Iceberg dynamics in a proglacial lake in Austria quantified by time-lapse photography

The recession of glaciers reveals a dynamic landscape exposed to high rates of hydrological and geomorphological modifications. Such deglaciation processes caused the formation of a 0.3 km² large proglacial lake (named Pasterzensee) near the terminus of Pasterze Glacier, Austria, during the last two decades. The evolution of the proglacial lake was accompanied by several buoyant calving events. The process of buoyant calving formed numerous floating dead ice bodies referred to as icebergs which covered a maximum of 7.3 % of the entire proglacial lake basin in November 2018.

Despite the existence of icebergs at some proglacial lakes in the European Alps, little is known about the evolution and life span of icebergs in proglacial lakes in the European Alps. The aim of this study was to reduce this research gap by (a) quantifying the evolution of such alpine icebergs during two different time scales and by (b) analysing the relationship between iceberg evolution and motion at the lake with meteorological conditions. At a long-term scale, one single iceberg was monitored during the period 01.09.2017-30.09.2019. At a short-term scale, all icebergs were studied during one single day (16.06.2019).

The most important data source for this study were time-lapse optical imagery from an automatic camera overlooking the entire proglacial lake (GROHAG). The used camera is a Roundshot Livecam Generation 2 (Seitz, Switzerland). Photographic imagery is captured every five minutes (during daylight) from a location 310 m above lake level and 450 m northeast of the lake margin. For the long-term analysis, a total number of 386 pictures of the lake were processed. For the short-term analysis, 97 pictures were analysed to reveal the dynamics of 84 icebergs during one single day. The oblique time-lapse images were transformed into orthorectified photos using a rectification algorithm which considers the camera properties and the lake surface geometry. Iceberg size and centroid coordinates were mapped in all generated orthophotos. In addition, meteorological data (ZAMG Vienna) was provided by a nearby automatic weather station, located at the glacier tongue of Pasterze Glacier some 1.1 km northwest of the lake margin.

Results indicate that the monitoring of one iceberg over a period of 25 months revealed highest melting rates from June to August, low melting rates from September to November and no measurable melting when the lake surface is frozen. Horizontal iceberg displacement is rising with decreasing iceberg size throughout the study period. The analysed iceberg formed during the detachment of a debris covered ice peninsula with an initial size of 7250 m² and was last identifiable at a size of 240 m². Monitoring lake-wide iceberg movement for one day shows that wind is the main influence on horizontal iceberg displacement. The existence of a strong valley wind, caused by a diurnal warming cycle, is observed. This wind system decouples the iceberg movement from the constant katabatic glacier wind, recorded by the weather station. Frequent jumps in movement rates, which are not explained by wind data, suggest that iceberg grounding is a common process influencing subaquatic lake morphology.