the Creative Commons Attribution 4.0 License.
the Creative Commons Attribution 4.0 License.
| 08 Mar 2024
How Many Parameters are Needed to Represent Polar Sea Ice Surface Patterns and Heterogeneity?
Abstract. Sea ice surface patterns encode more information than can be represented solely by the ice fraction. The aim of this paper is thus to establish the importance of using a broader set of surface characterization metrics, and to identify a minimal set of such metrics that may be useful for representing sea-ice in Earth System Models. Large-eddy simulations of the atmospheric boundary layer over various idealized sea ice surface patterns, with equivalent ice fraction and average floe area, demonstrate that the spatial organization of ice and water can play a crucial role in determining boundary-layer structure. Thus, different methods to quantify heterogeneity in categorical lattice spatial data, such as those done in landscape ecology and Geographic Information System (GIS) studies, are used here on a set of high-resolution, recently-declassified sea ice surface images. It is found that, in conjunction with ice fraction, the patch density (representing the fragmentation of the surface), the splitting index (representing the variability in patch size), and perimeter-area fractal dimension (representing the tortuosity of the interface) are all required to describe the two-dimensional pattern exhibited by a sea ice surface. Furthermore, for surfaces with anisotropic patterns, the orientation of the surface relative to the mean wind is needed. Furthermore, scaling laws are derived for these relevant landscape metrics to estimate them from aggregated spatial sea ice surface data at any resolution. The methods used and results gained from this study are a first step towards further development of methods to quantify the variability of non-binary surfaces, and for parameterizing mixed ice-water surfaces in coarse geophysical models.
- Preprint
(1777 KB) - Metadata XML
- BibTeX
- EndNote
Status: final response (author comments only)
- RC1: 'Comment on egusphere-2024-532', Ian Brooks, 21 Mar 2024
-
RC2: 'Comment on egusphere-2024-532', Christof Lüpkes, 26 Mar 2024
General
The marginal sea ice zone (MIZ) is characterized by strong surface inhomogeneity with respect to roughness and temperature. The typical scale of inhomogeneity is much smaller than the grid size of climate and weather prediction models, so that it is a challenge to parametrize turbulent fluxes close to reality. This paper attempts to study the impact of different ice floe patterns on domain averaged flux profiles over the MIZ by Large Eddy Simulation (LES).
The topic is challenging and important for polar climate modelling and weather prediction. In most parts the paper is well written, and the principal approach is adequate and can stimulate further scientific work. However, as explained below, there are some unclear points which should be considered before the paper is published. Qualitatively, the principle conclusions concerning the impact of sea ice patterns will probably not be affected by suggested modifications but their might be quantitative effects.
Major revisions
- My most important concern is the used grid size of 100 m. The problem of grid spacing in LES over sea ice with open water fraction is addressed in Gryschka et al. (2023), Lüpkes et al. (2008) and especially by Weinbrecht and Raasch (2001). The latter show that in LES 2 m grid spacing should be chosen when the width of open water leads is 200 m, Lüpkes et al. (2008) used in the LES 10 m for leads of 1 km width and a similar grid spacing (20 m) is chosen by Gryschka et al. (2023). Lüpkes et al. (2008) further show that for mesoscale simulations with coarser grid (200 m horizontal grid size) a lead specific nonlocal parametrization is necessary to obtain the correct plume inclination and vertical temperature gradients on the downstream side of open leads. The considered situation might be different due to the larger open water fraction but I recommend at least one model run with a strongly reduced grid size (e.g. 50 m) to test the sensitivity of the obtained flux profiles and thus main results on the resolution.
- Furthermore, in chapter 2.2 it is said that resolutions of the sea ice maps are based on much higher resolutions (2m, 10m, 20 m etc.). I cannot follow here, the maps shown in Figure 2 do not reflect these resolutions. In case of a higher resolution of the surface than of the LES one would need subgridscale flux parametrizations, which are not mentioned here. All this needs clarification.
- The authors use a roughness length of 1 cm for open water. This value is much too high (orders of magnitude) for open water surrounded by sea ice. More reasonable values can be found, e.g. in Andreas et al. (2010), Lüpkes et al. (2012), Lüpkes et al. (2008), Elvidge et al. (2016), and in Gryschka et al. (2023). The latter discussed the choice of 1 cm and consequences on results. Their findings concerning the effect of too large roughness lengths must at least be discussed but new simulations with realistic roughness length would be better.
- It is not enough to show just the flux profiles and wind. To understand the consequences of assumptions, it is necessary to show also the domain averaged profiles of potential temperature as in Michaelis and Lüpkes (2022). Also vertical cross-sections of temperature and wind at some positions and horizontal cross-sections could be helpful to understand differences in the ABL structure between model runs as well as possible difficulties of the LES. It is also necessary to mention the height averaged wind speed in the ABL. At least in conditions with high ice fraction and some open leads Lüpkes et al. (2008) found a strong dependence of the ABL development on wind speed due to their importance on plume inclination over and downstream of leads.
- Lines 61-63: It seems that the authors are not aware of the papers Lüpkes et al. (2008), Michaelis et al. (2021) and Michaelis and Lüpkes (2022). In all papers it is explained that a parametrization for orthogonal flow over leads in sea ice is developed and applied based on LES. Thus, although it is not LES, it is qualitatively different to other mesoscale model applications. Especially interesting for the submitted paper of Fogarty et al. is the work of Michaelis and Lüpkes (2022). They do very similar studies applying their LES-based turbulence parametrization over an ensemble of leads (see e.g. their figures 6 and 7). The main difference is that the sea ice fraction is 93 %, so much higher than in the present study and that a simpler (2D) geometry of the open water fraction is used. The new findings of the present study should be discussed considering this work.
- A Coriolis parameter is used for 90°N. This needs justification because it is not really realistic. In winterly temperature conditions prescribed in the model, a sea ice fraction of 50 % would be a rare event at North Pole. A more realistic choice would be 80°N, the typical latitude of the MIZ in the Fram Strait. What is the effect of this choice?
- The authors write always just ‘air temperature’. But I think at all occurrences, they mean air potential temperature (e.g. in equations B2, B3). This means, however, that the model is initialized with a neutral stratification throughout the atmosphere. I am afraid that this might lead to unrealistic boundary layers. Note that the usually found is for such ice fractions a convective layer that is capped by a very strong inversion somewhere between about 300 and 700 m condition (if not affected by a thick stratus layer). Such inversions cause entrainment and influence the ABL development (see e.g. Tetzlaff et al., 2015). This needs at least discussion.
Minor revisions
Line 38: the term MIZ was introduced some decades before the paper of Dumont (2022), so that more references than just this paper should be given.
Line 59: it should be even if some….
Section 2.1: More information is needed here (see above): Which lateral boundary conditions are used in the LES? How strong is geostrophic wind? What about humidity? Are these dry runs without clouds?
Figure 2: What is the unit of the axes? I suggest including two vectors illustrating the geostrophic wind and boundary layer wind.
Line 136: I am not sure if I understood Figure 2 and its relation to the different resolutions correctly. This should be better explained. It would be helpful to use kilometers as a unit for the axes and to give some distances between floes (or the width of leads).
Line 196: It is not the geostrophic wind alone. The near-surface wind is dominating the fluxes. However, the near-surface wind direction might differ from case to case for the same geostrophic wind.
Line 200: One could cite Michaelis et al (2021) in this connection (occurrence of LLJ) as well as Tetzlaff et al (2015). This would support the results.
Line 245: I would not write that differences are minimal. Note that smallest and highest surface fluxes differ by about 30 % from each other, which is a lot.
Line 368: The stability over only ice or water depends on many factors, especially on the air temperature and wind direction. It can happen that there is an unstable stratification over sea ice (cold-air advection) and a stable stratification over the open ocean (warm air advection).
References
Andreas, E. L., Horst, T. W., Grachev, A. A., Persson, P. O. G., Fairall, C. W., Guest, P. S., & Jordan, R. E. (2010). Parametrizing turbulent exchange over summer sea ice and the marginal ice zone. QJRMS, 136(649), 927-943.
Elvidge, A. D., Renfrew, I. A., Weiss, A. I., Brooks, I. M., Lachlan-Cope, T. A., & King, J. C. (2016). Observations of surface momentum exchange over the marginal ice zone and recommendations for its parametrisation. Atmospheric Chemistry and Physics, 16(3), 1545-1563.
Gryschka, M., Gryanik, V. M., Lüpkes, C., Mostafa, Z., Sühring, M., Witha, B., & Raasch, S. (2023). Turbulent heat exchange over polar leads revisited: A large eddy simulation study. J. Geophys. Res.: Atmospheres, 128(12), e2022JD038236.
Lüpkes, C., Gryanik, V. M., Witha, B., Gryschka, M., Raasch, S., & Gollnik, T. (2008). Modeling convection over arctic leads with LES and a non‐eddy‐resolving microscale model. J. Geophys. Res.: Oceans, 113(C9).
Lüpkes, C., Gryanik, V. M., Hartmann, J., & Andreas, E. L. (2012). A parametrization, based on sea ice morphology, of the neutral atmospheric drag coefficients for weather prediction and climate models. J. Geophys. Res.: Atmospheres, 117(D13).
Michaelis, J., Lüpkes, C., Schmitt, A. U., & Hartmann, J. (2021). Modelling and parametrization of the convective flow over leads in sea ice and comparison with airborne observations. QJRMS, 147(735), 914-943.
Michaelis, J., & Lüpkes, C. (2022). The impact of lead patterns on mean profiles of wind, temperature, and turbulent fluxes in the atmospheric boundary layer over sea ice. Atmosphere, 13(1), 148.
Tetzlaff, A., Lüpkes, C., & Hartmann, J. (2015). Aircraft‐based observations of atmospheric boundary‐layer modification over Arctic leads. Quarterly Journal of the Royal Meteorological Society, 141(692), 2839-2856.
Weinbrecht, S., & Raasch, S. (2001). High‐resolution simulations of the turbulent flow in the vicinity of an Arctic lead. J. Geophys. Res.: Oceans, 106(C11), 27035-27046.
Citation: https://doi.org/10.5194/egusphere-2024-532-RC2
Data sets
Large-Eddy Simulation and Statistical Metric Results for Patterned Sea Ice Surfaces Joseph Fogarty and Elie Bou-Zeid http://arks.princeton.edu/ark:/88435/dsp01rr1721506
Viewed
| HTML | XML | Total | BibTeX | EndNote | |
|---|---|---|---|---|---|
| 175 | 54 | 14 | 243 | 10 | 12 |
- HTML: 175
- PDF: 54
- XML: 14
- Total: 243
- BibTeX: 10
- EndNote: 12
Viewed (geographical distribution)
| Country | # | Views | % |
|---|
| Total: | 0 |
| HTML: | 0 |
| PDF: | 0 |
| XML: | 0 |
- 1