Integral equations for flexural-gravity waves: analysis and numerical methods
T. Askham, J.G. Hoskins, P. Nekrasov, M. Rachh · 2025
Read Integral equations for flexural-gravity waves: analysis and numerical methods by T. Askham, J.G. Hoskins, P. Nekrasov, M. Rachh (2025) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2501-00887.txt.
Introduction
Travis Askham, Jeremy Hoskins, Peter Nekrasov, and Manas Rachh build a fast solver for waves that bend thin ice plates floating on deep water, in a paper from 2025. These flexural-gravity waves arise around sea ice and ice shelves. Real plates vary in thickness, with ridges and rolls that rule out simple formulas.
The team recasts the problem as an integral equation and studies its math in depth. They solve it with a high-order method sped up by the fast Fourier transform. Tests on varied examples show the solver is fast, accurate, and able to scale to large problems. The work aims at realistic ice shapes that past methods could not handle.
Catalog record
| CREATOR | T. Askham, J.G. Hoskins, P. Nekrasov, M. Rachh |
|---|---|
| YEAR | 2025 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 117 KB |
| VERSION | 1.0 |
| LICENSE | http://creativecommons.org/licenses/by/4.0/ |
| RIGHTS | CC BY |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2501-00887
Browse the full library · About the Tower of Babel
Full text
The opening of the work, from the library's hosted text:
Integral equations for flexural-gravity waves: analysis and numerical methods T. Askham, J.G. Hoskins, P. Nekrasov, M. Rachh Abstract In this work, we develop a fast and accurate method for the scattering of flexural-gravity waves by a thin plate of varying thickness overlying a fluid of infinite depth. This problem commonly arises in the study of sea ice and ice shelves, which can have complicated heterogeneities that include ridges and rolls. With certain natural assumptions on the thickness, we present an integral equation formulation for solving this class of problems and analyze its mathematical properties. The integral equation is then discretized and solved using a high-order-accurate, FFT-accelerated algorithm. The speed, accuracy, and scalability of this approach are demonstrated through a variety of illustrative examples. 1 Introduction The motion of a thin plate coupled to a fluid is a subject of longstanding interest in the fields of geophysics and acoustics. In the polar regions, ice sheets floating atop the ocean are often modeled as thin plates that bend and flex in response to hydrodynamic forces. The coupling of plates to acoustic media results in similar wave phenomena and has important implications for aerodynamic stability [ 62 , 40 ] and sound radiation [ 38 , 54 ] . In complicated media, these waves exhibit dynamics that include branching [ 46 , 47 ] , bending [ 4 , 25 ] , and cloaking [ 55 , 87 ] . In the Arctic, ocean waves interact with sea ice in a dynamic region known as the marginal ice zone. Here, short period waves are attenuated by relatively thin ice, allowing longer periods to travel deeper into the ice pack [ 80 ] . The same is true in the Southern Ocean, where the opening of sea ice-free corridors allows the sea swell to make contact with Antarctic ice shelves [ 78 , 17 ] . Meanwhile, tsunami and infragravity waves, largely unattenuated by sea ice, are able to reach the ice shelf to induce significant flexure and calving [ 13 , 15 ] . While it is difficult to separate the impact of ocean wave forcing from thermodynamic processes, there is growing evidence to support the role of ocean waves in the breakup of Antarctic ice shelves and Arctic sea ice [ 5 , 49 , 48 , 81 ] . To weigh these factors, it is critical to understand and model interactions between ice and the ocean at various scales. The problem is inherently multi-scale, since sharp ice features like ridges account for a large portion of the total ice area [ 43 ] , and can significantly affect wave attenuation and propagation in the polar regions. In the small thickness limit, the ice-cover can be treated as the boundary of the fluid domain wherein the velocity potential satisfies a fourth order differential equation whose coefficients vary with the thickness of the ice-cover. Extensive work has been done to model the propagation of flexural-gravity waves across various obstructions in ice-covers of uniform thickness. These include narrow cracks [ 73 , 68 ] , wide cracks [ 22 ] , frozen leads [ 9 ] , and icebergs [ 75 ] ; see [ 76 ] for a comprehensive overview of these works. In this regime, semi-analytic solutions can be obtained using eigenfunction expansions [ 34 , 70 , 31 ] , Carlemann singular integral equations [ 18 ] , or the Wiener-Hopf method [ 50 , 21 , 20 , 8 ] . Interestingly, many of these works find that ice features act as low-pass filters, which may help to explain the attenuation patterns observed in the Arctic. Less attention has been given to the problem of wave scattering induced by continuous changes in the thickness of the ice, likely because they are more resistant to analytical techniques. Early attempts to solve this problem assume that ice features can be modeled as point inhomogeneities [ 3 , 59 ] or rely on the shallow-water approximation [ 58 ] . In three dimensions, this problem has been approximated using a ray method [ 42 ] , or by using a truncated modal decomposition of the vertical component of the velocity [ 67 ] ; though for the latter, numerical procedures have only been implemented in two dimensions [ 10 ] or for axisymmetric ice floes [ 11 ] . An alternate approach is to model the full-thickness of ice which has been used to study wave propagation in crevasses, see [ 71 , 61 , 63 ] , for example. However, these continuum models become prohibitively expensive in three dimensions, motivating the design of fast methods that work in the thin-plate regime. Perhaps the most closely related prior method to the present work is [ 83 ] . Using Green’s identities, the authors construct integral equations for unknowns supported on the water-ice interface to approximate the reflected and transmitted coefficients for incident plane waves in two dimensions. In this work, we derive a novel integral equation formulation for the fully three-dimensional scattering problem, assuming the thickness is smoothly varying and the region of inhomogeneity is compactly supported. We do this by reducing the problem to an associated boundary integral equation on the ice-covered surface. The reduction occurs in two steps. First, we use standard potential theory techniques to convert the three-dimensional boundary value problem into an integro-differential equation for an unknown defined on the surface. In the second step, we further reduce this integro-differential equation to an integral equation defined only on the compact support of the perturbation of the thickness. The resulting integral equation is an integral equation of the second kind and, for a large range of parameters, is provably well-posed. Conveniently, existing numerical methods can be easily adapted to discretize this integral equation with high order accuracy and it is straightforward to establish the convergence of the resulting scheme. Further, the discretized system is amenable to solution by fast algorithms, enabling the scalability of this approach to large scattering problems and (essentially) arbitrary accuracy. In the solution of the integro-differential equation resulting from the first step of the reduction, one obtains a slowly decaying surface density, a problem that often arises with infinite interfaces. An alternate strategy to the one presented here is the technique of PML-BIE [ 56 ] , or coordinate complexification [ 30 ] , which involves complexifying the boundary so that the density decays exponentially and the integration can be performed over a finite region. This method was recently applied in [ 27 ] for the solution of two-dimensional water waves with one-dimensional boundary, and we expect that the same can be readily extended to three-dimensions. For this approach, even though the coordinates are complex, high order quadrature schemes [ 45 ] and fast algorithms are available. In the present context, the approach adopted here has the advantage of being substantially simpler both to implement and analyze, as well as more amenable to acceleration via traditional fast algorithms. More broadly, integral equation methods of this flavor have been widely applied to scattering problems, and while we do not seek to review them here, we remark that these methods have appeared in several related contexts, including capillary