Crustal deformation across the Southern Patagonian Icefield observed by GNSS

Highlights

• Crustal deformation at the Southern Patagonian Icefield observed at 43 GNSS sites.

• Glacial-isostatic adjustment generates crustal uplift of up to 4 cm per year.

• Plate collision and the Patagonian slab window contribute also to horizontal strain.

Abstract

Geodetic GNSS observations at 43 sites well distributed over the Southern Patagonian Icefield region yield site velocities with a mean accuracy of 1 mm/a and 6 mm/a for the horizontal and vertical components, respectively. These velocities are analyzed to reveal the magnitudes and patterns of vertical and horizontal present-day crustal deformation as well as their primary driving processes. The observed vertical velocities confirm a rapid uplift, with rates peaking at 41 mm/a, causally related to glacial-isostatic adjustment (GIA). They yield now an unambiguous preference between two competing GIA models. Remaining discrepancies between the preferred model and our observations point toward an effective upper mantle viscosity even lower than $1.6\cdot {10}^{18}\text{ Pa}\text{s}$ and effects of lateral rheological heterogeneities. An analysis of the horizontal strain and strain-rate fields reveals some complex superposition, with compression dominating in the west and extension in the east. This deformation field suggests significant contributions from three processes: GIA, a western interseismic tectonic deformation field related to plate subduction, and an extensional strain-rate field related to active Patagonian slab window tectonics.

Introduction

The Southern Patagonian Icefield (SPI) has been identified as a locus of exceptionally rapid crustal uplift (Dietrich et al., 2010, Lange et al., 2014). Models show that observed vertical deformation in this region can be explained by glacial-isostatic adjustment (GIA) (Ivins and James, 1999, Ivins and James, 2004; Klemann et al., 2007). Vertical site velocities observed in Patagonia exceed those reported for other regions affected by glacier retreat such as Alaska (Larsen et al., 2005), Greenland (Khan et al., 2007, Dietrich et al., 2005) and West Antarctica (Groh et al., 2012).

The large-amplitude GIA response in Patagonia results from the coincidence of a large, rapidly changing temperate ice mass and a unique tectonic setting. The Patagonian icefields, representing the largest extra-polar ice mass in the southern hemisphere, have experienced substantial fluctuations in glacier extent and mass throughout the Pleistocene (e.g. Strelin et al., 2011, Strelin et al., 1999; Mercer, 1976) and Holocene (e.g. Strelin et al., 2014, Aniya, 2013, Masiokas et al., 2009) until present (e.g. Floricioiu et al., 2012, Willis et al., 2012, Glasser et al., 2011, Casassa et al., 2002). GIA models suggest, however, that the uplift observed today in Patagonia is caused by ice-mass changes since the Little Ice Age (LIA) with its maximum extent between AD 1630 and 1870 (Ivins and James, 2004, Lange et al., 2014). The uplift predicted from the elastic crustal response to ongoing ice loss shows a more localized pattern than the complete gravitational visco-elastic rebound, but may reach about half the total uplift rate close to fast-retreating glaciers like Upsala (Lange et al., 2014, Fig. 1c).

The short-lived memory of the solid earth for load changes in southern Patagonia, compared to other regions affected by GIA, is due to the peculiar regional rheology characterized by a thin lithosphere and very low viscosities in the asthenosphere and upper mantle (Lange et al., 2014). These rheological conditions, in turn, are imposed by the geologically young thermomechanics of the region. At the Chile Triple Junction, some 200 km NNW from the SPI, material of the relatively hot Chile Ridge is subducted beneath the South American plate (Fig. 1a). The subduction of this active spreading-ridge system between Nazca and Antarctic plates is accompanied by the opening of the Patagonian slab window. The existence and extent of this slab window has been established by geochemical analysis of Patagonian lavas and basalts (Gorring and Kay, 2001, Boutonnet et al., 2010), seismic imaging (Russo et al., 2010a, Gallego et al., 2010), shear wave splitting measurements (Murdie and Russo, 1999, Russo et al., 2010b) and plate kinematic modeling (Breitsprecher and Thorkelson, 2009). The slab window environment beneath the southernmost South American continent is characterized by upwelling of hot mantle material and enhanced mantle flow through the window with a strong ridge-parallel component (Russo et al., 2010b).

From structural-geological point of view, most of our sites pertain to the Southern Patagonian Andes fold-thrust belt. Most authors agree that fold deformation and activity of minor faults in our study area are of late-Miocene age or older (e.g. Giacosa et al., 2012; Fosdick et al., 2011, Ghiglione et al., 2009, Kraemer, 1998). Coutand et al. (1999) suspect that basement faults along the cordillera in the Lago Viedma area “appear to be active in strike-slip reverse mode”. Currently, evidence for active faulting in the SPI region is difficult to find. International seismological catalogues (ISC, 2015, USGS, 2015, Villaseñor et al., 1997, Sabbione et al., 2007) report also only a small number and low magnitude of seismic events for this area. There are active volcanoes along the crest of the Southern Patagonian Andes. However, the deformation associated with volcanic activity is usually restricted to the vicinity of the volcano edifice, and none of our sites are located close enough to be affected by volcanic deformation. We assume, therefore, that our observed site velocities essentially represent the long-wavelength deformation field.

Both plate collision and slab window opening are expected in addition to GIA to produce large-scale surface deformations affecting the SPI region. Tectonic deformations are reflected primarily in the horizontal components (e.g., Elliott et al., 2010). However, previous works on crustal deformation in southern Patagonia focused on the vertical component (Dietrich et al., 2010, Lange et al., 2014). Global compilations of the horizontal deformation field, in turn, expose Patagonia as a region lacking geodetic observations (Kreemer et al., 2014).

This paper presents results of GNSS observations in a regional network of 43 sites. Compared to previous work, the network extension and geometry are improved and allow, for the first time, a detailed analysis of the horizontal velocity components. The observed vertical and horizontal site velocities are interpreted with regard to the magnitude, patterns and driving processes of present-day crustal deformation across the SPI.