We use cookies to ensure that we give you the best experience on our website. You can change your cookie settings at any time. Otherwise, we'll assume you're OK to continue.

Durham University

Research & business

View Research Directory

Reconstructing Ice Sheet History Using Relative Sea-Level (RSL) Observations and Modelling

A research project of the Department of Geography.


Relative sea-level (RSL) data provide the most powerful constraints on Greenland Ice Sheet (GIS) history available. In recent research we have developed new RSL records using the “isolation basin” approach, significantly improving the precision of previous reconstructions and providing powerful new constraints on ice sheet models. Our work has focussed on the marine embayment of Disko Bugt and Jakobshavns Isbrae, one of the most important ice streams in Greenland. Using this approach, we have established a new chronology for ice stream retreat and developed a complementary record using offshore sedimentary sequences that tracks variations in Holocene ice stream behaviour.

We have applied the isolation basin approach elsewhere in Greenland, including in the Sisimiut area (outer Søndre Strømfjord) and also by developing the first RSL record from the southeast sector of the GIS at Ammassalik.


The “isolation basin” approach involves collecting sediment cores from lakes that were formerly below sea level but, because of glacio-isostatic rebound, have been raised up and isolated from the sea. The sedimentary sequence in these basins typically comprises a lower marine deposit which passes upwards through a brackish water phase that records the process of basin isolation and then a fully freshwater deposit that formed once isolation was complete. By analysing the microfossils (mostly diatoms) within the sediment sequences, we can constrain the details of the isolation process and determine when exactly the basins became freshwater and isolated from the sea. We then date this change using AMS radiocarbon dating. We can track how RSL has changed through time by collecting such data from a suite of basins at different elevations above present sea level.


We have developed local RSL records for seven field sites in the Disko Bugt embayment in West Greenland, enabling a comprehensive reconstruction of spatial and temporal variations in RSL and postglacial rebound across the region. The oldest parts of the records provide a basis for reconstructing a model for ice recession since the last glacial maximum. This shows that the bay became ice free much later than previously thought, with ice extending onto the continental shelf well up until the start of the Holocene, before retreating rapidly to the shallow eastern waters of the bay by c. 9-10 k cal. yr BP. When on the shelf, the ice sheet was grounded and able to withstand significant sea level rise, including meltwater pulse 1a, without undergoing catastrophic collapse.

One focus for our work has been to establish the Holocene history of the Jakobshavn ice stream. To this end, we have used RSL data to track ice stream retreat across Disko Bugt and also along the length of Jakobshavn Icefjord, developing a RSL record from mid-way down the fjord. A second focus has been on re-evaluating the age of the “Fjord Stade” moraines, a suite of moraines that occurs in many fjord heads in west Greenland and which have previously been dated to c. 8.2 ka cal. yr BP. We find that in Disko Bugt the ages of these moraines are diachronous and that they probably formed over several thousand years during the early Holocene in response to ice sheet interaction with local topographic controls. They do not record a single ice margin response to a climatic event.

A third area of effort has been to determine how far, and at what time, RSL fell below present riding again in the late Holocene. Several workers have argued that this turnaround in RSL reflects renewed crustal loading and glacio-isostatic depression caused by a readvance of the ice sheet in the “neoglacial”, from 5-4 k cal. yr BP and culminating in the Little Ice Age maximum. By coring drowned isolation basins, we are able to define spatial differences in the timing and magnitude of neoglacial RSL rise across Disko Bugt, identifying a greater rise (c. 5 m) in the east of the bay close to the present ice sheet margin compared with sites further west (c. 1-2 m). Such a pattern is commensurate with the neoglacial reloading hypothesis.

In southeast Greenland, a combined RSL and cosmogenic exposure dating programme provides the first constraints on post glacial rebound for this sector of the ice sheet. We show that the island of Ammassalik became ice free c. 11 k cal. yr BP and that RSL fell to reach close to present by c. 6-7 k cal. yr BP. Using a newly developed ice sheet model, we demonstrate that the ice sheet must have extended well onto the continental shelf at the last glacial maximum, before retreating quickly and relatively recently to reach its present position. The ice sheet thickness was well in excess of 750 m, requiring significant revision to existing ice sheet models.

Finally, in the Sisimiut area of central west Greenland, isolation basins provide new observations on the trend in mid and late Holocene RSL changes. Using drowned isolation basins we show that RSL here has rise by c. 3-4 m since 1-2 ka cal. yr BP, broadly similar to the patterns observed on the outer coast of Disko Bugt. However, the timing of the switch from RSL fall to rise is significantly later than the onset of the neoglacial climate deterioration, whilst elsewhere in south Greenland the equivalent change occurs well before the neoglacial deteriorationbegan. This implies that for these areas at least, non-Greenland processes, notably the collapse of the forebulge associated with the Laurentide Ice Sheet, have played important roles in controlling RSL trends.


From the Department of Geography