At Nunatakassap Sermia, we might document surprising calving from overhead satellites and even measure acceleration but to get at an explanation (and predict future behavior) we need improved Melville Bay bathymetry as well as better bedrock mapping under the glacier (which gives ice thickness).
As mentioned, E Rignot made some very significant findings last season about the actual bathymetry of a few West Greenland fjords and calving fronts of marine terminating glaciers using sonar on a volunteered small vessel. OMG (Oceans Melting Greenland) will follow up on that over the next four years but now it appears part of that agenda has been achieved by Boghosian 2015 using existing data.
The primary interest in better bathymetry is not improved nautical charts but rather the accessibility of deeper (here warmer) ocean waters to the calving front. Because of poor vertical mixing, deeper waters of greater density can't get pass obstructing sills. These barriers might be old underwater terminal moraines at the far end of previous glacier over-deepenings or just less erosive bedrock.
On the other hand, a larger glacier during the last ice age might have gouged unobstructed channels that provide a direct connection of the calving front to waters circulating in Baffin Bay.
That's what it's all about: distinguishing channels from sills and potholes. It's far safer and faster to get the necessary data from an airplane than a small boat dodging icebergs calving from an active front. In fact, the data has already been gathered over the years by Operation IceBridge. However it's not as precise as a half-million dollar sonar hanging off the side of an on-site ship.
Realistically, western Greenland has far too many glaciers with short melange-free seasonal windows to comprehensibly map bathymetry with sonar. However the ones that have been mapped just north of Jakobshavn show that previous depth maps were wholly unsatisfactory for modeling future calving rates, indeed downright misleading to the extent the current paradigm -- control of glacier advance by frontal turbulence at basal meltwater exits -- is valid.
Given several fjords with both types of data, very substantial improvements in AirGrav data interpretation might be possible over all of coastal Greenland using guidance from the much higher resolution sonar instances, in the sense of learning where and why AirGrav goes astray and how best to interpolate from its initial limited resolution.
Data is only collected along a sparse grid of intersecting flight lines; 'ordinary kriging' is used to fill in the blanks. This method is isotropic: all compass directions are on the same footing. We've seen before at Petermann that's untrue: basal ice upheavals follow contemporary surface flowlines and multiple paleo bedrock channels drain ancient gradients. Indeed all Greenland's ice flows away from the central ridge with implications for physics-based interpolation.
Here we have a natural symmetry-breaking vector supplied by the direction of glacier travel (and thus maximal erosion). Thinking now of classical U-shaped glacial valleys, this polarization vector also provides a twofold axis of transect symmetry to the channel going to the sea.
Incorporating the dominant physical process into interpolation cannot help but improve outcomes. Here we have something simple and specific sitting within the larger topic of 'anisotropic kriging' which could involve tensorial (simultaneous multi-linear directional) influences.
While applying this to Nunatakassap (aka Alison Glacier) is certainly within the collective technical competency of this forum, it's better to first take a detour to Store Glacier and its neighbors as these were studied by both Rignot and Boghosian groups. This can provides insights into improvements in AirGrav inversion and a validation area for resolution interpolation proposals.
Indeed ImageJ --> Plugins --> Integral Image Filters --> Normalize Local Contrast provides a low-budget variable anisotropic interpolation for geolocated grayscale data, which here is conveniently aligned east-west to begin with.
We've long needed a separate forum for the Uummannaq region. The Store glacier there is being intensively studied in the SAFIRE project -- four instrumented holes being drilled to bedrock to get at mechanical and hydrological conditions at the base. Store Gletscher (big glacier in Danish) drains 35,000 km
2 at 20 m per day by the calving front -- it's similar to but more favorable experimentally than Jakobshavn.
http://www.spri.cam.ac.uk/research/projects/safire/
