a vs b, c vs d), to allow immediate visual comparison.

MINOR COMMENTS:
p2L13: "UV and Vis/NIR" - suggest spelling out on first use

p2L26: Replace "fewer spectral channels available compared..." with "fewer spectral channels compared...".

p2L28: typo, need to remove "s" from "spectrometers" in the MODIS name.

p3l21: typo, "cloud edge[s]" (insert "s" since we're talking multiple).

p4L7: flipped words: "...the otherwise same..." --> "otherwise the same"

p11L4: "...as suggested in Eq. (9)". I don't believe the equation is making the suggestion, but you are making the suggestion and then applying Eq. (9)? Suggested text: "...for the simulations, with the sample restricted to a zenith angle difference of below 6$^\circ$".

p13L4: General comment for the section, could you add information on sample size? It would be more pleasant to know that without having to squint at figure labels and try to work it out in my head.

p14L19: "median" replace with "mean"? I'm nitpicking again, but OLS is more commonly introduced as mean unbiased estimator. Sure, if your Gaussian assumptions work then mean=median, but by mentioning median I start to think of other estimators, particularly non-parametric approaches to use medians.

p14L23-26: How did you define "best"? Maximise accuracy score, or what? Please specify.

p15L24-26: do LEO/GEO errors vary as you look at different parts of Antarctica? If you have a comment and/or reference here it would be nice. This is not vital though.

p19L15: I would delete "the summer month" and just say "in July". Avoids the boreal vs austral summer issue.

p19L25-26: "Therefore the method presented in this work provides a solution to polar cloud detection when infrared channels are not available"... I don't like the logic of this sentence. Your prior comment is that this approach is complementary to, and improves upon, infrared. So you don't have this *only* because of a lack of infrared channels. I'd change the last part to "...when infrared channels are not available, or struggle to distinguish between cloudy and clear scenes."

Figure 1c,d: the axis label still says dln, please change to dbln

General language comments:
There are a few grammatical article issues still. Mostly noted when standard nouns have adjectives on them, e.g. when "clouds" or "surfaces" are described. Changing snow/ice surface to "snow and ice surfaces" would fix a lot of these.

Other examples, I'll put <> to represent suggested deletion and [] for insertion:
p2L29: "<the> water and ice clouds" ("the" not necessary since "clouds" is plural)
p3L31: "...and [the] sea surface" (singular non-proper noun)
p4L1: "..at <the> 680 nm and 780 nm" (using 680 nm as a proper noun standing in for "the 680 nm channel")
p19L10: "[The] model derived algorithm" ("algorithm" singular)
p19L12: "...accuracy of [the] EPIC cloud mask" ("cloud mask" singular object)

Hide

ED: Publish subject to technical corrections (21 Feb 2020) by Piet Stammes

AR by Yaping Zhou on behalf of the Authors (26 Feb 2020) Author's response Manuscript

Short summary

Satellite cloud detection over snow and ice has been difficult for passive remote sensing instruments due to the lack of contrast between clouds and the bright and cold surfaces; the Earth Polychromatic Imaging Camera (EPIC) on board the Deep Space Climate Observatory (DSCOVR) has very limited channels. This study investigates the methodology of applying EPIC's two oxygen absorption band pair ratios for cloud detection over snow and ice surfaces.