Crustal and Upper Mantle Velocity Structure beneath Northwestern South America revealed by the CARMArray

The Caribbean plate (CAR) is a fragment of the Farallon plate heavily modified by igneous processes that created the Caribbean large igneous province (CLIP) between 110 and 80 Ma.The CAR collided with and initiated subduction beneath northwestern South America plate (SA) at about 60-55 Ma as a narrow flat-slab subduction zone with an accretionary prism offshore, but no volcanic arc. Large scale regional tomography suggests that 1000 km of the CAR has been subducted (Van Benthem et al., 2013, JGR). The flat slab has caused Laramide-style basement uplifts of the Merida Andes, Sierra de la Perija, and Santa Marta ranges with elevations >5 km. The details of subduction geometry of the CAR plate beneath northeastern Colombia and northwestern Venezuela are complicated and remain unclear. The region of slab steepening lies below the triangular Maracaibo block (Bezada et al, 2010, JGR), bounded by major strike slip faults and currently escaping to the north over the CAR. Geodetic data suggests the this region has the potential for a magnitude 8+ earthquake (Bilham and Mencin, 2013, AGU Abstract). To better understand the subduction geometry, we deployed 65 broadband (BB) stations across northeastern Colombia and northwestern Venezuela in April of 2016. The 65 stations interweave with the 32 existing Colombian and Venezuelan BB stations, forming a 2-D array (hereafter referred to as CARMArray) with a station spacing of 35-100 km that covers an area of 600 km by 400 km extending from the Caribbean coast in Colombia to the interior plains of Venezuela. With data from the first year of operation, we have measured the Rayleigh wave phase velocities and Z/H ratios in the period range of 8-40 s using both ambient noise and earthquake data recorded by the CARMArray. We also generated Ps receiver functions from waveform data of teleseismic events recorded by the array. We then jointly inverted the three datasets to construct a 3-D S-wave velocity model beneath the array. We will report the initial results of the inversion and discuss the lateral variations of crustal and upper mantle structure and their potential links with surface geology and regional tectonics.