Subduction is a fundamental component in plate tectonics. Here we focus on a prime example, the Sunda subduction zone that has been the location of several destructive mega earthquakes with magnitude greater than 8 in the past centuries (e.g Henstock et al. 2006). The subduction zone is formed by the Australian plate moving into the Eurasian plate and ranges from the Himalayas in the north, to Timor in the south (figure below), but several parts of the plate move independently and the fault is highly segmented (McCaffrey 2009). Moreover there is a lot of variation in type of motion and earthquake risk along strike (Newcomb and McCann 1987) The subduction zone shows the whole range of subduction events, including slow slip events (Tsang et al. 2015) and subduction-related volcanism.


South-east Asia has been dominated by subduction with rates between 5 and 10 cm/year (Hall and Spakman 2015). Information on the Sunda subduction zone is available from different measurements types. GPS measurements reveal change in plate coupling along the subduction zone (Prawirodirdjo et al. 1997). Horizontal GPS motion together with coral measurements of uplift rate demonstrated that the locked zone most likely extends into the mantle (Simoes et al. 2004). Detailed bathymetry reveals several smaller faults that link to the type of faulting below the surface (Henstock et al. 2006)

Information on the geometry of the plates is coming from earthquake hypocenters. At depths above 100 km these represent coupling between the subducting and overriding plate, and below the earthquakes arise from interpolate seismicity (e.g. Richards et al. 2007). Below hypocentre locations, P-wave tomography offers a view on the shape of the slab.  The region north of the Andaman islands, south of Myanmar, shows no seismicity, tomography anomaly (Richards 2007) or volcanoes. Richards et al. (2007) suggest an absence of the plate caused by a tear between the Burmese and Sumatran sections. This is disputed by Pesicek et al. (2010) who present a sharper P-wave tomography picture.


After subtracting topography, residual anomalies can represent deepening of the plate, and density variations due to erosion in the accretionary wedge (Kopp et al. 2001; Henstock et al. 2015). At the Andaman islands, the subduction processes are more complex (Subrahmanyam et al. 2008), making the gravity data more difficult to interpret. Because of good coverage with seismic data, coral data and GPS uplift rates we propose the area around Sumatra to be a test area for applying constraints for gravity data before tackling the more uncertain area north of Andaman islands. Magnetic anomaly maps have been used to identify smaller scale faults (Zubaidah et al. 2014).