Methodology:

Passive seismic surveys have been conducted in order to map and characterize the spatial distribution of subsurface layers across the site at Peninsula Point. Three-dimensional wave signals within the earths sub-surface propagate through the near surface layers and notable differences between amplitudes of horizontal and vertical motions are caused by acoustic impedance contrasts at layer boundaries. The spectral ratio of horizontal over vertical responses exhibit resonance frequencies (fr) that can be directly linked to the layer thickness (h). The layer thicknesses are estimated by the quarter wavelength resonance approximation. Measurements of visible contact layer depths exposed in permafrost cliff sections adjacent to a set of control reading locations have been used to characterize relative acoustic impedance contrast signatures and to calculate shear wave velocity (Vs) through each of the main material layers (see supporting information). This has enabled the determination of two distinct high frequency signal peaks associated with specific ground layer controls.

A single peak in the resonance frequency (usually around 20 Hz) denoted active layer or slurry contact directly onto massive ice, whereas traces with a high frequency peak around 30 - 50 Hz (usually with additional lower frequency peaks of 1 - 5 Hz) represented a permafrost to ice contact. Sub-divisions were also identified: Group 1 = single distinct peak representing a shallow contact directly on to ice (no lower frequency acoustic impedance contrast); Group 2 = large peak then a secondary lower peak that corresponds with the permafrost-ice contact and then the base of the massive ice (from boreholes we know this to mostly a sand layer) respectively; Group 3 = a muted high frequency peak that then curves up to a low frequency base line; Group 4 = two comparable peaks, which may indicate a lower contrast in the impedance so potentially frozen sand layers, although there is no validation for this interpretation. Once the impedance contrast signatures were classified, the following formula was applied to calculate and validate layer boundary depths (h):

h = Vs / 4fr

A shear-wave velocity of 920 ms-1 for the permafrost was determined from cliff outcrop data at the site and was used to derive depth to the massive ice contact. Validation sites have also been used to make sure that layer properties were sufficiently consistent over the survey area (see supporting information) and active layer probe depths have been used to map ice surface where it rose to very near the ground surface (and hence no permafrost was present). The 2018 massive ground ice surface has been interpolated to form a massive ice surface relative to sea-level (Ice_Surface.tif).

Data collection:

Tromino® 3G ENGY instruments have been used to record three component accelerometer measurements of ambient noise and have been processed in Grilla Software.

Data quality:

Surface to ice surface measurements from exposed ice cliffs have been used to calibrate and validate the shear wave velocities used to derive surface boundaries.