Mapping Greenland by ERS-1/2 InSAR for ice mass balance and dynamics studies

We are conducting an ESA ERS-AO3 study aiming at mapping ice elevations and flow rates for the Greenland ice sheet and glaciers. Such data are required by the Global Change community for monitoring changes of mass balance of ice sheets and glaciers, by ice modelers for verifying ice dynamics models, and for estimating changes in fresh-water input from land-ice masses to the sea, which have impact on global sea level and oceanic global circulation. There are many issues which needs to be considered when undertaking such a large scale mapping job. The data selection itself is a nontrivial matter. We have chosen an approach which utilizes the information in ESA's baseline files combined with automated selection filters and a graphical interface to support the final data selection. Also, to map Greenland's 2.3 mio. square kilometers, it is necessary to process long strips and to develop automated processing algorithms for data adjustment, interpretation, and mosaicking. In many areas we will never get optimal data coverage. Thus to complete a Greenland map, it is necessary to develop methods and models which provides solutions in areas where only one or two interferograms are available. At present the project is in an introductory phase, in which we focus on the development of the processing methods to be applied as well as on local phenomenological investigations. Introduction Ice flow velocities and elevations are key parameters in studying dynamics and mass balance of glaciers and ice sheets. The overall mass balance of the Greenland and Antarctic sheets is an important variable for climate and related sea-level change studies. The overall ice sheet mass balance can be determined by various approaches: 1) by measuring each term (snow accumulation, snow/ice melt, and iceberg calving) of the mass-balance budget; 2) by measuring the massor volume-change by monitoring ice sheet surface elevation change; or, 3) by combining surface velocity measurements with ice thickness information. For a short review of the various methods, see Reeh (in press). Ice flow measurement has previously been based on two techniques. Point measurements are provided by surveying poles placed in the ice. This technique is expensive and offers insufficient spatial sampling of flow fields. Using visual remote sensing techniques, e.g. Landsat, it has been possible to track recognisable features in the ice thereby measuring long term ice flow. This technique is hampered by clouds and depends on the existence of lasting characteristics recognizable in the images used. The accuracy of the technique is also limited by the spatial resolution of the imagery. The determination of the elevations of glaciers and ice sheets have been a major problem until recently, and height determinations by surface survey or airborne profiling have been very expensive and slow. Conventional mapping methods based on stereo photogrammetry are unsuitable for ice mapping, since ice sheets in polar regions seldom show sufficient contrast for stereo matching. In Greenland, for example, the existing maps of the margins of the ice sheet show height contours that can be hundreds of meters in error (Ekholm, 1996). With the advent of space-based satellite altimetry, notably by the ERS-1 satellite, the entire interior, flat parts of the Greenland ice sheet have been mapped accurately for the first time. Satellite altimetry does not work well in the marginal ice sheet zones, where large gradients, nunataks, outlet glaciers and local ice caps exist (Bamber et al. 1998). Thus InSAR is a unique tool for providing ice velocity data and digital elevation models (DEMs) on glaciers and ice sheets. Our project consist of 3 major phases: 1) technology/tools development, 2) operational processing, and 3) verification. The total duration of the project is estimated to be 4 years.