Evaluation of different earthquake scaling relations on the generation of tsunamis and hazard assessment

Earthquake source details can hardly be determined accurately in a short time window after an earthquake. For rapid tsunami warning purposes, rupture parameters for a given earthquake magnitude are estimated using empirical scaling relations. In this paper, we evaluate the effectiveness of several commonly-used scaling relations. It is found that, for the 2011 Tohoku earthquake, most scaling relations overestimate the rupture area and underestimate the average rupture slip, resulting in predictions of earlier wave arrivals and smaller wave amplitudes in comparison to recordings. The scaling relation that uses “asperity size” instead of actual rupture area, presents a smaller rupture area and a larger average slip, leading to better tsunami predictions. While for the 2014 Iquique and 2015 Illapel tsunamis of relatively smaller earthquake magnitude, different scaling relations lead to comparable tsunami predictions. We further implement these scaling relations to study a potential earthquake in the Manila subduction zone. Results show that different scaling relations cause significant differences in the tsunami arrival time and wave amplitude at coastal cities, especially in the along-trench direction. Thus, to access the tsunami threat from mega-earthquakes and build early warning systems, it is necessary to select the appropriate scaling relations.