NEW GPS EVIDENCE FOR CONTINENTAL TRANSFORM FAULT CREEP, CENTRAL RANGE FAULT, TRINIDAD, AND ITS GEOLOGICAL AND HAZARD IMPLICATIONS

Most continental plate boundary transform faults exhibit stick-slip behavior and are locked during the interseismic period. Important exceptions include the Hayward, Parkfield, and central segments of the San Andreas fault. Studying how and why fault creep occurs has implications for our understanding of how faults work in general and how to better assess their associated seismic hazards. Trinidad is located along the Caribbean-South American (CA-SA) dextral transform plate boundary. Using triangulation-to-GPS geodetic measurements, we previously discovered that the 072° striking Central Range Fault (CRF) is the active transform in Trinidad, accommodating ~two-thirds of the total ~20 mm/yr CA-SA plate motion with an interseismic slip rate of 12 ± 3 mm/yr. We build on this earlier work using new GPS data collected between 1994 and 2014 from a set of ~25 continuous and episodic GPS stations. 2-D elastic fault dislocation modeling of the new GPS velocity field gives a best-fit (reduced chi-squared = 4.119) interseismic slip-rate of 14 ± 1 mm/yr and a fault locking depth of 0.083 ± 0.153 km (model parameters given at 1σ uncertainty determined via empirical bootstrapping method). The shallow locking depth is well resolved, indicating that the CRF largely creeps. However, a sparsity of data along the eastern part of the Central Range fault leaves open the possibility of some interseismic locking there. Our new results are consistent with the one prehistoric (~2.7 - .55 ka) earthquake trenched along the CRF being of small magnitude or possibly related to a slow-slip event, and with the low permanent neotectonic, strike-slip strain recorded/observed in the walls of the fault. We propose that a rigid northern block is juxtaposed against a weak and thick Paleogene-shale-rich covered southern block along a sharp 072° striking inherited (?) boundary, along which the neotoectonic strain concentrates. The lack of sharp, conclusive creep offsets that we searched for, but not observed on the ground could mean that the CRF creeps across a relatively broad (dm-scale?) zone. We plan to establish a series of creep arrays and a network of microseismometers and additional cGPS stations to better quantify the behavior of the CRF.