Twentieth century constraints on sea level change and earthquake deformation at Macquarie Island

Abstract

SUMMARY Through the combination of rare historical sea level observations collected during Sir Douglas Mawson’s 1911–1914 Australasian Antarctic Expedition (AAE), together with modern sea level data, space geodetic estimates of crustal displacement and modelling of coseismic and post-seismic earthquake deformation, we present a contemporary analysis to constrain sea level and land level change over the twentieth century at Macquarie Island (54 ◦ 30 � S, 158 ◦ 57 � E). We combine 9 months of 1912–1913 sea level data with intermediate observations in 1969–1971, 1982 and 1998–2007 to estimate sea level rise relative to the land at +4.8 ± 0.6 mm yr −1 . Combined with estimates of global mean sea level rise, this value supports the geologically surprising notion of land subsidence, conflicting with longer term geological evidence that suggests uplift at ∼0.8 mm yr −1 over the last 400–300 Kyr. We investigate the current tectonic evolution of the Island through analysis of Global Positioning System (GPS) solutions that utilize data over the period 2000–2009. Importantly, this provides an opportunity to refine the source parameters of the M w ∼8.0 great earthquake of 2004 December 23 using estimates of coseismic displacements at regional GPS sites. We use the estimated earthquake source and GPS observations of four years of post-seismic deformation at Macquarie Island to infer the rheology of the oceanic upper mantle. We find that an asthenosphere bounded by stronger material above and below is required to produce the observed post-seismic deformation, particularly in the vertical component. Assuming a Maxwell rheology, the best fit is given by an asthenospheric viscosity of 3×10 19 Pa s. The inferred rheology determined from the 2004 earthquake is used to model long period post-seismic deformation from M w ∼8.0 earthquakes of 1989 and 1924. The 1924 earthquake is the closest of the three great earthquakes to Macquarie Island, and our modelling suggests that the majority of the vertical deformation at the tide gauge over the subsequent 80 years is related to ongoing viscoelastic relaxation from this thrust earthquake that ruptured south from an epicentre south of Macquarie Island. Assimilated time-series of land level change from the earthquake modelling (suggesting ongoing subsidence) constrained by the GPS estimate of vertical velocity of −2.46 ± 0.64 mm yr −1 combine with the relative sea level time-series to yield an estimate of absolute sea level change of +2.0 ± 0.8 mm yr −1 over the twentieth century. We conclude this is consistent with the upper bound of the global average rate of absolute sea level rise over the same period. This represents one of few estimates of observed sea level change in the Southern Ocean, re-emphasizing the importance of historical data and continued geodetic and oceanographic observation in remote areas.