GFAST-NZ: Rapid Geodetic Inversion for Finite Faults in New Zealand

Abstract New Zealand’s (NZ) unique tectonic setting generates a variety of natural hazards such as earthquakes and tsunamis. Previous studies have shown that tsunamis driven by slip at shallow depth during earthquake rupture are extremely dangerous to coastal populations. Therefore, rapid characterization of earthquakes and their tsunamigenic potential is necessary. In this context, the use of continuous Global Navigation Satellite System (GNSS) has demonstrated its efficacy in addressing these types of challenges, since, unlike velocity records, it does not saturate, and unlike accelerograms, it does not require a complex correction. In this study, we present the current status of the implementation for New Zealand of the geodetic first approximation of size and time (G-FAST) algorithm, referred to as GFAST-NZ. This algorithm is capable of estimating the moment magnitude, a moment tensor solution, as well as the fault geometry and corresponding slip. We demonstrate the capabilities and limitations in the context of NZ. This is relevant given NZ’s complex tectonic setting, which includes diverse fault geometries, such as the strike-slip Alpine fault and the Hikurangi subduction zone. We present the detailed results for one historic earthquake and two representative synthetic cases, that is, large earthquakes along the Hikurangi subduction zone and the Alpine fault. For the synthetic cases, we simulate real-time GNSS data for the GeoNet network, which is composed of approximately 200 GNSS stations, 62 of which are processing in real time. The results demonstrate that the GFAST-NZ algorithm applies to this representative set of geometries in NZ’s geological context. However, for the Alpine fault, there is considerable uncertainty due to sparse GNSS network coverage and complex along-strike segmentation. Although the results are satisfactory, robust rapid characterization of the full range of fault geometries in NZ is best achieved through using complementary tools and multidata approaches.