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.

Department of Geography

Departmental Research Projects

Publication details

Woodroffe, S. A., Long, A. J., Milne, G. A., Bryant, C. L. & Thomas, A. New constraints on late Holocene eustatic sea-level changes from Mahé, Seychelles. Quaternary Science Reviews. 2015;115:1-16.

Author(s) from Durham


This study provides new estimates of globally integrated ice sheet melt during the late Holocene (since 4 ka BP) from Seychelles in the western Indian Ocean, a tectonically stable, far field location where the necessary Glacial-Isostatic Adjustment (GIA) correction is small and is relatively insensitive to predictions using different Earth viscosity profiles. We compare sea level data from Seychelles to estimates of eustasy from two GIA models, ICE-5G and EUST3, which represent end-members in the quantity of global melt during the late Holocene. We use data from a range of coastal environments including fringing reef, present day beaches, fossil plateau and mangrove deposits on the largest island of the Seychelles archipelago, Mahé to reconstruct relative sea-level changes. Our data suggest that extensive coastal deposits of carbonate-rich sands that fringe the west coast formed in the last 2 ka and the horizontal nature of their surface topography suggests RSL stability during this period. Mangrove sediments preserved behind these deposits and in river mouths date to c. 2 ka and indicate that RSL was between −2 m and present during this interval. Correcting the reconstructed sea level data using a suite of optimal GIA models based on the two ice models mentioned above and a large number (c. 350) of Earth viscosity models gives a result that is consistent with the sedimentological constraints. When uncertainties in both model results and data are considered, it is possible to rule out eustatic sea levels below c. 2 m and more than a few decimetres above present during the past two millennia. This uncertainty is dominated by error in the reconstructions rather than the model predictions. We note, however, that our estimates of eustasy are more compatible with the EUST3 model compared to the ICE-5G model during the late Holocene (2–1 ka BP). Our evidence from Seychelles shows that the timing of when eustatic sea level first rose close to present is between the predictions of the two end-member GIA models presented here (4 ka BP for ICE-5G and 1 ka BP for EUST3). Using all lines of evidence currently available from Mahé we suggest that the eustatic contribution during the last 2 ka has been less than 2 m . This conclusion is drawn from a tectonically stable, far-field region that is relatively insensitive to earth and ice model uncertainties, and implies that global eustasy has been relatively insensitive to climate fluctuations over the pre-industrial part of the last 2 ka.

Department of Geography