Chris Petrich

Radiative transfer in sea ice is subject to anisotropic, multiple scattering. The impact of anisotropy on the light field under sea ice was found to be substantial and has been quantified. In this study, a large data set of irradiance and radiance measurements under sea ice has been acquired with a Remotely Operated Vehicle (ROV) in the central Arctic. Measurements are interpreted in the context of numerical radiative transfer calculations, laboratory experiments, and microstructure analysis. The ratio of synchronous measurements of transmitted irradiance to radiance shows a clear deviation from an isotropic under‐ice light field. We find that the angular radiance distribution under sea ice is more downward directed than expected for an isotropic light field. This effect can be attributed to the anisotropic scattering coefficient within sea ice. Assuming an isotropic radiance distribution under sea ice leads to significant errors in light‐field modeling and the interpretation of radiation measurements. Quantification of the light field geometry is crucial for correct conversion of radiance data acquired by Autonomous Underwater Vehicles (AUVs) and ROVs.

The amount of solar radiation transmitted through Arctic sea ice is determined by the thickness and physical properties of snow and sea ice. Light transmittance is highly variable in space and time since thickness and physical properties of snow and sea ice are highly heterogeneous on variable time and length scales. We present field measurements of under-ice irradiance along transects under undeformed land-fast sea ice at Barrow, Alaska (March, May, and June 2010). The measurements were performed with a spectral radiometer mounted on a floating under-ice sled. The objective was to quantify the spatial variability of light transmittance through snow and sea ice, and to compare this variability along its seasonal evolution. Along with optical measurements, snow depth, sea ice thickness, and freeboard were recorded, and ice cores were analyzed for chlorophyll a and particulate matter. Our results show that snow cover variability prior to onset of snow melt causes as much relative spatial variability of light transmittance as the contrast of ponded and white ice during summer. Both before and after melt onset, measured transmittances fell in a range from one third to three times the mean value. In addition, we found a twentyfold increase of light transmittance as a result of partial snowmelt, showing the seasonal evolution of transmittance through sea ice far exceeds the spatial variability. However, prior melt onset, light transmittance was time invariant and differences in under-ice irradiance were directly related to the spatial variability of the snow cover.

The location of snow dunes over the course of the ice‐growth season 2007/08 was mapped on level landfast first‐year sea ice near Barrow, Alaska. Landfast ice formed in mid‐December and exhibited essentially homogeneous snow depths of 4–6 cm in mid‐January; by early February distinct snow dunes were observed. Despite additional snowfall and wind redistribution throughout the season, the location of the dunes was fixed by March, and these locations were highly correlated with the distribution of meltwater ponds at the beginning of June. Our observations, including ground‐based light detection and ranging system (lidar) measurements, show that melt ponds initially form in the interstices between snow dunes, and that the outline of the melt ponds is controlled by snow depth contours. The resulting preferential surface ablation of ponded ice creates the surface topography that later determines the melt pond evolution.

Seasonal breakup of landfast sea ice consists of movement and irreversible ice detachment in response to winds or oceanic forces in the late stages of ice decay. The breakup process of landfast sea ice in the Chukchi Sea at Barrow, Alaska, was analyzed for the years 2000 through 2010 on the basis of local observations of snow and ice conditions, weather records, image sequences obtained from cameras, coastal X band marine radar, and satellite imagery. We investigated the relation of breakup to winds, tides, and nearshore current measurements from a moored acoustic Doppler current profiler. Two breakup modes are distinguished at Barrow on the basis of the degree of ice decay. Mechanical breakup due to wind and oceanic forces follows ablation and weakening of the ice. Thermal breakup is the result of ice disintegration under melt ponds, requiring little force to induce dispersion. Grounded pressure ridges are pivotal in determining the breakup mode. The timing of thermal breakup of the nearshore ice cover was found to correlate with the measured downwelling solar radiation in June and July. This linkage allows for the development of an operational forecast of landfast ice breakup. Results from forecasts during 2 years demonstrate that thermal breakup can be predicted to within a couple of days 2 weeks in advance. The cumulative shortwave energy absorbed by the ice cover provides for a measure of the state of ice decay and potential for disintegration. Discriminating between the two modes of breakup bears the potential to greatly increase forecasting skill.