the Creative Commons Attribution 4.0 License.
the Creative Commons Attribution 4.0 License.
Retrieval of snow layer and melt pond properties on Arctic sea ice from airborne imaging spectrometer observations
Sophie Rosenburg
Charlotte Lange
Evelyn Jäkel
Michael Schäfer
André Ehrlich
Manfred Wendisch
Abstract. A melting snow layer on Arctic sea ice, as a composition of ice, liquid water, and air, supplies meltwater that may trigger the formation of melt ponds. As a result, surface reflection properties are altered during the melting season and thereby may change the surface energy budget. To study these processes, sea ice surface reflection properties were derived from airborne measurements using imaging spectrometers. The data were collected over the closed and marginal Arctic sea ice zone north of Svalbard in May/June 2017. A retrieval approach based on different absorption indices of pure ice and liquid water in the near-infrared spectral range was applied to the campaign data. The technique enables to retrieve the spatial distribution of the liquid water fraction of a snow layer and the effective radius of snow grains. For observations from three research flights liquid water fractions between 8.7 % and 15.6 % and snow grain sizes between 115 μm and 378 μm were derived. In addition, the melt pond depth was retrieved based on an existing approach that isolates the dependence of a melt pond reflectance spectrum on the pond depth by eliminating the reflection contribution of the pond ice bottom. The application of the approach to several case studies revealed a high variability of melt pond depth with maximum depths of 0.33 m. The results were discussed considering uncertainties arising from the reflectance measurements, the setup of radiative transfer simulations, and the retrieval method itself. Overall, the presented retrieval methods show the potential and the limitations of airborne measurements with imaging spectrometers to map the transition phase of the Arctic sea ice surface, examining the snow layer composition and melt pond depth.
Sophie Rosenburg et al.
Status: open (until 09 Jun 2023)
-
RC1: 'Comment on amt-2023-64', Anonymous Referee #1, 19 May 2023
reply
I have only one critical comment on the manuscript:
The statements "the stronger a non-complete representation of the phase function will influence the simulated reflectance spectra" and "[the] sensitivity of the extraction method to the phase function" is one of the "strongest sources of uncertainty" are incorrect. In the case of multiple scattering, the details of the single scattering phase function are not important and in problems such as the one considered it is sufficient to use the so-called transport approximation with the correct value of the asymmetry factor of scattering
I would like to recommend the following literature on this subject:
- L.A. Dombrovsky, The use of transport approximation and diffusion-based models in radiative transfer calculations, Computational Thermal Sci. 4 (4) (2012) 297–315. http://doi.org/10.1615/ComputThermalScien.2012005050
- L.A. Dombrovsky and A.A. Kokhanovsky, Solar heating of the cryosphere: Snow and ice sheets, Ch. 2 in the book “Springer Series in Light Scattering”, edited by A. Kokhanovsky, Springer Nature, 2021, v. 6, pp 53-109. https://doi.org/10.1007/978-3-030-71254-9_2
Please consider including these articles in the reference list.
Citation: https://doi.org/10.5194/amt-2023-64-RC1 -
RC2: 'Comment on amt-2023-64', Christopher Donahue, 19 May 2023
reply
The article demonstrates airborne imaging spectroscopy retrievals over arctic sea ice. The retrievals include snow grain size, liquid water fraction, and ice pond depth. Generally, the methodology and retrievals presented are of interest to the cryosphere remote sensing community. However, there are several major items that need to be addressed before publication, listed here in no particular order:
- Throughout the paper ambiguous remote sensing terminology is used, e.g., “reflectance”. Refer to Schaepman-Strub et al (2006) and use correct terminology throughout. For example, the airborne spectrometer measures HDRF/HCRF while the paired albedometers measure BHR. Additionally, more details about the radiative transfer simulations need to be provided. Are the simulations representative of white or black sky albedo or HDRF? How is the illumination and observation angle geometry being handled in the radiative transfer modeling?
- There are 2 grain size retrieval methods presented (figure 3 and 4), however only results from one of the methods (figure 4) is presented. I am confused by this, where are grain size results from the methodology presented in Figure 3? Otherwise, the least-square fitting methodology for grain size should be removed from the manuscript.
- The grain size/LWC methodology presented is based on Green et al 2002, however the LWC and grain size retrievals are decoupled instead of retrieved simultaneously. Some discussion around the reasoning for this and its implications should be included. For example, the least squares methodology for grain size (using Parts 1-3) seems to include radiative transfer simulations that are based on wet and dry snow, though only the grain size is being retrieved. What uncertainties does this introduce? Would it be possible find a least square spectra that was simulated using a 200 µm grain size and 10% water and alternatively be very similar to 250 µm and 0% water?
- Line 64 – 68: This is somewhat of an overstatement. All three properties are not combined in a single flight line, though if this is possible and demonstrated it would greatly enhance the paper results. Furthermore, the retrievals individually (grain size and water content) have been demonstrated using airborne imaging spectroscopy (e.g., Bohn et al. 2021).
- Consider reorganizing the paper such that the methods are all presented together followed by the results. The current organization is hard to follow as a reader because it jumps around from methods to results. Additionally, the paper would greatly benefit from further editing to make the paper more concise and flow better.
Specific comments
Line 39-40 is the grain “size” a radius of diameter, should be defined?
Line 49: Green tested outside using a block of snow under natural solar conditions, not in a laboratory.
Line 54-56: Unclear, sentence needs to be restated.
Line 63: “In this study” is repetitive since the paragraph started with these 2 sentences prior.
Table 1: Under SMART should there be an F (upward arrow) as well? Line 81-82 states that it measures upward and downward looking irradiance.
Line 103: Example of ambiguous usage of “reflectance”. More details on the simulations should be included. How was the observation and illumination angle geometry represented in the modeling?
Line 139: How was this smoothing done?
Line 152: observation and illumination conditions?
Line 158-161: Does this normalization change the absolute magnitude of the reflectance spectrum? How might this impact the grain size retrieval? This should be mentioned in the discussion.
Figure 3. This figure could be improved. Consider showing a few selected simulations that include the full spectra and one or two measurements that match. You can highlight the regions (Part 1,2,3) that are used for the retrieval with vertical shading bars.
Figure 5. (a) Why is the RBG true color image green? I would expect it to be white.
Figure 5. A scale bar and north arrow should be added to retrieval maps.
Line 185-187: Consider moving to discussion section.
Line 200-202 and Table 2: Generally, the grain sizes retrieved seem low for water contents that are mostly >10%. For example, Green 2002 retrieved 10% LWC and grain radius of 550 um. Consider adding some discussion around this topic. How might the grain shape chosen influence the grain size retrieval. Further, How does decoupling the grain size and LWC retrieval effect the grain size retrieval?
Figure 7. It is hard to discern the difference between 30% and 100%. Consider removing 100% water fraction from the scale bar or making 100% a separate color and call it “open water”, or something similar.
Citation: https://doi.org/10.5194/amt-2023-64-RC2 -
RC3: 'Comment on amt-2023-64', Anonymous Referee #3, 25 May 2023
reply
The paper presents methods to retrieve a) effective radius of snow grains and liquid water fraction of a snow layer and b) melt pond depth during the first phase of the Arctic melt season by using airborne measurements with imaging spectrometers.
The data base consisting of three analysed flights is quite limited. However, a key aspect of the paper is the further development or modification of existing retrieval methods based on the use of additional instruments (e.g., the SMART albedometer) compared to previous studies.
Since retrieval of melt pond depth from airborne measurements is still rare, to my mind, the modified use of the model by König and Oppelt (2020) is most useful.
A weak point is the missing in situ data from ground-based measurements, a fact that prevents a thorough validation of the results.
However, the paper addresses very relevant questions within the scope of AMT. Substantial conclusions are reached. Scientific and technical methods are clearly outlined and the results are sufficient to support the conclusions.
The title reflects the content of the paper, the abstract provides a complete summary and the paper is generally well structured. The review of existing published work is very good, the number of references is appropriate. Overall, figures and tables are clear and their captions self-explanatory. Mathematical formulae, symbols and abbreviations are correctly defined and used. The use of the English language is very good.
Citation: https://doi.org/10.5194/amt-2023-64-RC3
Sophie Rosenburg et al.
Sophie Rosenburg et al.
Viewed
| HTML | XML | Total | BibTeX | EndNote | |
|---|---|---|---|---|---|
| 231 | 49 | 12 | 292 | 2 | 2 |
- HTML: 231
- PDF: 49
- XML: 12
- Total: 292
- BibTeX: 2
- EndNote: 2
Viewed (geographical distribution)
| Country | # | Views | % |
|---|
| Total: | 0 |
| HTML: | 0 |
| PDF: | 0 |
| XML: | 0 |
- 1