North Greenland's Ice Shelves and Ocean Warming

Abstract:

Rapid disintegration of ice shelves (the floating extensions of marine-terminating glaciers) can lead to increasing ice discharge, thinning upstream ice sheets, rising sea level. Pine Island Glacier, Antarctica, and Jacobshavn Isbrae, Greenland, provide prominent examples of these processes which evolve at decadal time scales. We here focus on three glacier systems north of 78 N in Greenland, each of which discharges more than 10 Gt per year of ice and had an extensive ice shelf a decade ago; Petermann Gletscher (PG), Niogshalvfjerdsfjorden (79N), and Zachariae Isstrom (ZI).

We summarize and discuss direct observations of ocean and glacier properties for these systems as they have evolved in the northwest (PG) and northeast (79N and ZI) of Greenland over the last two decades. We use a combination of modern and historical snapshots of ocean temperature and salinity (PG, 79N, ZI), moored observations in Nares Strait (PG), and snapshots of temperature and velocity fields on the broad continental shelf off northeast Greenland (79N, ZI) collected between 1993 and 2014.

Ocean warming adjacent to PG has been small relative to the ocean warming adjacent to 79N and ZI; however, ZI lost its entire ice shelf during the last decade while 79N, less than 70 km to the north of ZI, remained stable. In contrast, PG has thinned by about 10 m/y just prior to shedding two ice islands representing almost half its ice shelf area or a fifth by volume. At PG advective ice flux divergence explains about half of the dominantly basal melting while response to non-steady external forcing explains the other half. The observations at PG,79N, and ZI suggest that remotely sensed ambient surface ocean temperatures are poor proxies to explain ice shelf thinning and retreat. We posit that local dynamics of the subsurface ocean heat flux matters most. Ocean heat must first be delivered over the sill into the fjord and then within the ice shelf cavity to the base of the shelf near the grounding line. Models of glacier-ocean interaction must represent both bottom topography and closely related ocean dynamics and mixing at their dynamically relevant scales within a density stratified water column. Projects for such integrated ocean-glacier observations are in the planning stages for 79N and PG.