the Creative Commons Attribution 4.0 License.
the Creative Commons Attribution 4.0 License.
Quantifying the climatic impact of crude oil pollution on sea ice albedo
- Department of Earth Sciences, Royal Holloway, University of London, Egham, Surrey, TW20 0EX, UK
- Department of Earth Sciences, Royal Holloway, University of London, Egham, Surrey, TW20 0EX, UK
Abstract. Sea ice albedo is an important component of the Earth’s climate and is affected by low background concentrations of oil droplets within the ice matrix that absorb solar radiation. In this study the albedo response of three different types of sea ice (multi-year, first-year, and melting sea ice) are calculated at increasing mass ratios (0–1000 ng g−1) of crude oil by using a coupled atmosphere-sea ice radiative-transfer model (TUV-snow) over the optical wavelengths 400–700 nm. The different types of quasi-infinite thickness sea ice exhibit different albedo responses to oil pollution, with a 1000 ng g−1 mass ratio of oil causing a decrease to 70.9 % in multi-year sea ice, 47.9 % in first-year sea ice, and 22 % in melting sea ice relative to the unpolluted albedo at a wavelength of 400 nm. The thickness of the sea ice is also an important factor, with realistic thickness sea ices exhibiting similar results, albeit with a weaker albedo response for multi-year sea ice to 75.3 %, first-year sea ice to 66.7 %, and melting sea ice to 35.7 %. The type of oil also plays a significant role on the response of sea ice albedo, with a relatively opaque and heavy crude oil (Romashkino oil) causing a significantly larger decrease in sea ice albedo than a relatively transparent light crude oil (Petrobaltic oil). The size of the oil droplets polluting the oil also plays a minor role in the albedo response, with weathered submicron droplets (0.05–0.5 µm radius) of Romashkino oil being the most absorbing across the optical wavelengths considered. Therefore, the work presented here demonstrates that low background concentrations of small submicron to micron-sized oil droplets have a significant effect on sea ice albedo. All three types of sea ice are affected, however first-year sea ice and particularly melting sea ice are very sensitive to oil pollution; thus, the Arctic may become more vulnerable to oil pollution as the ice becomes progressively thinner and younger in response to a changing climate.
Benjamin Heikki Redmond Roche and Martin D. King
Status: final response (author comments only)
-
RC1: 'Comment on tc-2021-372', Anonymous Referee #1, 02 Mar 2022
This manuscript addresses the effects of included droplets of oil on the albedo of sea ice. The authors have assembled data on the inherent optical properties of two types of crude oil and used these values, along with information about the optical properties of three types of sea ice (multiyear, first-year, melting) to assess the effects of included droplets of oil on the spectral albedo of the ice. The authors assess these effects as functions of loading, ice type, oil droplet size, background carbon loading, and ice thickness.
The text is very easy to read as the writing is crisp and organized. However, I have concerns about the substance of this manuscript and its conclusions. This manuscript describes a carefully executed sensitivity study and I find no fault here in the mechanics of the sensitivities assessed. However, these studies offer little new physical insight or understanding. I find the title somewhat misleading, as the authors really have not quantified the climatic impact of oil pollution—that would be a much more far reaching study—likely involving a climate model. My more significant concern however is the conclusions that are sensitive to the inherent optical properties chosen for the three ice types. I don’t believe that the optical descriptions given here for first-year, multiyear, and melting sea ice are realistic or representative in the context of this study. It is well established that the scattering coefficients for sea ice display significant variability, including between ice types, within a single ice column, and for the same ice type (e.g., FY, MY, melting) at different times and locations. This variability poses problems for the types of conclusions stated in this manuscript.
In fact, I’m a bit confused by the FY, MY, melting classifications. If Arctic sea ice isn’t melting, it is likely snow covered. Do the authors intend for this study to treat bare, non-melting FY, MY ice? And, if the snow is implicitly included, then the dynamics of snow-oil interactions need to be accounted for. I see that snow cover is mentioned at the end of the manuscript as an uncertainty, but it’s not clear that it is being accounted for in the optical properties listed in Table 1. I don’t think the corresponding clean ice albedos generated from these properties are representative of ice conditions that are commonly observed in nature.
183: how do authors justify 201 layers?
200 (Eq1): What is the prime symbol for?
212: why bother modeling the atmosphere here? Seems extraneous.
299: “The oil is most absorbing at 400 nm, where ice is the least absorbing…” Perhaps this is true, but this is the cause, not a result.
313: “the effect of oil significantly decreases as wavelength increases over the region studied and the oil becomes less absorbing whilst the ice becomes more absorbing.” Same as previous comment, this is a physical cause, not a result.
328: “The three types of sea ice have different unpolluted albedos: melting sea ice 0.72, first-year sea ice 0.87, and multi-year sea ice 0.94 at a wavelength of 400 nm respectively, owing to their different scattering cross sections (Perovich, 1996; Marks and King, 2014).” I think this overstates the differences between these three ice types if indeed it is intended that all are snow free.
606: “As these data shows, this decline in perennial types of sea ice renders the Arctic much more vulnerable to increased oil pollution in the region...” I don’t think this conclusion is supported by this study. For example, the high scattering prevalent in the surface layers of multiyear ice, and the larger thickness of this layer in thicker ice is not accounted for in this study. Also, there is no attempt to simulate how oil droplets respond to summer freshwater flushing that is a key factor that distinguishes FY ice from MY ice.
621: “Therefore, it appears that the type of oil has the biggest effect on how responsive sea ice is to increasing mass ratios of oil as opposed to the type of sea ice and in contrast with the findings of a comparable study into the effects of mineral dust on sea ice albedo (Lamare et al., 2016).” I don’t understand what this means. Is it saying that there is larger variability in oil inherent optical properties than in the optical properties of mineral dust? That may be so, but it is not a result or a conclusion.
627: “First-year and particularly melting sea ice are more responsive to oil pollution than multi-year sea ice, so these trends indicate that sea ice albedo in the Arctic may become more vulnerable to background levels of oil 630 pollution as the ice becomes progressively thinner and younger.” I find this conclusion unsubstantiated, because I don’t think differences in the optical properties of FY / MY ice types is treated in a realistic way here. This simplification may be well justified for the purposes of a sensitivity study such as carried out here, but I think it’s a stretch to draw conclusions such as stated here from this type of sensitivity exercise.
- AC1: 'Reply on RC1', Ben Redmond Roche, 08 Apr 2022
-
RC2: 'Comment on tc-2021-372', Anonymous Referee #2, 10 Apr 2022
This manuscript provides a theoretical / modelling study of the impact of crude oil droplets on sea ice albedo. The authors use a coupled atmosphere-snow/sea ice radiative transfer model to simulate the wavelength-dependent albedo of three common types of sea ice in response to increasing mass ratios of two different types of crude oil, whilst accounting for different optical and physical parameters. The main findings of the article are that: 1) sea ice properties play an important role when considering the impact of oil pollution, 2) the thickness of sea ice is an important factor to be considered, 3) the type of crude oil dispersed in the sea ice also significantly affects the response of sea ice albedo, and 4) the size of the oil droplets polluting the oil only plays a small role in albedo response.
The manuscript appears to be a “logical” continuation of the previous works of the research group who used a similar modelling approach to characterise the effects of black carbon [Marks et al. 2014] and mineral dust deposits [Lamare et al., 2016] on the albedo of sea ice. The article is clearly written, well structured and the results shown in the figures are easy to understand. Although the findings are not ground-breaking, the authors provide valuable insights into the impacts of increased pollution on the albedo of sea ice, which is a timely subject to be addressed in the current Arctic context. Moreover, the parametrisations used in this study are of value to the scientific community, for further modelling studies. For these reasons, I recommend this manuscript for publication, after the authors have address a few minor comments stated below.
General comments
There is a misalignment between the title of the article and certain statements (see detailed comments below) and the actual purpose of the proposed work. Indeed, the climatic impact of crude oil pollution is not addressed in this sensitivity study. I would recommend that the authors rephrase the title of the manuscript and correct the statement line 641.
The impacts of crude oil droplets dispersed in sea ice are investigated in the visible wavelengths (400-700 nm). Could the authors please expand on the reasons for this wavelength range? Is it a limitation of the model or a deliberate choice? Most observation tools (ground-based instruments, drone or aircraft mounted sensors, Earth Observation satellite) cover a larger spectrum, generally from the visible to the shortwave infra-red. Furthermore, climate models (e.g. CMIP models) consider the longwave radiative balance to monitor the Earth’s energy balance. Although longwave radiation is most likely out of scope for this study, it is surprising that such a small range of wavelengths is being considered here. Extending the range to 2500 nm would allow direct comparisons with observations.
The authors state that “the effects that oil pollution has upon sea ice albedo have not previously been considered in literature”. Although this statement holds, the authors disregard the existing corpus of works that investigate the effects of oil pollution on sea ice reflectance (e.g. 1-5 in the reference section below). Despite the quantities being different, albedo can be derived from reflectance using a BRDF model, and it is widely accepted that reflectance may be used as an approximation for albedo. A short review of the existing studies would be desirable in the introduction.
In this study the authors have chosen to distribute the oil evenly throughout the sea ice. While this may be realistic in certain conditions (particularly for low oil concentrations), how plausible is this to occur at higher loadings (1000 ng g-1)? In the discussion (line 496) the authors describe the different scenarios of how oil entrains itself into sea ice. From these comments, it is clear that layering of the oil is a common situation encountered in sea ice. The model used allows the definition of layers throughout the ice pack: rather than address the relationship between black carbon and oil loadings, would it not have been of value to consider the effects of oil located in specific (e.g. surface, or sub-surface) layers?
In link to the paragraph above, the concentrations of the oil would deserve more clarifications. Indeed, the article focusses on “microscopic sized background concentrations of oil” (line 509) but in the introduction it is mentioned that after the Deepwater Horizon incident, mass ratios of 100 ng g-1 were found. Can mass ratios of 1000 ng g-1 still be considered as “background”?
More information about the relationship between the oil droplet size and the mass ratios would be important to better understand which scenarios in the paper are most plausible. Have the authors investigated if there is a relationship between droplet size and mass ratios between 1 and 1000 ng g-1 or is it likely to find all sizes within the loading range?
Section 4.6 on the implications of the study is quite light and could be fleshed out more. It would be insightful to read the authors thoughts on the implications in terms of how the sea ice melting rates and extent in summer will be affected by oil pollution. How does the increased oil pollution impact the energy balance of the Arctic, and are the effects sufficient to be considered in General Climate Models?
Lastly, as a side note, I would suggest that the authors make the input data available through an open repository, which would benefit the modelling community greatly (e.g. use of the parametrisations of crude oil in climate models) and allow for further inter-comparisons of modelling approaches.
Detailed comments
l46: What do the authors mean by: “The wavelength integrated and spectral albedos for different types of sea ice have previously been considered [...]: this study focuses on three types of sea ice: melting, first-year, and multi-year sea ice.” It is not clear if the authors are referring to the literature or if they are stating that they have considered a wide variety of sea ice types before settling for the three mentioned.
Table 1: One would expect the density of first-year, multi-year and melting sea ice to be different owing to differences in structure (brine channels, air bubbles...). The reference cited by the authors [Marks and King, 2014] states that the density of sea ice ranges 700–950 kg m-3. How do the authors justify the same fixed value for all sea ice types?
Table 2: On what basis were the number of layers (201) and the increments chosen for the model?
l236: In this paragraph the authors describe the optical properties of the two types of crude oil used in the study. Although it is stated that “Whilst both crude oils have a variety of uses, including as marine engine fuels, Romashkino can be considered a typical marine engine Heavy Fuel Oil”, the reasons for selecting Romashkino and Petrobaltic oil is not sufficiently clear to the reader, who has to wait until line 578 to understand that “both Romashkino and Petrobaltic can be regarded as the upper and lower respective brackets of the effect that oil pollution can have on sea ice albedo”. Furthermore, it would be useful to understand if these oil types are only representative of pollution that may occur from shipping activities, or can also be used to understand the impacts of oil spills from drilling activities.
l275: There seems to be a repetition in the first and second sentences. In the second sentence, the author state the same elements as in the first sentence but with melting and multi-year sea ice in addition. Please fix or clarify.
l279: In the sentence concerning the effect of oil droplet size, it would be useful to explicitly state the sea ice type considered.
Figure 1: I suggest using a Y axis ranging from 0.3 to 0.9, and putting the legend outside the figure for more clarity, if this is allowed by editing rules.
Section 3.3 Could the authors specify why melting sea ice was chosen for the analysis of the effects of oil droplet size on albedo? Are the implications similar for other types of sea ice?
Section 3.4: A reference to ΔA/Δm used in Figure 5 is expected in the text.
l627: “Arctic multi-year and first-year sea ice are declining at 17.5% and 13.5% respectively...” is not clear. Please rephrase.
l641: “[...] this is the first instance that the climatic effect of oil pollution on sea ice has been considered.” This sentence is misleading and implies the use of a climate model or conclusions on the large scale impact of oil pollution in the Arctic which is not the case here. Please rephrase.
l659: “[...] the findings of this study may only be valid during the ablation season when snow cover has melted or been removed by wind.” In this case why consider different types of sea ice? I believe the value of this paper lies in the sensitivity study considering a variety of optical and physical parameters. I suggest to add that this may be the case in practise and that the authors restate the main purpose of the study..
References
- [1] Ivanov, B., Bezgreshnov, A., Kubyshkin, N., & Kursheva, A. (2005). Spreading of oil products in sea ice and their influence on the radiation properties of the snow-ice cover. In 18th International Conference on Port and Ocean Engineering under Arctic Conditions (pp. 853-862). (Proceedings of the International Conference on Port and Ocean Engineering under Arctic Conditions, POAC).
- [2] Liu, B.; Li, Y.; Liu, C.; Xie, F.; Muller, J.-P. Hyperspectral Features of Oil-Polluted Sea Ice and the Response to the Contamination Area Fraction. Sensors 2018, 18, 234. https://doi.org/10.3390/s18010234
- [3] Bingxin Liu, Ying Li, Qiang Zhang, Liang Han, "Assessing Sensitivity of Hyperspectral Sensor to Detect Oils with Sea Ice", Journal of Spectroscopy, vol. 2016, Article ID 6584314, 9 pages, 2016. https://doi.org/10.1155/2016/6584314
- [4] Praks, M. Eskelinen, J. Pulliainen, T. Pyhalahti and M. Hallikainen, "Detection of oil pollution on sea ice with airborne and spaceborne spectrometer," IGARSS 2004. 2004 IEEE International Geoscience and Remote Sensing Symposium, 2004, pp. 276, doi: 10.1109/IGARSS.2004.1369014.
- [5] Chao, J., Liu, C., Li, Y. et al. Characteristics of the sea ice reflectance spectrum polluted by oil spills based on field experiments in the Bohai Sea. Acta Oceanol. Sin. 36, 73–79 (2017).
-
AC2: 'Reply on RC2', Ben Redmond Roche, 05 May 2022
The comment was uploaded in the form of a supplement: https://tc.copernicus.org/preprints/tc-2021-372/tc-2021-372-AC2-supplement.pdf
Benjamin Heikki Redmond Roche and Martin D. King
Benjamin Heikki Redmond Roche and Martin D. King
Viewed
| HTML | XML | Total | BibTeX | EndNote | |
|---|---|---|---|---|---|
| 414 | 143 | 20 | 577 | 10 | 8 |
- HTML: 414
- PDF: 143
- XML: 20
- Total: 577
- BibTeX: 10
- EndNote: 8
Viewed (geographical distribution)
| Country | # | Views | % |
|---|
| Total: | 0 |
| HTML: | 0 |
| PDF: | 0 |
| XML: | 0 |
- 1
