Limitations of the Resolvability of Finite‐Fault Models Using Static Land‐Based Geodesy and Open‐Ocean Tsunami Waveforms

Abstract

AbstractFinite‐fault inversions are a common technique, employed following large earthquakes, used to understand the nature of slip along a fault. Using multiple data sets, including static offsets from geodetic instruments and tsunami wave heights from open‐ocean gauges, a richer perspective on the expected slip distribution than using a singular tool is created. However, the model resolution obtained from open‐ocean tsunami data and techniques used to subsample that data have not been widely evaluated. Static geodetic data can provide near‐complete model resolution of the subduction megathrust near the trench, if data are local. However, model resolution falls off precipitously as distance between instrument and fault increases when geodetic data are limited to more distal on‐land sites. Tsunami data derived from open‐ocean waveforms are less dependent on station to fault distance, but offshore model resolution is lost due to necessary data processing, such as windowing, often necessary to avoid coastal reflections. This primarily affects the resolution in the downdip direction, which often arrives at open‐ocean stations later in the waveform. Spatial detail is also limited by the minimum subfault size that will satisfy the longwave approximation, which is dependent on the water depth. For most cases, this subfault limit is approximately 20 by 20 km. In many environments, the sparsity of offshore geodetic instruments, and the large distances between estimated slip and coastal geodetic gauges, makes the inclusion of open‐ocean data, if available, highly advantageous. Still it is possible even with the pairing of on‐land and offshore data sets for poorly resolved zones to exist. In these cases further resolution can be recovered through the incorporation of additional data sets such as strong‐motion data and seismic waveforms through a seismic‐geodetic inversion.