Analysis of the Illapel Mw = 8.3 Thrust Earthquake Rupture Zone Using GOCE-Derived Gradients ORLANDO A ´ LVAREZ, 1,2 AGUSTINA PESCE, 1,2 MARIO GIMENEZ, 1,2 ANDRES FOLGUERA, 3 SANTIAGO SOLER, 4 and WENJIN CHEN 5,6 Abstract—Satellite gravimetry has proven to be a useful tool to identify mass anomalies along a subduction interface, interpreted as heterogeneities related to the rupture process during megathrust earthquakes. In the last years, different works, reinforced with data derived from satellite gravity missions as GRACE and now GOCE, have analyzed not only the static component of the Earth gravity field, but also its temporal variations and relation to the seismic cycle. In particular, during the last decade, the Chilean margin has been affected by three megathrust earthquakes (with Mw[ 8): Maule 2010 Mw = 8.8, Pisagua 2014 Mw = 8.2 and recently the Mw = 8.3 Illapel event. Then, the recently completed GOCE mission (November 2009 to November 2013) offered a unique opportunity to study the Maule February 2010 and Pisagua April 2014 events by means of gravity gradients, directly measured at satellite height altitudes, which allowed mapping density heterogeneities with greater detail than the gravity anomaly which has been used in most studies up to now. In the present work, we use the last GOCE model (GO_CONS_GCF_2_DIR_R5), the one of higher spatial resolution (N = 300, k/2 & 66 km) derived from satellite-only data. The methodology used is the same as that to study the previous events, with the addition that now we derived a relation between the asso- ciated depths of a causative mass with a determined degree of the spherical harmonic expansion. This allowed to ‘‘decompose’’ the gravimetric signal, by cutting off the degree/order of the harmonic expansion, as depth increases. From this analysis, we found that prominent oceanic features such as the Challenger fracture zone and the Juan Fernandez ridge played a key role in latitudinal seismic segmentation for the Illapel earthquake rupture zone, acting as bar- riers/attenuators to the seismic energy release. We compared the slip model from Tilmann et al. (Geophysical Research Letters 43: 574–583. doi:10.1002/2015GL066963, 2016) for the Illapel earth- quake with vertical gravity gradient with and without sediment correction, and at different degree/order of the harmonic expansion. From this analysis, we inferred that prominent oceanic features over the subducting Nazca plate play a key role in seismic segmentation not only at heavily sedimented trenches, but also at sediment-starved segments. Key words: Vertical gravity gradient, Megathrust earthquakes, South central Andes, Spherical harmonics, GOCE. 1. Introduction Rupture areas related to large subduction earth- quakes have been studied by means of gravity since the pioneer works of Song and Simons (2003) and Wells et al. (2003), among others. Then the different vari- ables governing this relationship have been analyzed by different authors (e.g., Llenos and Mc Guire 2007; Sobiesak et al. 2007; Tassara 2010; Maksymowicz et al. 2015; among others). Fuchs et al. (2013) observed coseismic gravity changes from the Japan Tohoku-Oki 2011 earthquake with GOCE (Gravity Field and Steady State Ocean Circulation Explorer) gravity gradiometry, concluding that these variations left a statistically significant signal in the GOCE-measured gravity gradients. This work indicated that it was possible to detect coseismic gravity changes by spaceborne gradiometry as GOCE. More recently, slip distribution for great megathrust earthquakes along the Peru–Chile margin has been correlated to GOCE- 1 Instituto Geofı ´sico Sismolo ´gico Ing. F.S. Volponi, Universidad Nacional de San Juan, Ruta 12, Km 17, Jardı ´n de los Poetas, Marquesado, Rivadavia, San Juan, Argentina. E-mail: or- lando_a_p@yahoo.com.ar; orlando.alvarez@conicet.gov.ar; pesce.agustina@gmail.com; gimmario@gmail.com 2 Consejo Nacional de Investigaciones Cientı ´ficas y Te ´cnicas (CONICET), Buenos Aires, Argentina. 3 Departamento de Cs, Geolo ´gicas, INDEAN, Instituto de Estudios Andinos ‘‘Don Pablo Groeber’’, FCEN, Universidad de Buenos Aires, Buenos Aires, Argentina. E-mail: andresfolguera2@yahoo.com.ar 4 Facultad de Ciencias Exactas, Ingenierı ´a y Agrimensura, Universidad Nacional de Rosario, Rosario, Argentina. E-mail: santiago.r.soler@gmail.com 5 School of Geodesy and Geomatics, Institute of Geodesy and Geophysics, Wuhan, China. E-mail: chenwenjinwhu@gmail.com 6 Department of Mathematics and Geosciences, University of Trieste, Trieste, Italy. Pure Appl. Geophys. Ó 2016 Springer International Publishing DOI 10.1007/s00024-016-1376-y Pure and Applied Geophysics