Multiseasonal, multifrequency, multipolarization, and multiresolution radar speckle and feature tracking for Arctic glacier velocities

The project RASTAR (Multiseasonal, multifrequency, multipolarization, and multiresolution RAdar Speckle and feature Tracking for ARctic glacier velocities) continued with technical developments and further worked on several major results: - A number of special radar image acquisitions have been programmed over Svalbard: Austfonna (TerraSAR-X; Radarsat Fine, Radarsat WideFine, CosmoSkymed), Kronebreen/Ny Ålesund (Radarsat Wide, Fine, WideFine) and entire Svalbard (Radarsat Wide). A number of algorithms were implemented to measure glacier displacements over time from repeat spaceborne radar images. A largely automated software and workflow was written to process large data volumes. Based on this work the following three major works were conducted: - The total mass balance of calving glaciers consist mainly of their surface mass balance, such as from snow accumulation and ice melt, and the calving flux, i.e. the ice transported through the glacier front into the sea. For Svalbard, the calving component is very uncertain but estimated to account for several ten percents of the total glacier mass balance. The large uncertainty in calving fluxes (two estimates exist so far, at 3 and 5.2 km3 per year) leads to a large uncertainty for estimates of total Svalbard glacier mass balance and thus sea level contribution. Within RASTAR, calving velocities of all calving glaciers have been estimated based on displacements from repeat Radarsat Wide images, arriving at a calving flux of 9 km3 per year, and thus exceeding previous estimates. These estimates are continuously refined in the project. One reason for the high estimate is the massively increased (but temporary) contributions of just two glaciers, Nathorstbreen and Basin-3 of Austfonna. - Basin-3 of Austfonna started a major increase in ice flux (surge) during the project period. Due to the massive impact of this instability on the mass balance and sea level contribution of Austfonna and Svalbard as a whole, special focus was given to this event. Special acquisitions of TerraSAR-X and Radarsat-2 high resolution radar data were programmed and are still obtained in regular intervals, and combined with automatic GPS stations on the ground. Displacement measurements based on the repeat radar data revealed a massive increases in surface velocity in 2012, with surface speeds of almost 20 metres per day (!) in winter 2012/2013. Currently, this one ice cap basin causes a sea-level contribution of 7.2 Gigatons/yr, that is about as much as the annual ice loss from the rest of the Svalbard archipelago, and amounts to about 3% of the current sea-level contribution from all glaciers and ice caps (except the Greenland and Antarctic ice sheets). - An in-depth study was conducted for Kronebreen and Kongsbreen (Ny Ålesund), two of the fastest flowing glaciers on Svalbard. An unprecedented series of monthly high-resolution radar data and automatic ground-GPS was compiled and analysed to gain new insights in the calving flux of glaciers in general, and the current response of Kronebreen and Kongsbreen in particular. We found maximal speeds close to the calving front of 3.2 metres/day at Kronebreen in July 2013 and 2.7 metres/day at Kongsbreen in December 2012. The inter-seasonal and inter-annual variations in surface velocities are high. The ice-flow variations are closely linked to the amount and timing of surface meltwater production and rainfall, both of which have a strong influence on the hydrological system of the glacier at its base. Kronebreen retreated up to 850 m (2.8 km2) and Kongsbreen up to 1800 m (2.5 km2). This makes both glaciers major contributors to the overall mass loss of the Svalbard archipelago through frontal ablation with shares of 4.0% (Kronebreen) and 2.5 % (Kongsbreen).

Glacier flow is a fundamental Earth surface process. Connecting ice accumulation and ablation, it is an integral element of glacier systems and their response to climate and its changes. Quantification of ice velocities on glaciers is a crucial step towar ds (i) understanding and modeling of the dynamic processes involved, (ii) estimating the system responses to external forcing, such as changes in climatic conditions, and (iii) assessing related impacts. With the Arctic being a region presently particular ly affected by atmospheric warming, monitoring of Arctic glaciers is of critical importance for understanding regional climate change and its impacts, such as sea level contribution. In particular on Svalbard with its high percentage of surge-type glacier s, knowledge of spatio-temporal glacier flow is crucial for interpreting observed glacier volume changes correctly. Radar satellite sensors with their all-weather and night-time capability are the only means to continuously monitor glacier flow in polar e nvironments. While the motion of slow-moving glaciers can in theory be derived through synthetic aperture radar (SAR) interferometry, fast-flowing glaciers have to be measured using correlation between repeat data (offset tracking). The present project RA STAR will explore the dependency of trackable SAR backscatter features from different seasons and weather conditions, different radar bands (L, C, X), and different SAR polarizations and spatial resolutions, focusing in particular on new and upcoming sens ors like TerraSAR-X, RADARSAT-2 or ESA Sentinel-1. The findings will be compiled to an integral methodology and observation strategy. From that, the first multitemporal glacier velocity map over entire Svalbard will be produced and the results be analysed towards Svalbard calving flux, surge activities, and velocity trends over time. The map will further be a key to understand current and forthcoming (ESA CRYOSAT-2) elevation change data over Svalbard.