The depositional architecture of valley-fill systems is controlled by the interaction of boundary conditions: (1) ice sheet and drainage area variations which influence initial topography, river discharge and sediment supply, and (2) sea-level fluctuation and isostatic rebound which control the available accommodation space. This complex interplay between supply and accommodation conditions for glacio-fluvial valley-fills leads to a typical stratigraphy that is the focus of study for this project. Glacio-fluvial systems respond rapidly to local and global environmental changes during deglaciation. Eustatic sea level changes and isostatic uplift interact to form a complicated relative sea-level history. This study aims to relate the spatial distribution of sedimentary units (laterally and vertically) to the complex depositional history of the area, in order to better understand the mechanisms that force the glacio-fluvial sedimentary system. Sedimentary, morphological and geophysical data is used to constrain the size and spatial distribution of the main depositional units.

The study area is located on the ice-free margin of Western Greenland near Kangerlussuaq (Søndre Strømfjord). The ice-carved Sandflugtdalen valley (Figure 1) is characterized by several sediment outwash plains divided by bedrock features. The valley length measures about 30 kilometers between the ice front and the present-day fjordhead. At its widest, the valley is about 4 kilometers wide. Evidence for the upper marine limit (fine-grained marine sediments; figure 1) is found at 55 m above present sea-level. River discharge is dominated by glacial meltwater from the hinterland ice sheet.

In the summer of 2007 a field campaign was held which included geomorphological mapping, sedimentary logging and ground penetrating radar measurements. Sedimentary information was gathered based on 4 vertical log sections and 7 shallow soil pits (e.g. Figure 2). A total length of over 2000 meters of GPR profile was gathered using shielded 225 MHz and 450 MHz antennas. Survey depth was limited to the uppermost 4 meters due to the presence of an active layer-permafrost contact at 2-3.5 m depth.

Glacio-fluvial braid plain deposits and glacio-marine mud deposits are the main sedimentary units that make up the valley-fill of Sandur 1 and 2 (Figure 2). The fluvial deposits exhibit discrete changes in fluvial styles along the longitudinal valley, imposed by bedrock morphology and related transport capacity. The narrow stretches of the valley, which also include bedrock steps, increase the river transport capacity which results in the formation of river terraces. Several incised terrace levels are present testifying of sediment fluctuations over the longer term evolution. At the braid plain furthest from the ice margin (sandur 2) terrace incision does not occur. Braid plain deposits located near the fjordhead are finer and better sorted than deposits near the ice margin. The marine mud deposits contain evidence of the sea level history of the system.

Using ground penetrating radar the sedimentary architecture was imaged at both outwash plains, along the length as well as perpendicular to the river system. Differences in depositional elements between sandur 1 and 2 have been described in terms of sedimentary and radar facies. The braid plain of sandur 1 (closest to the ice margin) is characterized by thick packages of braided stream deposits. GPR records reveal braided stream architectural elements as channel fills and bars. Several incised terrace levels are present testifying of sediment fluctuations over the longer term evolution. At sandur 2, radar data suggests that the sandy braid delta deposits erosively overlie fine-grained marine mud deposits at a depth of approximately 4 m.

Based on our observations, we reason that the coarse sands of sandur 2 were deposited under conditions of slow relative sea-level rise. This is different from the common hypothesis that recently deglaciated regions should experience a relative sea-level fall due to isostatic uplift. This coincides well with the observation of Dietrich et al. (2005) who report inverted isostatic movement (subsidence) related to ice sheet re-expansion in Western Greenland following the Holocene Thermal Maximum.