betaFeedback

WorksPhysics and Astronomy

Wave Attenuation in Drifting Sea Ice: A Mechanistic Model for Observed Decay Profiles

Abstract

Wave-sea ice interactions shape the transition zone between open ocean and pack ice in the polar regions. Most theoretical paradigms, implemented in coupled wave-sea ice models, predict exponential decay of the wave energy but some recent observations deviate from this behaviour. Expanding on a framework based on wave energy dissipation due to ice-water drag, we account for drifting sea ice to derive an improved model for wave energy attenuation. Analytical solutions replicate the observed non-exponential wave energy decay and the spatial evolution of the effective attenuation rate in Antarctic sea ice.

Introduction

Sea ice that seasonally forms in the polar regions regulates heat and momentum fluxes between the ocean and the atmosphere over large scales, thereby playing a crucial role in the global weather and climate system (e.g. [14, 5, 27]). The transition zone between the open ocean and the pack ice, often referred to as the marginal ice zone (MIZ; [35]), is the most dynamic, yet poorly understood, component of such system ([21, 31, 8]).

The interaction of ocean waves propagating into the sea ice is a defining feature of the Arctic and Antarctic MIZ ([18, 30, 2, 6, 8]), and in the Southern Ocean this region extends for hundreds of kilometres. Wave energy decays with distance into the MIZ, with the attenuation rate α\alpha, i.e. E(x)∝exp⁡(−αx)E(x) \propto\exp(-\alpha x), assumed to be proportional to the wave frequency ff to some power (α∝fp\alpha\propto f^p; e.g. [23]). The exact frequency dependency varies according to the dominant attenuation mechanism, and various options are implemented in operational third generation spectral wave models (e.g. WaveWatchIII; [24]) that are being coupled into global- and hemispheric-scale ocean-sea ice models. Most models predict a constant attenuation rate in homogeneous sea ice conditions and, consequently, exponential wave energy decay.

Overall, measurements conform reasonably well to the exponential attenuation paradigm. and agree on the magnitude of the attenuation rate α\alpha, within the range (10−4 m−1<α<10−5 m−110^{-4}\,\mathrm{m}^{-1} < \alpha< 10^{-5}\,\mathrm{m}^{-1}; [17, 25]). However, recent satellite measurements of wave height in the Antarctic MIZ across seasons revealed an almost linear increase of the average attenuation rate with distance from the sea ice edge (up to ≈10×\approx10\times; [33]) with individual transects showing sharp peaks in attenuation rate. Possible causes of deviation from the exponential profile were attributed to the misalignment between the transect and wave direction, and inhomogeneous sea ice conditions (sea ice concentration and thickness). Lacking a firm theoretical foundation, Voermans et al. [33] propose an empirical wave attenuation formulation as an interim solution for implementation in wave models, i.e. α=βx/xMIZ−γ\alpha= \beta x/x_{\mathrm{MIZ}} - \gamma in which β\beta and γ\gamma are empirical fitting parameters and xMIZx_{\mathrm{MIZ}} is the extent of the MIZ. The formulation leads to exponential energy decay at a rate βx2\beta x^{2}.

Amongst the various theoretical wave attenuation models, the ones that attribute wave energy dissipation to quadratic ice-water drag deviate from exponential decay in form. Ice-water drag models capture the bulk of the energy loss when sea ice comprises floes with small diameter, relative to wavelength, in high concentration ([28, 16, 12]) as further supported by experimental observations ([29]). This makes these models well suited to the Antarctic MIZ. The attenuation rate due to drag is proportional to the square of the relative velocity between fluid orbital motion and sea ice ([12]). Assuming a stationary semi-infinite sea ice domain, analytical, non-exponential, solutions for the amplitude profile have been obtained under slightly different approaches by Kohout et al. [16], for the amplitude envelope, and Herman et al. [12], for the phase-averaged amplitude. The former is a lower bound for the amplitude, i.e. a higher bound for the attenuation rate.

The drift of sea ice is a feature of MIZ dynamics throughout the year ([20, 8]). The Southern Ocean, under persistent waves ([9]), is characterised by unconsolidated small floes (≈10\approx10 m in diameter) in high concentration (>80%> 80\%; [3, 8]) with little internal stresses ([1, 40]). In these conditions, wind is the main driver of sea ice drift ([13, 20]). Through air-ice drag coefficients of order 10−310^{-3} ([36, 10]), sea ice drifts at 0.1–0.4 m/s with low variability over large scales (10–100 km; [13]). Nevertheless, the drift of the MIZ is particularly intense during extreme events when concurrent energetic ocean wave are able to penetrate hundreds of kilometres into the sea ice, as demonstrated by recent observations ([32, 1, 40, 39]). These events are likely the main contributors to the overall MIZ dynamics. However, to the best of our knowledge, the interplay between drift and wave attenuation has not been explored before.

Several of the world’s leading meteorological agencies have moved to using coupled atmosphere–ocean–sea-ice forecasting systems in the last few years to improve forecasting of near-surface temperature, humidity, winds and low-level clouds, as well as sea-ice distribution in the polar regions. In this regard, prediction in the MIZ is proving a particularly difficult challenge ([7, 4]). One aspect of focussed development is the coupling of physics-based ocean wave and sea-ice models, within a dynamical ocean model, so subject to drift ([19]), making this study timely with the potential to inform this development.

Here, we develop a new analytical framework for the attenuation of monochromatic waves due to drag, explicitly accounting for constant sea ice drift.

We extend the previous works of Kohout et al. [16] and Herman et al. [12] where drift was neglected on the grounds that, in typical oceanic conditions, drift velocities are two orders of magnitude smaller than the wave group velocity. Despite its apparently negligible effect, we reveal distinctive consequences of including such drift, most notably the emergence of a location into the ice field where the wave amplitude vanishes and the attenuation rate becomes unbounded. We further demonstrate the capability of this enhanced model to reproduce recent observations of wave attenuation in the Antarctic MIZ.

Mathematical Formulation and Analytical Solutions

The one-dimensional wave energy transport equation for energy density per unit surface area, E(x,t)E(x,t), in sea ice-covered regions and excluding all other source terms is:

∂E∂t+cg∂E∂x=Sice=Ssd+Sexp,(1)\frac{\partial E}{\partial t}+c_g\frac{\partial E}{\partial x}=S_{\mathrm{ice}}=S_{\mathrm{sd}}+S_{\mathrm{exp}}, \tag*{(1)}

where cgc_g denotes the group velocity and, similarly to Kohout et al. [16], the effect of the ice (SiceS_{\mathrm{ice}}) is explicitly expressed as a linear superposition of all the mechanisms with constant decay, that lead to exponential attenuation (SexpS_{\mathrm{exp}}), and the ones due to ocean-ice skin drag (SsdS_{\mathrm{sd}}). Note that the assumption of exponential decay for all the mechanisms other than drag will facilitate analytic handling. Here, in agreement with most wave-in-ice parametrisations in WaveWatchIII [24], we assume the open water dispersion relation between angular frequency ω\omega and wave-number kk to hold in sea ice.

The transport equation in terms of wave amplitude a(x,t)a(x,t), using E=12ρga2E=\frac{1}{2}\rho g a^2 where ρ\rho is water density and gg acceleration due to gravity, becomes:

a∂a∂t+acg∂a∂x=sice=sexp+ssd,(2)a\frac{\partial a}{\partial t}+ac_g\frac{\partial a}{\partial x}=s_{\mathrm{ice}}=s_{\mathrm{exp}}+s_{\mathrm{sd}}, \tag*{(2)}

where s=S/ρgs=S/\rho g. In agreement with previous work [e.g. 16, 12], in ice covered regions we define the exponential source term as

sexp=−cgαexpa2,(3)s_{\mathrm{exp}}=-c_g\alpha_{\mathrm{exp}}a^2, \tag*{(3)}

where αexp\alpha_{\mathrm{exp}} is the constant amplitude attenuation rate for all other contributing non skin-drag mechanisms e.g. wave scattering [34], viscous dissipation [38, 15] and visco-elasticity [26, 11, 37]. Its inclusion provides flexibility for future operational model implementations.

In ice covered regions, for a monochromatic wave, the skin drag stress at the ice-ocean interface follows a standard quadratic law, τ=ρCd∣uorb−v∣(uorb−v)\tau=\rho C_d\lvert u_{\mathrm{orb}}-v\rvert(u_{\mathrm{orb}}-v) [28], with CdC_d the drag coefficient. The rate of change of energy per unit area due to skin drag is proportional to ∣uorb−v∣3\lvert u_{\mathrm{orb}}-v\rvert^3 [28], where uorb=Usin⁡(ϕ)u_{\mathrm{orb}}=U\sin(\phi) is the wave orbital velocity at the ice-water interface, here evaluated at the free surface under the hypothesis of thin sea ice, and vv is the ice drift velocity. The magnitude of the orbital velocity is U=aΩU=a\Omega, where Ω=gk/ω\Omega=gk/\omega, and ϕ\phi is time-dependent phase. Phase averaging, we obtain:

ssd=−Cdg12π∫02π∣aΩsin⁡(ϕ)−v∣3 dϕ.(4)s_{\mathrm{sd}}=-\frac{C_d}{g}\frac{1}{2\pi}\int_0^{2\pi}\lvert a\Omega\sin(\phi)-v\rvert^3\,d\phi. \tag*{(4)}

We consider a thin semi-infinite homogenised sea ice layer drifting at a constant velocity vv and arbitrarily set a frame of reference moving at vv, i.e. x^=x−vt\hat{x}=x-vt, with x^\hat{x} the distance from the edge of the sea ice. Taking the steady state system in the moving frame with a constant wave forcing at the ice edge a(0)=a0a(0)=a_0, the one-dimensional wave energy transport equation Eq. 2.2 in the ice region (x^≥0\hat{x}\geq0) reduces to:

dadx^=−αa−Γa∫02π∣aΩsin⁡(ϕ)−v∣3 dϕ(5)\frac{\mathrm{d}a}{\mathrm{d}\hat{x}}=-\alpha a-\frac{\Gamma}{a}\int_{0}^{2\pi}\lvert a\Omega\sin(\phi)-v\rvert^{3}\,\mathrm{d}\phi \tag*{(5)}

with

Γ=Cd2πg(cg−v),α=cgcg−vαexp⁡(6)\Gamma=\frac{C_d}{2\pi g(c_g-v)},\qquad\alpha=\frac{c_g}{c_g-v}\alpha_{\exp} \tag*{(6)}

being the attenuation, therefore remaining exponential in the moving frame.

Solving the Amplitude Profile

Equation 2.5 represents a challenging nonlinear ordinary differential equation. Here we aim to derive an analytical solution and to facilitate the mathematical treatment we assume the parameters CdC_d and α\alpha to be spatially homogeneous (constant). The model can be extended for heterogeneous conditions but numerical solutions should be sought. A polynomial in aa for the integral of the skin drag (Eq. 2.4) can be found explicitly in closed form for ∣v∣≥aΩ\lvert v\rvert\geq a\Omega, see Appendix A. In effort to enable closed form analytical solutions we seek to an approximate polynomial form using an asymptotic expansion of the integrand for ∣v∣≪aΩ\lvert v\rvert\ll a\Omega that are then extended into the region ∣v∣<aΩ\lvert v\rvert< a\Omega (see Appendix A). We obtain:

I=∫02π∣aΩsin⁡(ϕ)−v∣3 dϕ≈{83(aΩ)3+12aΩv2,∣v∣<aΩ;3π(aΩ)2∣v∣+2π∣v∣3,∣v∣≥aΩ.(7)I=\int_{0}^{2\pi}\lvert a\Omega\sin(\phi)-v\rvert^{3}\,\mathrm{d}\phi\approx \begin{cases} \frac{8}{3}(a\Omega)^{3}+12a\Omega v^{2}, & \lvert v\rvert<a\Omega;\\ 3\pi(a\Omega)^{2}\lvert v\rvert+2\pi\lvert v\rvert^{3}, & \lvert v\rvert\geq a\Omega. \end{cases} \tag*{(7)}

Note that, in oceanic conditions, cgc_g is two orders of magnitude lager than the wind and current induced drift, i.e. O(10)O(10) m/s versus O(0.1)O(0.1) m/s, therefore ∣v∣≪cg\lvert v\rvert\ll c_g implying that, with α\alpha and Γ\Gamma positive, Eq. 2.7 is strictly positive, therefore the amplitude aa decreases monotonically in Eq. 2.5. For a0Ω>∣v∣a_0\Omega>\lvert v\rvert, we define x^∗\hat{x}^{*} to be the location into the ice domain such that a(x^∗)Ω=∣v∣a(\hat{x}^{*})\Omega=\lvert v\rvert. Therefore, the condition ∣v∣≤aΩ\lvert v\rvert\leq a\Omega corresponds to x^≤x^∗\hat{x}\leq\hat{x}^{*} in the physical domain. In contrast, for a0Ω≤∣v∣a_0\Omega\leq\lvert v\rvert, one obtains x^∗=0\hat{x}^{*}=0.

Solving Eq. 2.5 with the integral approximation Eq. 2.7, the amplitude solutions are written in piecewise form as:

a(x^)≈aA(x^) 1{x^<x^∗}+aB(x^) 1{x^≥x^∗}(8)a(\hat{x})\approx a_A(\hat{x})\,\mathbf{1}_{\{\hat{x}<\hat{x}^{*}\}}+a_B(\hat{x})\,\mathbf{1}_{\{\hat{x}\geq\hat{x}^{*}\}} \tag*{(8)}

where 1\mathbf{1} is the indicator function, equal to 1 if the condition is true and 0 otherwise. The piecewise solution components result in:

aA(x^)={3Δ16ΓΩ3tan⁡(−Δ2x^+arctan⁡(16ΓΩ3a0+3α3Δ))−3α16ΓΩ3,δ>1,32ΓΩ3a0−(16αΓΩ3a0+3α2)x^32ΓΩ3+(32ΓΩ3a0+6α)x^,δ=1,r−(a0−r+)exp⁡(8ΓΩ33(r−−r+)x^)−r+(a0−r−)(a0−r+)exp⁡(8ΓΩ33(r−−r+)x^)−(a0−r−),δ<1.(9)a_A(\hat{x})= \begin{cases} \frac{3\sqrt{\Delta}}{16\Gamma\Omega^{3}}\tan\left(-\frac{\sqrt{\Delta}}{2}\hat{x}+\arctan\left(\frac{16\Gamma\Omega^{3}a_0+3\alpha}{3\sqrt{\Delta}}\right)\right)-\frac{3\alpha}{16\Gamma\Omega^{3}}, & \delta>1,\\ \frac{32\Gamma\Omega^{3}a_0-(16\alpha\Gamma\Omega^{3}a_0+3\alpha^{2})\hat{x}}{32\Gamma\Omega^{3}+(32\Gamma\Omega^{3}a_0+6\alpha)\hat{x}}, & \delta=1,\\ \frac{r_{-}(a_0-r_{+})\exp\left(\frac{8\Gamma\Omega^{3}}{3}(r_{-}-r_{+})\hat{x}\right)-r_{+}(a_0-r_{-})}{(a_0-r_{+})\exp\left(\frac{8\Gamma\Omega^{3}}{3}(r_{-}-r_{+})\hat{x}\right)-(a_0-r_{-})}, & \delta<1. \end{cases} \tag*{(9)}
aB(x^)=[e−(6πΩ2∣v∣Γ+2α)(x^−x^∗)(v2Ω2+2π∣v∣3Γ3πΩ2∣v∣Γ+α)−2π∣v∣3Γ3πΩ2∣v∣Γ+α]12.(10)a_B(\hat{x})=\left[e^{-(6\pi\Omega^{2}\lvert v\rvert\Gamma+2\alpha)(\hat{x}-\hat{x}^{*})}\left(\frac{v^{2}}{\Omega^{2}}+\frac{2\pi\lvert v\rvert^{3}\Gamma}{3\pi\Omega^{2}\lvert v\rvert\Gamma+\alpha}\right)-\frac{2\pi\lvert v\rvert^{3}\Gamma}{3\pi\Omega^{2}\lvert v\rvert\Gamma+\alpha}\right]^{\frac{1}{2}}. \tag*{(10)}

Here Δ≡α2(δ2−1)\Delta\equiv\alpha^{2}(\delta^{2}-1), r±≡−3(α±Δ)/(16ΓΩ3)r_{\pm}\equiv-3(\alpha\pm\sqrt{\Delta})/(16\Gamma\Omega^{3}), and

δ=82ΓΩ2∣v∣α(11)\delta=\frac{8\sqrt{2}\Gamma\Omega^{2}\lvert v\rvert}{\alpha} \tag*{(11)}

which controls the roots of the quadratic polynomial in aa affecting the solution in the x^<x^∗\hat{x} < \hat{x}^{*} region (see Appendix § 2). Note that the physical interpretation of δ\delta will be discussed in §2.2.

The solution has a few interesting features. The amplitude profile is continuous in x^\hat{x} (see Appendix § 2). The solution is nonlinear, i.e. drag effects and other mechanisms cannot be added as linear superposition. For x^<x^∗\hat{x} < \hat{x}^{*}, the different branches of the solution in Eq. (9) differ in form but possess similar qualitative behaviour, where we note that x^∗\hat{x}^{*} depends on δ\delta. For a wave with initial orbital velocity magnitude U0<∣v∣U_{0} < \lvert v\rvert we obtain the solution in region B, aBa_{B}, applies to the entire domain x^≥0\hat{x} \ge0.

A distinctive consequence of such solution is the existence of an extinction location

x^end=x^∗+832παδ+8log⁡(52+42πδ),(12)\hat{x}_{\mathrm{end}} = \hat{x}^{*} + \frac{8}{3\sqrt{2}\pi\alpha\delta+ 8}\log\left(\frac{5}{2} + \frac{4\sqrt{2}}{\pi\delta}\right), \tag*{(12)}

such that aB(x^=x^end)=0a_{B}(\hat{x} = \hat{x}_{\mathrm{end}}) = 0. The extinction location is always in region B because it is where amplitude tends to zero and the condition aΩ<∣v∣a\Omega< \lvert v\rvert is satisfied. The extinction location is unique to dissipation due to drift-induced skin drag and defines the extent of the wave affected region. It moves closer to the sea ice edge for increasing CDC_{D} coefficient and α\alpha (see solution in red, for v=0.05 m/sv = 0.05\ \mathrm{m/s} and α=7×10−6 m−1\alpha= 7 \times10^{-6}\ \mathrm{m}^{-1} [δ=3.4\delta= 3.4], vs in blue, for v=0.05 m/sv = 0.05\ \mathrm{m/s} and α=0\alpha= 0 [δ=∞\delta= \infty], in Figure 1a, while its dependence on drift is not trivial. Such extinction locations are not present in the solutions of Kohout et al. (2011) and Herman et al. (2019), shown dashed green and orange respectively in Figure 1a, for which α=0\alpha= 0 and v=0v = 0.

Sample amplitude and corresponding attenuation profiles

Figure 1. Sample amplitude (a) and corresponding attenuation (b) profiles according to our model and formulations by Kohout et al. (2011) and Herman et al. (2019). Dashed lines denote x∗x^{*} which denotes the separation between region A and B. Amplitude is normalised by the initial amplitude and distance by xendx_{\mathrm{end}} for α=0\alpha= 0. Attenuation rate is normalised by α\alpha. Solutions are shown for a0=1 ma_{0} = 1\ \mathrm{m}, Cd=0.05C_{d} = 0.05, v=0.053 m s−1v = 0.053\ \mathrm{m}\ \mathrm{s}^{-1}, ω=0.63 s−1\omega= 0.63\ \mathrm{s}^{-1}, α=7.2×10−6 m−1\alpha= 7.2 \times10^{-6}\ \mathrm{m}^{-1}.

In the special case of no drift, i.e. v=0v = 0, the solution reduces to:

a(x^)=3a0αe−αx^3α+8a0ΓΩ3−8a0ΓΩ3e−αx^.(13)a(\hat{x}) = \frac{3a_{0}\alpha e^{-\alpha\hat{x}}}{3\alpha+ 8a_{0}\Gamma\Omega^{3} - 8a_{0}\Gamma\Omega^{3}e^{-\alpha\hat{x}}}. \tag*{(13)}

Taking Cd=0C_{d} = 0, i.e. no drag, we recover the exponential attenuation at a rate α\alpha whereas for α→0\alpha\to0 the solution converges to the one of Herman et al. (2019). Therefore, Eq. (13) is an extension Herman et al. (2019) that accounts for additional mechanisms. As a result, our solution predicts lower amplitudes compared to Herman et al. (2019), see Figure 1a.

Figure 1a displays the normalised amplitude as a function of the distance into the MIZ normalised by the value xendx_{\mathrm{end}} computed for the non zero drift and exponential attenuation case (v=0.053 m s−1v = 0.053\ \mathrm{m\ s}^{-1} and α=0\alpha= 0; the case displayed in blue). The attenuation is stronger when drift is present as shown in red (v>0v > 0 and α=7×10−6\alpha= 7 \times10^{-6}) compared to the solution in purple with (v=0v = 0 m/s and α=7×10−6\alpha= 7 \times10^{-6}). The increased attenuation due to other mechanisms also manifests when drift is present, see solutions in blue vs in red.

Predicting the Attenuation Rate

To allow better comparison with attenuation results for non-drifting sea ice as they appear in the literature [16, 12] and better highlight the attenuation characteristics, we compute the effective amplitude attenuation rate in the moving frame in the ice covered region defined as

αeff=−1adadx^=α+Γa2I.(14)\alpha_{\mathrm{eff}} = -\frac{1}{a}\frac{\mathrm{d}a}{\mathrm{d}\hat{x}} = \alpha+ \frac{\Gamma}{a^{2}}I . \tag*{(14)}

Using the integral estimation (Eq. 2.7) we obtain:

αeff(x^)≈{α+83aΩ3Γ+12Ωv2Γa,if x^<x^∗;α+3πΩ2∣v∣Γ+2π∣v∣3Γa2,if x^≥x^∗.(15)\alpha_{\mathrm{eff}}(\hat{x}) \approx \begin{cases} \alpha+ \frac{8}{3}a\Omega^{3}\Gamma+ \frac{12\Omega v^{2}\Gamma}{a}, & \text{if } \hat{x} < \hat{x}^{*};\\ \alpha+ 3\pi\Omega^{2}|v|\Gamma+ \frac{2\pi|v|^{3}\Gamma}{a^{2}}, & \text{if } \hat{x} \ge\hat{x}^{*}. \end{cases} \tag*{(15)}

Note that, by extending asymptotic solution of II for ∣v∣<aΩ|v| < a\Omega to ∣v∣=aΩ|v| = a\Omega, the decay rate inherits a jump discontinuity at x^∗\hat{x}^{*}, shown in Figure 1b, with the suitability of the integral approximation justified by discontinuity magnitude being only 7%. The nonlinearity between skin drag and other exponential decay mechanisms leads to the lower bound

αeff≥α+82ΓΩ2∣v∣=α(1+δ),(16)\alpha_{\mathrm{eff}} \ge\alpha+ 8\sqrt{2}\Gamma\Omega^{2}|v| = \alpha(1+\delta), \tag*{(16)}

with the global minimum located in the region x^<x^∗\hat{x} < \hat{x}^{*} if a0Ω≥32∣v∣/2a_{0}\Omega\ge3\sqrt{2}|v|/2, whereas for lower orbital velocity at the sea ice edge the attenuation monotonically increases. This result allows for a physical interpretation of the dimensionless parameter δ\delta: for δ>1\delta> 1 wave attenuation is dominated by drift-induced skin drag and for δ<1\delta< 1 all other mechanisms dominate.

For drifting sea ice, the attenuation rate grows unbounded in the neighbourhood of the extinction location (see solutions in red and blue in Figure 1b), i.e. αeff(x^)∼[2(xend−x^)]−1\alpha_{\mathrm{eff}}(\hat{x}) \sim[2(x_{\mathrm{end}}-\hat{x})]^{-1}, therefore the extinction location affects the magnitude of the attenuation rate in this region. When drift is absent, x^∗→∞\hat{x}^{*} \to\infty and the solution for x^<x^∗\hat{x} < \hat{x}^{*} applies to the entire domain. As a consequence of our model, skin drag dominates attenuation close to the sea ice edge whereas other processes dominate deeper into the sea ice, where we retrieve the classical exponential attenuation, i.e. αeff→α\alpha_{\mathrm{eff}} \to\alpha for x^→∞\hat{x} \to\infty (see solution in purple in Figure 1b; the grey dashed line denotes αeff=α\alpha_{\mathrm{eff}} = \alpha).

Comparisons to Antarctic Measurements

In this section comparisons are made with measured amplitudes and derived attenuations in the Antarctic sea ice [6, 33]. The observations comprises quality controlled ICESat-2 satellite transects of wave height (both the total wave heights and the spectral wave heights corresponding to frequency bands at periods T=9,12,15,18T = 9, 12, 15, 18 s) across the MIZ throughout 2019. Note that transects seldom start at the ice edge (possibly because they fail quality control in intermediate sea ice concentrations) and, in computing attenuation rates, transects are assumed to align with the wave direction [33]. Drift is not measured by ICESat-2.

To compare observations to model results, which are derived for a monochromatic wave, we attribute all energy to the most energetic frequency component measured over the transect, normalising amplitude and attenuation rate by the first available measurement from the ice edge, aV1a_{\mathrm{V}}^{1} and αV1\alpha_{\mathrm{V}}^{1}. The ice edge is taken to coincide be the start of each transect. To allow for a straightforward comparison of the model with the observations we use xx in place of x^\hat{x} noting that their difference is negligible for small drift velocities. Moreover, in agreement with the observations, transects are normalised by MIZ extent xMIZx_{\mathrm{MIZ}}. In this regard is important to note that MIZ extent is reported independently from the wave amplitude, i.e. low amplitude readings were excluded due to quality control ([33]).

Drift and drag are used as fitting parameters, within the range of available Antarctic sea ice observations, i.e. v=O(10−1)v = O(10^{-1}) m/s and Cd=O(10−2)C_d = O(10^{-2}), together with α\alpha. While least squares optimisations was attempted, the noise in the observational data prevented reasonable fitting throughout the transect. For illustrative purposes, we performed manual tuning for each transect of the parameters within ranges of characteristic values for the MIZ. The drift is assumed to be constant along the transect, justified by slow spatial variation over 10–100 km in Antarctic sea ice ([13]). The water density is set ρ=1025\rho= 1025 kg/m3^3 and deep water dispersion relation used, namely Ω=ω\Omega= \omega.

Individual Transects

The two transects reported in Voermans et al. [33] are analysed in detail. Transect I, taken on 26 December 2019 at 85°W, 69°S, is 61.5 km long and waves have peak period 15 s with amplitude aV1=0.45a_{\mathrm{V}}^{1} = 0.45 m and attenuation, αV1=1.42×10−5\alpha_{\mathrm{V}}^{1} = 1.42 \times10^{-5} m−1^{-1}. Transect II, taken on 24 May 2019 at 88°W, 68°S, is longer (xMIZ=137.5x_{\mathrm{MIZ}} = 137.5 km) and waves have a shorter peak period (T=12T = 12 s) with amplitude aV1=0.55a_{\mathrm{V}}^{1} = 0.55 m and attenuation, αV1=2.43×10−6\alpha_{\mathrm{V}}^{1} = 2.43 \times10^{-6} m−1^{-1}.

For Transect I the parameters v=0.26v = 0.26 m/s, Cd=6.0×10−3C_d = 6.0 \times10^{-3} and α=5.0×10−6\alpha= 5.0 \times10^{-6} provide good qualitative and quantitative agreement in terms of amplitude decay (see Figure 2a). Similar performances are achieved by the the non-drifting model, i.e. v=0v = 0, and the exponential decay. However, the model with drift outperforms the others reproducing the observed increase in effective attenuation rate along the transect (between x/xMIZ≈0.2x/x_{\mathrm{MIZ}} \approx0.2 and 0.6; see Figure 2b). Note that for the chosen parameters, drift is stronger than the orbital velocity throughout the transect (∣v∣/U0≈1.2|v|/U_0 \approx1.2), i.e. only the branch x>x^∗x > \hat{x}^{*} of Eq. (15) is used. Remarkably, the model predicts an extinction location consistent with the observed wave-affected sea ice extend, i.e. at x/xMIZ≈1x/x_{\mathrm{MIZ}} \approx1.

Figure 2. Amplitude (left) and corresponding attenuation profiles (right) for Transect A (top) and B (bottom). Measurements (red dots) are shown against model predictions. Distance is normalised with respect to the wave-affected sea ice extent xMIZx_{\mathrm{MIZ}} and amplitude/attenuation with respect to the measurement closest to the sea ice edge as reported in Voermans et al. [33].

For Transect II the parameters v=0.49v = 0.49 m/s, Cd=2.8×10−3C_d = 2.8 \times10^{-3} and α=1.0×10−7\alpha= 1.0 \times10^{-7} qualitatively reproduce the measured mild amplitude decay up to x/xMIZ<0.6x/x_{\mathrm{MIZ}} < 0.6 followed by a sharp decline (see Figure 2c). Other models, i.e. the exponential and non-drifting model better reproduce the attenuation near to the ice edge, but fail to capture the sharp decline deeper into the sea ice. The predictive performance of the drifting model is clearly highlighted by the attenuation rate, shown in Figure 2d, where steady attenuation rate near the ice edge followed by rapid growth for x/xMIZ>0.6x/x_{\mathrm{MIZ}} > 0.6 are well captured qualitatively and quantitatively. Note that the observed attenuation rate fluctuates resulting in negative values close to the edge and again at x/xMIZ≈0.4x/x_{\mathrm{MIZ}} \approx0.4, and models can not capture this behaviour. A jump discontinuity appears at x/xMIZ≈0.1x/x_{\mathrm{MIZ}} \approx0.1, consistent with the choice of vv resulting in ∣v∣/U0≈0.9|v|/U_0 \approx0.9 and the presence of x^∗\hat{x}^{*} within the domain. Unlike our formulation, all other models predict either a decreasing attenuation rate (no drift model) or constant attenuation (exponential model), as shown in Figure 2d.

In summary, Transect I and II showcase two different regimes, in Transect I drag attenuation is weak (δ=0.6\delta= 0.6) whereas in Transect II it is the dominant mechanism (δ=8.0\delta= 8.0). Therefore, as expected, in Transect I deviations from exponential attenuation are weaker than in Transect II however, also in Transect II, inclusion of exponential attenuation through α\alpha provides significantly improved predictions. When drag dominates, the extinction location xendx_{\mathrm{end}} occurs earlier and the spike in attenuation is more prominent.

Our model is developed for homogeneous sea ice conditions, implying constant CdC_d along the transect. While this hypothesis is more likely to hold in the winter season (Transect II) its applicability to Transect I is more debatable. Moreover, direct comparison with individual observations is made difficult by inherent measurement uncertainty, including misalignement of the transect with the wave direction Voermans et al. [33], lack of information on the local sea ice properties, and possible presence of other forcing mechanisms in the wave transport equation, e.g. wind forcing. Nevertheless, we highlight that our simple model captures the observed feature of wave attenuation in the MIZ.

Averaged Attenuation

Voermans et al. [33] report average attenuation profiles across seasons and sectors of the Antarctic MIZ. To reproduce the observed variability, we drive our model, i.e. Eq. (15), with N0=3×105N_0 = 3 \times10^5 random realisations with Gaussian distributed drift velocity (with mean μv\mu_v and standard deviation σv\sigma_v), constant throughout the transect. Each realisation corresponds to an individual transect and NN denotes the number of transects at any given location according to our model, the decrease is due to individual transects reaching the extinction location. The simulations replicate the variability of drift around Antarctica ([20]).

The attenuation rate is computed for the wave component T=12T = 12 s (angular frequency Ω=0.52\Omega= 0.52 s−1^{-1} and wavelength λ=225\lambda= 225 m) and we set the wave amplitude at the sea ice edge to a0=1a_0 = 1 m. The other parameters are Cd=0.02C_d = 0.02 and α=7.0×10−6\alpha= 7.0 \times10^{-6} m−1^{-1}, to match the averaged measured attenuation rate at the sea ice edge. While the drag coefficient is higher than the one chosen for the individual transects in the previous section, its value is within the range of observed values for the MIZ ([22]). Moreover, the chosen value allows to quantitatively match the observed attenuation rate close to the sea ice edge as reported in Voermans et al. (2025). Smaller values of drag would give rise to a similar qualitative behaviour. For the comparison, only a≥0.05ma \ge0.05\mathrm{m} are averaged, in accordance with the methodology used by Voermans et al. (2025) to exclude observations with high uncertainty.

For drift μv=0.22 m/s\mu_v = 0.22\ \mathrm{m/s} (μv/U0=0.42\mu_v/U_0 = 0.42) and standard deviation σv=0.03 m/s\sigma_v = 0.03\ \mathrm{m/s} (Figure 3a) the attenuation only slightly increases up to x/λ≈550x/\lambda\approx550, consistent with the fact that simulations are in weak drag regime (δ=0.42\delta= 0.42). From x/λ≈550x/\lambda\approx550 to x/λ≈700x/\lambda\approx700 the increase is more noticeable, up to ≈3\approx3 times the attenuation rate at the sea ice edge, in qualitative agreement with simulations for Transect I (cf. Figure 2b), before stabilising farther in the sea ice domain. The spread of individual profiles, shaded in grey, also grows when attenuation rate increases. Note that the increase in attenuation corresponds to a reduction in available measurements of amplitude over threshold (shown in red) and, similarly, individual amplitude profiles reaching the extinction location (distribution shaded in blue).

Averaged attenuation rates and distributions

Figure 3. Averaged attenuation rates (in black; for negative velocity in green) and interquartile range (shaded in grey). On the right axis, number of measurements (in red) and distribution of the extinction location (shaded in blue; not to scale). Insets show the corresponding drift velocity distribution: (a) μv=0.22 m s−1\mu_v = 0.22\ \mathrm{m\,s^{-1}} σv=0.03 m s−1\sigma_v = 0.03\ \mathrm{m\,s^{-1}}; (b) μv=0.22 m s−1\mu_v = 0.22\ \mathrm{m\,s^{-1}} σv=0.01 m s−1\sigma_v = 0.01\ \mathrm{m\,s^{-1}}; (c) μv=0.29 m s−1\mu_v = 0.29\ \mathrm{m\,s^{-1}} σv=0.03 m s−1\sigma_v = 0.03\ \mathrm{m\,s^{-1}}.

With the same average drift (μv=0.22 m/s\mu_v = 0.22\ \mathrm{m/s}) but lower variability (from σv=0.03 m/s\sigma_v = 0.03\ \mathrm{m/s} to σv=0.01 m/s\sigma_v = 0.01\ \mathrm{m/s}; Figure 3b) the same qualitative behaviour is observed. The increase in attenuation is sharper, i.e. it starts deeper in the sea ice (at ≈650\approx650 wavelengths) and reaches higher values (αeff≈5.5×10−4\alpha_{\mathrm{eff}} \approx5.5 \times10^{-4} vs αeff≈3.5×10−4\alpha_{\mathrm{eff}} \approx3.5 \times10^{-4}).

The increase in drift (μv=0.29 m/s\mu_v = 0.29\ \mathrm{m/s}, μv/U0=0.55\mu_v/U_0 = 0.55 and σv=0.03 m/s\sigma_v = 0.03\ \mathrm{m/s}; Figure 3c), results in higher δ=0.55\delta= 0.55, still in weak drag regime, and moves the increase in attenuation rate closer to the sea ice edge, from x/λ≈450x/\lambda\approx450 to x/λ≈550x/\lambda\approx550. A slightly higher attenuation is also achieved compared to the previous cases, up to αeff≈6×10−4\alpha_{\mathrm{eff}} \approx6 \times10^{-4}, after which decreases slightly. For smaller drag coefficients results the increase in attenuation occurs further into the ice field. The number of observations decreases sharply due to individual transects reaching their extinction location (cf. Figure 1b), and only less than 0.1% remain after 800 wavelengths, also explaining the larger variance in effective attenuation.

In Figure 3 the mean attenuation rates for negative mean velocity are also reported in dashed green line. The direction of the drift has only marginal effect on the effective attenuation profile, i.e. the growth of attenuation rate is deeper into the sea ice (by ≈10\approx10 wavelengths) but does not affect its magnitude, with similar effect on available measurements and extinction location distributions. This is expected for oceanic conditions, as α≈αexp⁡\alpha\approx\alpha_{\exp} and Γ≈Cd/(gcg2π)\Gamma\approx C_d/(g c_g 2\pi), hence the dependence of direction of the drift is negligible for operational purposes.

Figure 3c is remarkably similar to the measured averaged attenuation in Voermans et al. [33], i.e. a linear increase followed by slight decline, also matching the range of observed attenuation rates. Unlike in the measurements, in the model all individual transects are available from the edge but the decrease of available observations in Voermans et al. [33] deeper into the sea ice is also captured. It can be argued that the presence of drift limits the extent of the wave-affected area, and its extent in our model (≈185\approx185 km; Figure 3c) is consistent to the MIZ width of the order of 200 km as reported in observations ([6]) and model simulations ([8]).

Conclusions

We derived a mechanistic model for wave attenuation that accounts for both phenomenological exponential attenuation mechanisms and dissipation due to drag at the ocean-sea ice interface, as well as the effect of sea ice drift, therefore combining and extending previous models in which drift was excluded (cf. [28, 16, 12]). We identified the parameter δ\delta which governs attenuation dominated by drift-induced drag mechanisms (δ>1\delta> 1) and exponential mechanisms (δ<1\delta< 1). Analytical solutions reveal that in drift-induced drag dominated regime, the amplitude eventually dies off, unlike prediction based on exponential attenuation, and the attenuation profile spikes close to the extinction location.

The model agrees well, both qualitatively and quantitatively with pan-Antarctic measurements of attenuation rate recently reported by Voermans et al. [33] when sea ice is allowed to drift slowly (μv/U0=0.55\mu_v/U_0 = 0.55), even for small δ\delta, i.e. δ=0.55\delta= 0.55. In particular, drift is responsible for the increase in wave attenuation rate deeper into the sea ice and, to a certain extent, regulates the width of the wave-affected sea ice which is in quantitative agreement to the MIZ extent (100–200 km) reported for observations ([6]) and model simulations ([8]).

The quantitative agreement against individual transects is reasonable with limitations. Differences with the measured observations can be attributed to experimental uncertainty and current model assumptions, i.e. no interaction between wave components, no presence of other forcing mechanisms, homogeneous drift and sea ice properties along the transect. Some of the limitations of the model could be sought by seeking numerical solutions for the governing equation (Eq. 2.5) and allowing for spatially heterogenous parameters. In this regard, our modelling framework allows for easy implementation into operational models and the biggest limitation, common to many wave attenuation models, is the lack of physically-informed sea ice parameters along the MIZ.

Ultimately the model provides a solid theoretical foundation to better interpret observations of wave attenuation in drifting sea ice as found around Antarctica, and prompts new concurrent measurements of waves, drift, and sea ice properties for further validation. Moreover, while only drag has been studied explicitly, the approach can be extended to better specify other wave attenuation mechanisms for integration in emerging coupled ocean-wave-sea ice models as the ones currently being developed in most weather forecasting agencies ((Kousa et al. 2026)), e.g. ECMWF.

Appendix A

The integral II in Eq. 2.7:

I=∫02π∣Usin⁡ϕ−v∣3 dϕ,(17)I = \int_{0}^{2\pi} \lvert U \sin\phi- v \rvert^{3}\,d\phi, \tag*{(17)}

is treated differently for ∣v∣≪aΩ\lvert v\rvert\ll a\Omega (or equivalently ∣v∣≪U\lvert v\rvert\ll U) and ∣v∣>aΩ\lvert v\rvert> a\Omega (or equivalently ∣v∣>U\lvert v\rvert> U).

Case ∣v∣≪U\lvert v\rvert\ll U.

We expand the integrand as:

∣Usin⁡ϕ−v∣3=∣Usin⁡ϕ∣3−3vU2sin⁡2ϕ sgn⁡(sin⁡ϕ)sin⁡ϕ+3v2U∣sin⁡ϕ∣+O(v3).(18)\lvert U\sin\phi-v\rvert^{3}=\lvert U\sin\phi\rvert^{3}-3vU^{2}\sin^{2}\phi\,\operatorname{sgn}(\sin\phi)\sin\phi+3v^{2}U\lvert\sin\phi\rvert+O(v^{3}). \tag*{(18)}

The second term has zero contribution towards II. Therefore, we approximate the integral as a polynomial in UU:

I≈∫02π(U3∣sin⁡ϕ∣3+3Uv2∣sin⁡ϕ∣)dϕ=83U3+12Uv2(19)I \approx\int_{0}^{2\pi}\left(U^{3}\lvert\sin\phi\rvert^{3}+3Uv^{2}\lvert\sin\phi\rvert\right)d\phi=\frac{8}{3}U^{3}+12Uv^{2} \tag*{(19)}

While strictly valid only for ∣v∣≪U\lvert v\rvert\ll U we use this approximation for ∣v∣≤U\lvert v\rvert\leq U.

Case: ∣v∣>U\lvert v\rvert> U

The integrand only depends on the sign of vv. Therefore it can be written as:

∣Usin⁡ϕ−v∣3=sign⁡(v)(v−Usin⁡ϕ)3.(20)\lvert U\sin\phi-v\rvert^{3}=\operatorname{sign}(v)\left(v-U\sin\phi\right)^{3}. \tag*{(20)}

Note that some of the terms have zero contribution towards II due to symmetry. We write the integral as a polynomial:

I=∫02π(∣v∣3+3∣v∣U2sin⁡2ϕ)dϕ(21)I=\int_{0}^{2\pi}\left(\lvert v\rvert^{3}+3\lvert v\rvert U^{2}\sin^{2}\phi\right)d\phi \tag*{(21)}
I=2π∣v∣3+3πU2∣v∣(22)I=2\pi\lvert v\rvert^{3}+3\pi U^{2}\lvert v\rvert \tag*{(22)}

Appendix 2

The integral approximation II is piecewise, i.e. depends on the relationship between UU and vv. As a result, different governing equations emerge for solutions in region A, corresponding to aΩ≥∣v∣a\Omega\geq\lvert v\rvert, and region B for aΩ<∣v∣a\Omega< \lvert v\rvert. We first solve the governing equation for aa in region A and then impose continuity on the amplitude to obtain the solution in region B.

Region A.

In Region A, defined as the region in which orbital velocity is greater than the drift (aΩ≥∣v∣a\Omega\geq\lvert v\rvert), if it exists we impose the initial condition a(x^=0)=a0a(\hat{x}=0)=a_{0}.

The governing equation in its differential and integral form is:

daAdx^=−αaA−83ΓΩ3aA2−12ΓΩ∣v∣2≡−f(a),(23)\frac{d a_{A}}{d\hat{x}}=-\alpha a_{A}-\frac{8}{3}\Gamma\Omega^{3}a_{A}^{2}-12\Gamma\Omega\lvert v\rvert^{2}\equiv-f(a), \tag*{(23)}
∫a0aAda′83ΓΩ3a′2+αa′+12ΓΩ∣v∣2=−x^.(24)\int_{a_{0}}^{a_{A}}\frac{d a'}{\frac{8}{3}\Gamma\Omega^{3}a'^{2}+\alpha a'+12\Gamma\Omega\lvert v\rvert^{2}}=-\hat{x}. \tag*{(24)}

The sign of the discriminant of ff, denoted as Δ\Delta, sets the form of solution. The roots of ff are r±r_{\pm}.

The solutions are categorised according to Δ\Delta:

aA(x^)={3Δ16ΓΩ3tan⁡(−Δ2x^+arctan⁡(16ΓΩ3a0+3α3Δ))−3α16ΓΩ3,Δ>0,32ΓΩ3a0−(16αΓΩ3a0+3α)x^32ΓΩ3+(32ΓΩ3a0+6α)x^,Δ=0,r−(a0−r+)exp⁡(8ΓΩ33(r−−r+)x^)−r+(a0−r−)(a0−r+)exp⁡(8ΓΩ33(r−−r+)x^)−(a0−r−),Δ<0.(25)a_A(\hat{x}) = \begin{cases} \dfrac{3\sqrt{\Delta}}{16\Gamma\Omega^3}\tan\left(-\dfrac{\sqrt{\Delta}}{2}\hat{x}+\arctan\left(\dfrac{16\Gamma\Omega^3 a_0+3\alpha}{3\sqrt{\Delta}}\right)\right)-\dfrac{3\alpha}{16\Gamma\Omega^3}, & \Delta>0,\\ \dfrac{32\Gamma\Omega^3a_0-(16\alpha\Gamma\Omega^3a_0+3\alpha)\hat{x}}{32\Gamma\Omega^3+(32\Gamma\Omega^3a_0+6\alpha)\hat{x}}, & \Delta=0,\\ \dfrac{r_-(a_0-r_+)\exp\left(\dfrac{8\Gamma\Omega^3}{3}(r_- - r_+)\hat{x}\right)-r_+(a_0-r_-)}{(a_0-r_+)\exp\left(\dfrac{8\Gamma\Omega^3}{3}(r_- - r_+)\hat{x}\right)-(a_0-r_-)}, & \Delta<0. \end{cases} \tag*{(25)}

Because the relationship between Δ\Delta and δ\delta, we can rewrite the condition as function of δ\delta as in Eq. (9).

We define x^∗\hat{x}^* as the location where U(x^∗)=∣v∣U(\hat{x}^*)=|v|, or equivalently a(x^∗)Ω=∣v∣a(\hat{x}^*)\Omega=|v|. Therefore:

x^∗=max⁡{a−1(∣v∣Ω),0},(26)\hat{x}^*=\max\left\{a^{-1}\left(\frac{|v|}{\Omega}\right),0\right\}, \tag*{(26)}

denotes the extent of region A.

Region B

In region B we enforce continuity of the amplitude at x^∗\hat{x}^*. Note that if the orbital velocity at the ice edge is lower that the drift only region B exists. Using the integral approximation, the differential and integral form of the governing equation are:

daBdx^=−(α+3πΓΩ2∣v∣)aB−2πΓ∣v∣3aB,(27)\frac{d a_B}{d\hat{x}}=-\left(\alpha+3\pi\Gamma\Omega^2|v|\right)a_B-\frac{2\pi\Gamma|v|^3}{a_B}, \tag*{(27)}
∫aA(x∗)aBa′(α+3πΓΩ2∣v∣)a′2+2πΓ∣v∣3 da′=x∗−x^.(28)\int_{a_A(x^*)}^{a_B}\frac{a'}{\left(\alpha+3\pi\Gamma\Omega^2|v|\right)a'^2+2\pi\Gamma|v|^3}\,da' = x^*-\hat{x}. \tag*{(28)}

The integrand yields:

aB(x^)=[e−(6πΩ2∣v∣Γ+2α)(x^−x^∗)(v2Ω2+2π∣v∣3Γ3πΩ2∣v∣Γ+α)−2π∣v∣3Γ3πΩ2∣v∣Γ+α]12,x^≥x^∗.(29)a_B(\hat{x})=\left[e^{-(6\pi\Omega^2|v|\Gamma+2\alpha)(\hat{x}-\hat{x}^*)}\left(\frac{v^2}{\Omega^2}+\frac{2\pi|v|^3\Gamma}{3\pi\Omega^2|v|\Gamma+\alpha}\right)-\frac{2\pi|v|^3\Gamma}{3\pi\Omega^2|v|\Gamma+\alpha}\right]^{\frac{1}{2}},\qquad\hat{x}\ge\hat{x}^*. \tag*{(29)}

Acknowledgements

Acknowledgements. This work was supported by the Natural Environment Research Council and the ARIES Doctoral Training Partnership [grant number NE/S007334/1]. RR and AA acknowledge funding from Royal Society (IEC\R3\243016). This work was stimulated by discussions during the “Maths of Sea Ice” meeting, made possible by the Isaac Newton Institute for Mathematical Sciences. DP is supported by EPSRC Grant No. EP/Y021118/1 and by the ExtreMe Matter Institute EMMI at the GSI Helmholtzzentrum fuer Schwerionenphysik, Darmstadt, Germany.

Declaration of interests

Declaration of interests. The authors report no conflict of interest.

References

  1. [1]ALBERELLO, ALBERTO, BENNETTS, LUKE, HEIL, PETRA, EAYRS, CLARE, VICHI, MARCELLO, MACHUTCHON, KEITH, ONORATO, MIGUEL & TOFFOLI, ALESSANDRO 2020 Drift of pancake ice floes in the winter antarctic marginal ice zone during polar cyclones. Journal of Geophysical Research: Oceans 125 (3), e2019JC015418, e2019JC015418 10.1029/2019JC015418.DOI
  2. [2]ALBERELLO, A., BENNETTS, L. G., ONORATO, M., VICHI, M., MACHUTCHON, K., EAYRS, C., NTAMBA, B. N., BENETAZZO, A., BERGAMASCO, F., NELLI, F., PATTANI, R., CLARKE, H., TERSIGNI, I. & TOFFOLI, A. 2022 Three-dimensional imaging of waves and floes in the marginal ice zone during a cyclone. Nature Communications 13 (4590), 1–11.DOI
  3. [3]Alberello, A., Onorato, M., Bennetts, L., Vichi, M., Eayrs, C., MacHutchon, K. & Toffoli, A. 2019 Brief communication: Pancake ice floe size distribution during the winter expansion of the Antarctic marginal ice zone. The Cryosphere 13 (1), 41–48.DOI
  4. [4]Barrell, Christopher, Renfrew, Ian A, Bennett, Miriam, Elvidge, Andrew D, Weiss, Alexandra, Methven, John, Blockley, Ed, Copsey, Dan, Graham, Tim & Krumpen, Thomas 2025 Rapid summertime sea ice melt in a coupled numerical weather prediction system. Journal of Advances in Modeling Earth Systems 17 (10), e2025MS004945.DOI
  5. [5]Bennetts, Luke G., Shakespeare, Callum J., Vreugdenhil, Catherine A., Foppert, Annie, Gayen, Bishakhdatta, Meyer, Amelie, Morrison, Adele K., Padman, Laurie, Phillips, Helen E., Stevens, Craig L., Toffoli, Alessandro, Constantinou, Navid C., Cusack, Jesse M., Cyriac, Ajitha, Doddridge, Edward W., England, Matthew H., Evans, D. Gwyn, Heil, Petra, Hogg, Andrew McC., Holmes, Ryan M., Huneke, Wilma G. C., Jones, Nicole L., Keating, Shane R., Kiss, Andrew E., Kraitzman, Noa, Malyarenko, Alena, McConnochie, Craig D., Meucci, Alberto, Montiel, Fabien, Neme, Julia, Nikurashin, Maxim, Patel, Ramkrushnbhai S., Peng, Jen-Ping, Rayson, Matthew, Rosevear, Madelaine G., Sohail, Taimoor, Spence, Paul & Stanley, Geoffrey J. 2024 Closing the loops on Southern Ocean dynamics: From the circumpolar current to ice shelves and from bottom mixing to surface waves. Reviews of Geophysics 62 (3), e2022RG000781, e2022RG000781 2022RG000781.DOI
  6. [6]Brouwer, Jill, Fraser, Alexander D, Murphy, Damian J, Wongpan, Pat, Alberello, Alberto, Kohout, Alison, Horvat, Christopher, Wotherspoon, Simon, Massom, Robert A, Cartwright, Jessica & others 2022 Altimetric observation of wave attenuation through the Antarctic marginal ice zone using ICESat-2. The Cryosphere 16 (6), 2325–2353.DOI
  7. [7]Day, Jonathan, Keely, Sarah, Arduino, Gabriele, Magnusson, Linus, Mogensen, Kristian, Rodwell, Mark, Sandu, Irina & Tietsche, Steffen 2022 Benefits and challenges of dynamic sea-ice for weather forecasts. Weather and Climate Dynamics Discussions 2022, 1–28.DOI
  8. [8]Day, Noah S., Bennetts, Luke G., O’Farrell, Siobhan P., Alberello, Alberto & Montiel, Fabien 2024 Analysis of the Antarctic marginal ice zone based on unsupervised classification of standalone sea ice model data. Journal of Geophysical Research: Oceans 129 (8), e2024JC020953, e2024JC020953 2024JC020953.DOI
  9. [9]Derkani, Mikhail Z. H., Alberello, Alberto, Nelli, Filippo, Bennetts, Luke G, Hessner, Katrin G, MacHutchon, Keith, Reichert, Konny, Aouf, Lotfi, Khan, Salman Saeed & Toffoli, Alessandro 2020 Wind, waves, and surface currents in the Southern Ocean: observations from the Antarctic Circumnavigation Expedition. Earth System Science Data Discussions 2020, 1–22.DOI
  10. [10]Elvidge, AD, Renfrew, IA, Weiss, AI, Brooks, IM, Lachlan-Cope, TA & King, JC 2016 Observations of surface momentum exchange over the marginal ice zone and recommendations for its parametrisation. Atmospheric Chemistry and Physics 16 (3), 1545–1563.
  11. [11]Fox, Colin & Squire, Vernon A 1994 On the oblique reflexion and transmission of ocean waves at shore fast sea ice. Philosophical Transactions of the Royal Society of London. Series A: Physical and Engineering Sciences 347 (1682), 185–218.DOI
  12. [12]Herman, Agnieszka, Cheng, Sukün & Shen, Hayley H 2019 Wave energy attenuation in fields of colliding ice floes–part 1: Discrete-element modelling of dissipation due to ice–water drag. The Cryosphere 13 (11), 2887–2900.DOI
  13. [13]Holland, Paul R & Kwok, Ron 2012 Wind-driven trends in Antarctic sea-ice drift. Nature Geoscience 5 (12), 872–875.
  14. [14]IPCC 2021 Climate Change 2021: The Physical Science Basis. Contribution of Working Group I to the Sixth Assessment Report of the Intergovernmental Panel on Climate Change. Cambridge, UK and New York, NY, USA: Cambridge University Press.
  15. [15]Keller, Joseph B 1998 Gravity waves on ice-covered water. Journal of Geophysical Research: Oceans 103 (C4), 7663–7669.DOI
  16. [16]Kohout, Alison L, Meylan, Michael H & Plew, David R 2011 Wave attenuation in a marginal ice zone due to the bottom roughness of ice floes. Annals of Glaciology 52 (57), 118–122.DOI
  17. [17]Kohout, Alison L., Smith, Madison, Roach, Lettie A., Williams, Guy, Montiel, Fabien & Williams, Michael J. M. 2020 Observations of exponential wave attenuation in Antarctic sea ice during the PIPERS campaign. Annals of Glaciology 61 (82), 196–209.DOI
  18. [18]Kohout, Alison L, Williams, MJM, Dean, SM & Meylan, MH 2014 Storm-induced sea-ice breakup and the implications for ice extent. Nature 509 (7502), 604–607.DOI
  19. [19]Kousal, Josh, Pelletier, Charles, Denissen, Jasper M.C., Schulte, Luise, Keeley, Sarah, Dueben, Peter, Penny, Stephen G., Park, Rae-Seol, Miralles, Diego G., Shapkalievskiy, Metodija M., Valmassoi, Arianna, Renfrew, Ian, Zampieri, Lorenzo, Kumar, Rajesh, Ruiz-Vásquez, Melissa, Schulz, Jan-Peter, Wu, Lichuan & Bagazgotta, Xabier Pedruzo 2026 Numerical weather prediction model coupling—strategies, challenges, and outlook. Bulletin of the American Meteorological Society 107 (1), E183 – E189.DOI
  20. [20]Kwok, Ron, Pang, Shirley S & Kacimi, Sahra 2017 Sea ice drift in the Southern Ocean: Regional patterns, variability, and trends. Elem Sci Anth 5, 32.DOI
  21. [21]Landwehr, S. & al. 2021 Exploring the coupled ocean and atmosphere system with a data science approach applied to observations from the Antarctic Circumnavigation Expedition. Earth System Dynamics 12 (4), 1295–1369.DOI
  22. [22]McPhee, Miles G 1979 The effect of the oceanic boundary layer on the mean drift of pack ice: Application of a simple model. Journal of Physical Oceanography 9 (2), 388–400.doi.org/10.1175/1520-0485(1979)009<0388:teotob>2.0.co;2
  23. [23]Meylan, M. H., Bennetts, L. G., Mosig, J. E. M., Rogers, W. E., Doble, M. J. & Peter, M. A. 2018 Dispersion relations, power laws, and energy loss for waves in the marginal ice zone. Journal of Geophysical Research: Oceans 123 (5), 3322–3335.DOI
  24. [24]Montiel, Fabien, Forbes, Martin, Echevarria, Emilio, Rapizo, Henrique & Gamble, Carlo 2025 Evaluation of ice dissipation parameterizations in spectral ocean wave model WAVEWATCH III: An intercomparison analysis. Journal of Geophysical Research: Oceans 130 (9), e2024JC022113, e2024JC022113.DOI
  25. [25]Montiel, Fabien, Kohout, Alison L. & Roach, Lettie A. 2022 Physical drivers of ocean wave attenuation in the marginal ice zone. Journal of Physical Oceanography 52 (5), 889 – 906.DOI
  26. [26]Robinson, NJ & Palmer, SC 1990 A modal analysis of a rectangular plate floating on an incompressible liquid. Journal of sound and vibration 142 (3), 453–460.
  27. [27]Selivanova, Julia, Iovino, DoroteaCiro & Cocetta, Francesco 2024 Past and future of the Arctic sea ice in high-resolution model intercomparison project (HighResMIP) climate models. The Cryosphere 18 (6), 2739–2763.DOI
  28. [28]Shen, Hayley H & Squire, Vernon A 1998 Wave damping in compact pancake ice fields due to interactions between pancakes. Antarctic Sea Ice: Physical Processes, Interactions and Variability 74, 325–341.
  29. [29]Smith, Madison & Thomson, Jim 2020 Pancake sea ice kinematics and dynamics using shipboard stereo video. Annals of Glaciology 61 (82), 1–11.DOI
  30. [30]Stopa, Justin E., Sutherland, Peter & Ardhuin, Fabrice 2018 Strong and highly variable push of ocean waves on Southern Ocean sea ice. Proceedings of the National Academy of Sciences 115 (23), 5861–5865.
  31. [31]Vichi, M. 2022 An indicator of sea ice variability for the Antarctic marginal ice zone. The Cryosphere 16 (10), 4087–4106.DOI
  32. [32]Vichi, Marcello, Eayrs, Clare, Alberello, Alberto, Bekker, Anriëtte, Bennetts, Luke, Holland, David, de Jong, Ehlke, Joubert, Warren, Machutchon, Keith, Messori, Gabriele, Mojica, Jhon F., Onorato, Miguel, Saunders, Clinton, Skatulla, Sebastian & Toffoli, Alessandro 2019 Effects of an explosive polar cyclone crossing the antarctic marginal ice zone. Geophysical Research Letters 46 (11), 5948–5958.
  33. [33]Voermans, Joey J, Fraser, Alex, Alexander D. Brouwer, Jill, Meylan, Michael H. Liu, Qingxiang & Babanin, Alexander V 2025 Finely resolved along-track wave attenuation estimates in the Antarctic marginal ice zone from ICESat-2. The Cryosphere 19 (8), 3381–3395.DOI
  34. [34]Wadhams, Peter 1978 Wave decay in the marginal ice zone measured from a submarine. Deep Sea Research 25 (1), 23–40.DOI
  35. [35]Wadhams, Peter 1986 The Seasonal Ice Zone, pp. 825–991. Boston, MA: Springer US.DOI
  36. [36]Wamser, Christian & Martinson, Douglas C 1993 Drag coefficients for winter Antarctic pack ice. Journal of Geophysical Research: Oceans 98 (C7), 12431–12437.DOI
  37. [37]Wang, Ruixue & Shen, Hayley H 2010 Gravity waves propagating into an ice-covered ocean: A viscoelastic model. Journal of Geophysical Research: Oceans 115 (C6).DOI
  38. [38]Weber, Jan Erik 1987 Wave attenuation and wave drift in the marginal ice zone. Journal of physical oceanography 17 (12), 2351–2361.doi.org/10.1175/1520-0485(1987)017<2351:waawdi>2.0.co;2
  39. [39]Womack, Ashleigh, Alberello, Alberto, de Vos, Marc, Toffoli, Alessandro, Verrinder, Robyn & Vichi, Marcello 2024 A contrast in sea ice drift and deformation between winter and spring of 2019 in the Antarctic marginal ice zone. The Cryosphere 18 (1), 205–229.DOI
  40. [40]Womack, Ashleigh, Vichi, Marcello, Alberello, Alberto & Toffoli, Alessandro 2022 Atmospheric drivers of a winter-to-spring lagrangian sea-ice drift in the eastern antarctic marginal ice zone. Journal of Glaciology 68 (271), 999–1013.DOI

Paper details

vlnc:W_rhyb53ap
Version
v1 · latest
Published
Fresh — added to the Hub within the last 30 days
Added
License
CC BY 4.0
Usage

views · 0

downloads · 0

A view counts once per browser per day, a download each time the PDF is requested.

  • Physics and Astronomy

Not recorded for this paper

View statement
Looking for related papers…
Contents