High-resolution measurement of sea ice mechanical characteristics using Distributed Acoustic Sensing

TL;DR

The study demonstrates that Distributed Acoustic Sensing (DAS) can map spatial variations in sea ice mechanical properties along a 600 m fiber by jointly analyzing active-source and swell-induced hydroelastic signals. Using active hammer and jump sources, the ice Young's modulus $E$ and thickness $h$ are inverted from dispersion curves via the relation $E = \rho (1-\nu^2) \frac{k_L^2}{\omega^2}$ and flexural-wave fitting, achieving $E$ values of $4.5$–$5.7\ \mathrm{GPa}$ and $h$ from $0.30$ to $0.67\ \mathrm{m}$. Passive swell analysis with Continuous Wavelet Transform yields local wavenumbers and the hydroelastic dispersion $\omega^2 = \left(gk + \frac{D}{\rho_w} k^5\right) \tanh(Hk)$, enabling estimation of flexural rigidity $D$ that varies across three ice morphologies. The results are in good agreement with geophone and drill-hole measurements, illustrating DAS as an effective tool for high-resolution, spatially resolved sea ice property mapping and history of formation and dynamics, albeit with coupling and power challenges for longer deployments.

Abstract

Sea ice mechanical properties are involved in dynamical processes acting from the scale of meters to several hundred kilometers. The current rapid changes in the state of polar sea ice require a better understanding and modeling of these processes and, therefore, accurate measurements of properties including sea ice thickness, density, Young's modulus and Poisson's ratio. These properties can be measured by tracking the propagation of elastic waves within the ice. Recent technological advances have enabled the use of fiber-optic cables as cost-effective, dense seismic arrays. Once connected to an interrogator unit and mechanically coupled to a medium, here the ice cover, these cables can monitor strain field propagation, using a technique called Distributed Acoustic Sensing (DAS). In this work, we describe the use of such an array of sensors in the coastal ice of the St. Lawrence Estuary, Canada, where a 600 m long optical fiber was deployed across three different morphological sea ice conditions. During hour-long recordings, we measured the propagation of both multi-modal seismic signals generated by active sources and hydro-elastic swell. We computed dispersion curves of active signals and used Continuous Wavelet Transform (CWT) to observe the evolution of swell characteristics in the different ice areas. The dispersion curves were successfully inverted to measure the spatial evolution of ice thickness, and Young's and flexural rigidity in each of these areas. We observed ice thicknesses from 25 cm to 68 cm and Young's modulus values between 4.5 GPa and 5.7 GPa, in good agreement with values derived from collocated geophone arrays and drill hole thickness measurements. DAS systems therefore appear to be effective in evaluating heterogeneous sea ice mechanical properties and thus sea ice formation history and dynamics.

Figures (5)

• Figure 1: (a) Bathymetry of Baie du Ha! Ha! relative to the tidal water chart datum, as obtained from the 10-m resolution non-navigational data product of the Canadian Hydrographic Service (CHS NONNA10). The black solid line shows the location of the optical fiber, coordinate system $(X,Y)$ used in (b) is also represented and accurately oriented. (b) A georectified aerial picture of Baie du Ha! Ha! taken on 5 February 2025 at an altitude of 111 m overlaid with drill hole thickness and geophones measurements made on 10 and 11 February, the day the 600-m long optical fiber was installed (black line). The cartesian coordinate system is associated with the DAS fiber with the origin located at one end and the fiber at $Y=0$. Vertical dotted lines separate three regions with different morphologies and histories.
• Figure 2: a) A 10 minute-long space-time plot of the strain rate signal measured on 11 February 2025. b) Zoom on the space-time strain rate signal corresponding to active noise generation following the ZEN procedure. c) Zoom on the space-time strain rate signal related to the generation of a flexural out-of-plane wave (Z) at $X=$83 m.
• Figure 3: Illustration of the inversion process for obtaining the sea ice flexural rigidity $D$ from the hydro-elastic signal induced by swell. First, the strain rate signal is split into each frequency component (a). Then a Continuous Wavelet Transform (CWT) is computed for each frequency component, leading to a scaleogram for each frequency (b). Finally, a section of all scaleograms (shown in red in panel b) for a given position $x$ is selected in order to get the space-time spectrum at this position (c). The hydro-elastic dispersion relation is then obtained by fitting the maxima of the space-time spectrum (red dots) for each position along the fiber.
• Figure 4: Space-time spectra $|\hat{\dot{\epsilon}}|(k,f)$ obtained at $x=$ 80 m from active seismic (a), and passive hydroelastic signals (b). Subpixel position of each spectrum maxima are detected and represented as color dots. Combination of these two sets of $(k,f)$ components are shown in (c). The dash line represents the hydro-elastic dispersion relation fitted in the $(k,f)$ space obtained from active seismic noise (blue dots).
• Figure 5: Variations of (a) effective Young's modulus $E$, (b) ice thickness $h$ and (c) flexural rigidity $D$ along the fiber. In each subplot, vertical dotted lines represent the frontier between the different ice regions identified in Figure \ref{['fig:situation_map_bathymetry']}. On 11 February, longitudinal in-plane waves created by active sources could not be measured efficiently, which explains why only the ice thickness could be inverted on this day with the active method.