Snow and sea ice temperature profiles from satellite data and ice mass balance buoys

Abstract: The sea ice covers approximately 5% of the Earth’s surface at any given time and it plays an important role in the polar climate system affecting the heat, mass and momentum exchange between the atmosphere and the ocean. The snow cover on top of the sea ice affects its insulating and reflective properties and thus key figures in the climate system feedback loop. Sea ice and snow is of significant importance for our global climate system. However, it is difficult to effectively and accurately access data relating to snow and sea ice properties in the vast and remote Arctic region, especially during the winter, and snow is poorly constrained in current climate models. Improved information on snow and sea ice properties and thermodynamics from satellite observations could give valuable information in the process of validating, optimizing and improving these sea ice models and thereby the future predictions of sea ice growth and related climate variables. This project examines the possibility of deriving the temperature profile through the snow and ice layers, from the surface down to 0.5 m into the ice, from a combination of available satellite data. Satellite data used are thermal infrared (TIR) and microwave radiation at different wavelengths and polarisations. The satellite data are compared with coincident data from ice mass balance buoys (IMB) and numerical weather prediction (NWP) data. This combined dataset are analysed for possible and theoretically derived relationships between the satellite measurements and different snow and ice parameters. Different empirical models are used in this study to derive the mean snow temperature, snow density and snow and ice thickness, with various degree of success. It is clear that more advanced models are needed to accurately predict the observed variations of the snow and ice parameters. From the analysis it is clear that the satellite channels of lower frequencies are able to retrieve temperature measurements from deeper levels in the snow and ice than the higher frequencies. It is also clear that the satellite sensors are sensitive to changes in snow emissivity, associated with melting processes initiated by surface air temperatures around the freezing point, as the penetration depth is significantly decreased. The models derived in the multiple regression analysis, performed on one of the four IMB buoys available, show a higher level of confidence for the deeper levels in the sea ice. When the models are tested on the remaining three IMB buoys the correlation for the lower levels in the sea ice are stronger. The comparisons between measured and theoretically derived temperatures show a generally strong correlation with R2-values ranging from 0.43 to 0.90. It is evident that the models without TIR are superior to those including TIR measurements. The differences in correlation between the IMB buoys indicate a spatial dependency, as well as a strong dependency on differences in snow and ice thickness. The models derived in this study are based on conditions with relatively thick snow and ice covers. Further studies would need to be conducted in order to improve and generalize the models derived in this project, in order to implement the empirical models in operating, global sea ice models.