Multi-sensor detection of glacial lake outburst floods in Greenland from space

Abstract:

GLOFs cause substantial erosion, transport and delivery of sediment along the river system from the glaciated parts of the hydrologic catchment to the sea, and have been found to control the riverine export dynamics of some pollutants like mercury in NE Greenland. GLOFs also pose a risk to human presence and infrastracture.

Ice-dammed lakes at the margin of the ice sheet and of local glaciers and ice caps are common features of Greenland’s landscape. The occasional or periodic emptying of some of these lakes have been described as early as the 18^thcentury. Thinning glaciers in a warming climate are already changing the behaviour of some of these lakes. However, little is known of the frequency and seasonality of glacier lake outburst floods (GLOF) outside of the relatively more densely populated parts of West and South Greenland.

This contribution demonstrates automatic multi-sensor detection of ice-dammed lake emptying events from space for three test regions in West, South and Northeast Greenland, using visible imagery from Landsat, ASTER, PROBA-V and MODIS. The current detection algorithm relies on prior knowledge of lakes location and approximate shape from a topographic map at the scale of 1:250.000, and it is meant as a prototype for a future operational product. For the well documented case of the glacier-dammed lake of A.P. Olsen Ice Cap (NE Greenland), where GLOF’s observations at Zackenberg Research Station started in 1996, the remote sensing and in situ records are compared, showing good agreement. ICESat altimetry, MODIS and AVHRR thermal imagery, and the ENVISAR ASAR signature of two detected GLOFs that took place late autumn and winter are also discussed to demonstrate the potential for successful retrievals during the polar night.

The upcoming Sentinel-3 missions will alleviate what is currently the major drawback of implementing this prototype into an operational service, namely the limited availability of high resolution imagery. This is of special importance for applications characterizing and monitoring GLOF hazard in areas of Greenland of interest for mining, hydropower development, tourism and and other human activities.