Morphology of sea ice pressure ridges in the northwestern Weddell Sea in winter

Abstract

[1] To investigate the morphology and distribution of pressure ridges in the northwestern Weddell Sea, ice surface elevation profiles were measured by a helicopter-borne laser altimeter during Winter Weddell Outflow Study with the German R/V Polarstern in 2006. An optimal cutoff height of 0.62 m, derived from the best fits between the measured and theoretical ridge height and spacing distributions, was first used to separate pressure ridges from other sea ice surface undulations. It was found that the measured ridge height distribution was well modeled by a negative exponential function, and the ridge spacing distribution by a lognormal function. Next, based on the ridging intensity R i (the ratio of mean ridge sail height to mean spacing), all profiles were clustered into three regimes by an improved k-means clustering algorithm: R i ≤ 0.01, 0.01 < R i ≤ 0.026, and R i > 0.026 (denoted as C 1 , C 2 , and C 3 respectively). Mean (and standard deviation) of sail height was 0.99 (AE0.07) m in Regime C 1 , 1.12 (AE0.06) m in C 2 , and 1.17 (AE0.04) m in C 3 , respectively, while the mean spacings (and standard deviations) were 232 (AE240) m, 54 (AE20) m, and 31 (AE5.6) m. These three ice regimes coincided closely with distinct sea ice regions identified in a satellite radar image, where C 1 corresponded to the broken ice in the marginal ice zone and level ice formed in the Larsen Polynya, C 2 corresponded to the deformed first-and second-year ice formed by dynamic action in the center of the study region, and C 3 corresponded to heavily deformed ice in the outflowing branch of the Weddell Gyre. The results of our analysis showed that the relationship between the mean ridge height and frequency was well modeled by a logarithmic function with a correlation coefficient of 0.8, although such correlation was weaker when considering each regime individually. The measured ridge height and frequency were both greater than those reported by others for the Ross Sea. Compared with reported values for other parts of the Antarctic, the present ridge heights were greater, but the ridge frequencies and ridging intensities were smaller than the most extreme of them. Meanwhile, average thickness of ridged ice in our study region was significantly larger than that of the Coastal Ross Sea showing the importance of deformation and ice age for ice conditions in the northwestern Weddell Sea.

Characterizing Arctic sea ice topography using high-resolution IceBridge data

“…An ordinary exponential distribution of sail heights was proposed by Wadhams (1980), which has been validated (to varying degrees) by further observations of sail/feature height (e.g. Tucker et al, 1979;Dierking, 1995;Martin, 2007;Rabenstein et al, 2010;Tan et al, 2012). As discussed earlier (Sect.…”

“…The choice of cut-off height can provide a significant impact on the sail/feature height distributions (e.g. Tan et al, 2012) and should be considered when analysing the surface feature data derived in this study.…”

“…In this study we employ the elevation threshold approach, which has been used extensively in previous studies (e.g. Wadhams, 1980;Dierking, 1995;Martin, 2007;Tan et al, 2012;Castellani et al, 2014). Typically, a ridge (or surface feature) is detected if it has a height above the local level ice/snow surface greater than a chosen elevation threshold.…”

The Weddell Gyre, Southern Ocean: Present Knowledge and Future Challenges

“…Upward‐looking sonars on moorings (e.g., Behrendt et al, ; Harms et al, ) and with autonomous vehicles (e.g., Dowdeswell et al, ) have been used to study long‐term changes of sea‐ice flux and thermodynamic processes in the WG. Airborne electromagnetic (EM) induction sounding provides coincident thickness and roughness information (Haas et al, ; Tan et al, ). Very promising efforts to obtain better snow thickness measurements have been presented by Kwok and Maksym (), using an airborne broadband Frequency‐Modulated Continuous Wave radar.…”

A Variational Method for Sea Ice Ridging in Earth System Models

“…The statistical ridging relationship that emerges from is not an empiricism but instead a derived physical property. The power law suggested by that equation is not comparable with the log‐normal spacing of ridges commonly measured over A ( x ) (e.g., Davis & Wadhams, ; Tan et al, ; Timco & Burden, ; Wadhams & Davis, ) because gives the spatial occurrence of ridging at an instant, whereas observed log‐normal ridge distributions are a hysteresis. Rather, one expects at an instant that the spatial distribution of ridging will observe a power law due to the concentration of deformation at the edge of floes, which themselves observe spatial scaling by way of the floes size distribution (e.g., Paget et al, ; Perovich & Jones, ).…”