the Creative Commons Attribution 4.0 License.
the Creative Commons Attribution 4.0 License.
| 02 Nov 2023
The Radiative and Geometric Properties of Melting First-Year Sea Ice
Abstract. In polar regions, sea ice is a crucial mediator of the interaction between earth's atmosphere and oceans. Its formation and breakup is intimately connected with local weather patterns and larger-scale climatic processes. During the spring melt and breakup period, snow-covered ice transitions to open water in a matter of weeks. This has a profound impact on the use of sea ice in coastal Arctic regions by Indigenous People, where activities such as hunting and fishing are central to community livelihood. In order to investigate the physical phenomena at the heart of this process, a set of targeted, intensive observations were made over Spring sea ice melt and breakup in Kotzebue Sound, Alaska. This program is part of the Ikaaġvik Sikukun project, a collaborative effort in which an Indigenous Elder advisory council from Kotzebue and scientists participated in co-production of hypotheses and observational research, including a stronger understanding of the physical properties of sea ice during spring melt. Data were collected using high-endurance, fixed-wing uncrewed aerial vehicles (UAVs) containing custom-built scientific payloads. Here we present the results of these measurements. Repeated flights over the measurement period captured the early stages of the transition from a white, snow-covered state to a broken up, bare/blue-green state. We found that the reflectance of sea ice features depend strongly on their size. Snow patches get darker as they get smaller, an effect owed to the geometric relationship between bright interior and the darker, melting feature edges. Conversely, bare patches get darker as they get larger. For the largest ice features observed, bare blue-green ice patches were found to be ~20 % less reflective than average, while large snowy/white ice patches were found to be ~20 % more reflective than average.
Status: final response (author comments only)
-
RC1: 'Comment on egusphere-2023-2541', Anonymous Referee #1, 27 Nov 2023
This manuscript presented a study on the radiative and geometric properties of sea ice observed by a fixed-wing UAV, which is interesting to Arctic sea ice society. The geometric properties part (Figures 9, 10) was something common study, which didn’t present novel information. The second part, the dependence of reflectance on ice size is indeed novel and could be useful as thinking about the impact of incorporating ice size into new sea ice models. However, I am confused by the motivation of this work. The problem that this manuscript seeks to solve needs to be clarified.
Major concerns:
Figures 11, 12, 13: It is reasonable that ice reflectance increases with the floe size, because of the lateral loss of light out the ice boundary. Furthermore, when the size is big enough (e.g. > 15 m), the reflectance is nearly identical. This result agrees with other simulated results. However, I don’t understand why the reflectance of ice with size was 20 and > 20 became smaller.
It seems that the manuscript didn’t present the observing time of UAV flights. The figure shows ice reflectance but not albedo. So, there is a question, whether the effects of solar zenith angle on ice bidirectional reflectance been considered?
I also didn’t see the view angle of the radiance sensor. At a flight altitude of 1000 m, its field of view may be over 40 m (assume the view angle is 5°). So, it is doubly that if the observed reflectance is from bare ice or ice and seawater. If it is the latter one, the word ‘size’ used here is misleading. The changing radiance is mostly due to the changing ice concentration but not size. Then, the novelty of this work disappeared.
Other concerns:
The abstract need re-write, and there are lots of introduction but less results.
There are too many abbreviations in the manuscript that weren’t defined in the main text. So, it is difficult to understand the figures.
L36-39: The Arctic amplification can’t be regarded as a total result of the changing ice situation.
Figure 3: It seems this figure didn’t show any useful information.
Section 3.1 (Figure 8): I don’t understand what this figure seeks to present. The difference in ΔEs was mostly controlled by the reflected radiance. It is a matter of course that mean or spectral reflectance decreases with increasing ΔEs.
Citation: https://doi.org/10.5194/egusphere-2023-2541-RC1 -
AC1: 'Reply on RC1', Nathan Laxague, 25 Jan 2024
(Reviewer's comments provided here as Italicized text)
This manuscript presented a study on the radiative and geometric properties of sea ice observed by a fixed-wing UAV, which is interesting to Arctic sea ice society. The geometric properties part (Figures 9, 10) was something common study, which didn’t present novel information. The second part, the dependence of reflectance on ice size is indeed novel and could be useful as thinking about the impact of incorporating ice size into new sea ice models. However, I am confused by the motivation of this work. The problem that this manuscript seeks to solve needs to be clarified.
We are grateful to the reviewer for their feedback. After reading the feedback from both reviewers (and the editor), we determined that we need to more effectively communicate the fact that observations were made over landfast sea ice- and that all of the feature spatial/geometric analysis was performed with respect to regions in/on that landfast ice, not floes within a marginal ice zone.
We have responded to each of the reviewer’s concerns on a point-by-point basis:
Major concerns:
Figures 11, 12, 13: It is reasonable that ice reflectance increases with the floe size, because of the lateral loss of light out the ice boundary. Furthermore, when the size is big enough (e.g. > 15 m), the reflectance is nearly identical. This result agrees with other simulated results. However, I don’t understand why the reflectance of ice with size was 20 and > 20 became smaller.
This was identified by both reviewers, and we believe it is an important point for discussion. As shown in our Figure 13, for bare ice, both small features and large features appear to be darker than features of moderate size. We do not have an explanation for 18 m being a lengthscale of particular significance. It is likely the case that the phenomena responsible for the relative feature darkness at small and large sizes differ from one another. For example, “small” features are often complex in shape (small area/perimeter ratio), whereas for large features, light is expected to transmit more freely through the bare ice without impediment by edge effects. This content will be added to the manuscript’s Discussion section.
It seems that the manuscript didn’t present the observing time of UAV flights. The figure shows ice reflectance but not albedo. So, there is a question, whether the effects of solar zenith angle on ice bidirectional reflectance been considered?
The approximate flight times are indicated in Figure 2. The exact times will be given in the text of section 2.1 along with the associated solar zenith angles.
I also didn’t see the view angle of the radiance sensor. At a flight altitude of 1000 m, its field of view may be over 40 m (assume the view angle is 5°). So, it is doubly that if the observed reflectance is from bare ice or ice and seawater. If it is the latter one, the word ‘size’ used here is misleading. The changing radiance is mostly due to the changing ice concentration but not size. Then, the novelty of this work disappeared.
The imaging radiance sensor view angle (both full swath and pixel IFOV) is provided in Table 1. As mentioned in the surrounding text, the ground sample distance is approximately 50 cm at a flight altitude of 1000 m.
Regarding the possibility of including seawater in our observations- open water is not included in analysis, with a rejection criterion specified in L125.
Other concerns:
The abstract need re-write, and there are lots of introduction but less results.
We agree that the abstract is heavily laden with introductory material. We will adjust the ratio between background and results in the abstract. In order to avoid over-burdening the abstract, we propose to condense the introductory material, keeping the five lines of results as-is.
There are too many abbreviations in the manuscript that weren’t defined in the main text. So, it is difficult to understand the figures.
The VNIR and RAD payloads were not explicitly defined in the text, but will be at the beginning of section 2.1. Furthermore, mentions of GPS (US-based constellation) will be replaced with GNSS (the general term), with the latter acronym defined in section 2.1 in the vicinity of L95.
L36-39: The Arctic amplification can’t be regarded as a total result of the changing ice situation.
We will re-word this sentence, stating that the changing ice situation "contributes to..." Arctic amplification.
Figure 3: It seems this figure didn’t show any useful information.
Figure 3 was placed there to demonstrate that the marine-atmospheric boundary layer was changing on top of the changing sea ice state. It was intended as a (vertical profile) accompaniment to the time series of Figure 2.
Section 3.1 (Figure 8): I don’t understand what this figure seeks to present. The difference in ΔEs was mostly controlled by the reflected radiance. It is a matter of course that mean or spectral reflectance decreases with increasing ΔEs.
The left panel provides reader with a visual depiction of the sea ice color and shows the range of conditions (in reflectance and delta E_s) observed. Right panel provides spectral reflectance binned by net shortwave irradiance; reader can connect level and shape of spectra with marker color/darkness in panel A. As pointed out by the reviewer, the relationships between these quantities are a matter of course. We included this figure not for the purpose of presenting novel results, but to provide background and context. Perhaps the figure could be moved earlier in the manuscript?
Also- we note that the ordinate of panel (a) is properly labeled (rho_s, sr^-1), but the ordinate of panel (b) should be labeled as rho_s(lambda), sr^-1 nm^-1.
Citation: https://doi.org/10.5194/egusphere-2023-2541-AC1
-
AC1: 'Reply on RC1', Nathan Laxague, 25 Jan 2024
-
RC2: 'Comment on egusphere-2023-2541', Anonymous Referee #2, 02 Dec 2023
Review of “The Radiative and Geometric Properties of Melting First-Year Sea Ice” This manuscript presents observational data from a field campaign deploying UAV with payloads designed to characterize the radiative properties of the sea ice surface. Data were collected over landfast Arctic sea ice during the melt season as the ice transitioned from snow covered to bare and melting. During this time the floe size distribution was rapidly evolving from larger floes to smaller floes.
The text is written very clearly and concisely. The figures are clear and informative. There is a lot of information here and the results are of interest to the sea ice radiative transfer modeling community.
I find the conclusions intriguing, but would be informative to have more explanation of the effects that were documented. For example, I can think of a variety of explanations for the results showing sensitivity of reflectivity to feature size:
- Smaller features are more influenced by their perimeter, perimeters are subject to some strong effects that can alter their reflectivity, such as lateral melting, wave wash.
- Light leakage: do photons propagate laterally out of floe boundary before they can be backscattered? Likewise, do photons enter the ice from the side and increase the albedo?
It would be helpful to the reader to consider these mechanisms (and there may be more?) and either substantiate them or refute them. I think it might not be that the larger floes are darker, but rather that the smaller floes get light leaked in from the sides?
I do agree that the reflectivity is reduced when light propagation out the sides becomes significant, so, yes less light backscattered to atmosphere, but does that mean that these snowy features are absorbing more solar radiation? The surface is absorbing more radiation, but I wouldn’t say that the ice is.
Am I interpreting correctly that 18 m is the size scale that roughly divides whether a floe is on the “large” or “small” size? From a radiative transport perspective, that “boundary” sounds very large to me.
I am a bit surprised that dust and sediment were mentioned on a couple of occasions, but not explicitly considered in the explanation. Further, not much was said about the green appearance of the ice cover. Is this a result of biologic activity? Or dust/sediment?
There are a lot of interesting results and intriguing discussions in this manuscript. I feel that the overall story of how the pieces fit together could be tightened up.
Minor comments:
Title: Add words “Arctic” and “landfast”
Line 3: matter of weeks? Sure, some areas undergo the transition from snow covered sea ice to open water in a matter of weeks, but that process takes a lot longer in other regions, depends on location and ice type.
14-15: “than average” than average of what? All features?
40: “Arctic system” less resilient to change? Of should this say “Arctic ice cover” is less resilient?
Fig 1: this figure shows the color of the ice, but why choose the same color for the mask?
160: estimation, retrieval of feature length scale hasn’t been sufficiently motivated, why should it be measured? Section 2.2 is ‘how’, but there is no ‘why’
Fig 11: “thickness of each trace corresponds to the mean effective feature diameter…” not clear how “thickness” is being used here. all look same thickness (width?) to me. Or does ‘thickness’ refer to “D” and hence the gray/yellow shade?
329-331: “…to strengthen positive feedbacks associated with radiation uptake. In short: the tendency of large blue-green features to absorb radiation increases with their size, while large snowy features absorb more solar radiation as they are subdivided and split by melt and degradation.” I don’t understand the mechanism whereby an increase in absorption with increasing size is a positive albedo feedback. That would suggest that as the floe size decreased (due to melt, increased absorption), the absorption would also decrease (which sounds like a negative feedback)
Citation: https://doi.org/10.5194/egusphere-2023-2541-RC2 -
AC2: 'Reply on RC2', Nathan Laxague, 25 Jan 2024
(Reviewer's comments provided here as Italicized text)
Review of “The Radiative and Geometric Properties of Melting First-Year Sea Ice” This manuscript presents observational data from a field campaign deploying UAV with payloads designed to characterize the radiative properties of the sea ice surface. Data were collected over landfast Arctic sea ice during the melt season as the ice transitioned from snow covered to bare and melting. During this time the floe size distribution was rapidly evolving from larger floes to smaller floes.
The text is written very clearly and concisely. The figures are clear and informative. There is a lot of information here and the results are of interest to the sea ice radiative transfer modeling community.
We are grateful to the reviewer for their feedback. We have responded to each of the reviewer’s concerns on a point-by-point basis:
I find the conclusions intriguing, but would be informative to have more explanation of the effects that were documented. For example, I can think of a variety of explanations for the results showing sensitivity of reflectivity to feature size:
- Smaller features are more influenced by their perimeter, perimeters are subject to some strong effects that can alter their reflectivity, such as lateral melting, wave wash.
- Light leakage: do photons propagate laterally out of floe boundary before they can be backscattered? Likewise, do photons enter the ice from the side and increase the albedo?
It would be helpful to the reader to consider these mechanisms (and there may be more?) and either substantiate them or refute them. I think it might not be that the larger floes are darker, but rather that the smaller floes get light leaked in from the sides?
We need to more effectively communicate the fact that observations were made over landfast sea ice- and that all of the feature spatial/geometric analysis was performed with respect to regions in/on that landfast ice, not floes within a marginal ice zone.
Am I interpreting correctly that 18 m is the size scale that roughly divides whether a floe is on the “large” or “small” size? From a radiative transport perspective, that “boundary” sounds very large to me.
As shown in our Figure 13, for bare ice, both small features and large features appear to be darker than features of moderate size. We do not have an explanation for 18 m being a lengthscale of particular significance. It is likely the case that the phenomena responsible for the relative feature darkness at small and large sizes differ from one another. For example, “small D” features are often not small in area, but complex in shape (small area/perimeter ratio); however, for large features, transmission through the ice is expected to be substantial. This content will be added to the manuscript’s Discussion section.
I am a bit surprised that dust and sediment were mentioned on a couple of occasions, but not explicitly considered in the explanation. Further, not much was said about the green appearance of the ice cover. Is this a result of biologic activity? Or dust/sediment?
Within the Discussion section (L219-224), we posit that the presence of sediment may have been partially responsible for the rapid degradation of the sea ice in Kotzebue Sound during our study period. We will bolster this discussion point: spatially-averaged values of sea ice color (Figure 8a) show that brown/dark grey patches were ubiquitous, corroborating our eyewitness observation.
Regarding the blue/green appearance of the bare ice: this may be a consequence of the thin ice present during our field operations (Witte et al., 2021; Mahoney et al., 2021). In these conditions, light may transmit through the ice and reflect off the bottom. We also suspect that CDOM concentrations were moderate to high in the coastal waters of Kotzebue Sound. However, it is difficult to say whether or not these effects (sediment, ice thickness, CDOM) had bearing on the scale-dependent reflectance behavior observed by our airborne instrumentation.
There are a lot of interesting results and intriguing discussions in this manuscript. I feel that the overall story of how the pieces fit together could be tightened up.
We are grateful for the reviewer’s suggestions for improving the manuscript. We feel that the expanded discussion points mentioned here will help to tie some of the disparate components together.
Minor comments:
Title: add words "Arctic" and "landfast"
This will be done.
Line 3: matter of weeks? Sure, some areas undergo the transition from snow covered sea ice to open water in a matter of weeks, but that process takes a lot longer in other regions, depends on location and ice type.
We will rephrase this line to communicate that the melt and breakup process is location and ice type-dependent.
14-15: “than average” than average of what? All features?
We need to clarify this within the text. Yes, compared to the average reflectance observed across all observational cases.
40: “Arctic system” less resilient to change? Of should this say “Arctic ice cover” is less resilient?
We will re-word as "Arctic ice cover is less resilient"
Fig 1: this figure shows the color of the ice, but why choose the same color for the mask?
We will change the land color to a sandy (light brown) color
160: estimation, retrieval of feature length scale hasn’t been sufficiently motivated, why should it be measured? Section 2.2 is ‘how’, but there is no ‘why’
We recognize that this component of the study has not been adequately justified in the present text. In the introduction (~L160), we will refer to Popovic et al. [2018] (specifically, in reference to their language “These results demonstrate that the geometry and abundance of Arctic melt ponds can be simply described, which can be exploited in future models of Arctic melt ponds that would improve predictions of the response of sea ice to Arctic warming.”) and Horvat et al. [2020] (specifically, in reference to their language “We find that aggregate properties of the instantaneous sub-ice light field, such as the enhancement of available solar energy under bare ice regions, can be described using a new parameter closely related to pond fractal geometry.”).
Fig 11: “thickness of each trace corresponds to the mean effective feature diameter…” not clear how “thickness” is being used here. all look same thickness (width?) to me. Or does ‘thickness’ refer to “D” and hence the gray/yellow shade?
That legend corresponds to a previous version of figure; we will update the legend to properly match the figure.
Citation: https://doi.org/10.5194/egusphere-2023-2541-AC2
Data sets
Codes and Data for "The Radiative and Geometric Properties of Melting First-Year Sea Ice", pre-submission Nathan J. M. Laxague and Christopher J. Zappa https://doi.org/10.7916/rrbv-k026
Viewed
Since the preprint corresponding to this journal article was posted outside of Copernicus Publications, the preprint-related metrics are limited to HTML views.
| HTML | XML | Total | BibTeX | EndNote | |
|---|---|---|---|---|---|
| 138 | 0 | 0 | 138 | 0 | 0 |
- HTML: 138
- PDF: 0
- XML: 0
- Total: 138
- BibTeX: 0
- EndNote: 0
Viewed (geographical distribution)
Since the preprint corresponding to this journal article was posted outside of Copernicus Publications, the preprint-related metrics are limited to HTML views.
| Country | # | Views | % |
|---|
| Total: | 0 |
| HTML: | 0 |
| PDF: | 0 |
| XML: | 0 |
- 1