Forward and inverse numerical modelling: complementary approaches to
better understand palaeotsunamis
Abstract
The 1755 Lisbon earthquake triggered the largest historical tsunami ever
registered in Western Europe. Despite the recent efforts to better
understand this event, there are still questions to be answered.
Understanding the past tsunami intensity is key to assessing tsunami
hazard. Sedimentary imprints are the only evidence in the geological
record able to quantify onshore tsunami flow characteristics through
inverse modelling. On the other hand, forward numerical modelling is a
powerful tool capable of simulating tsunami hydrodynamics and the
induced sediment transport. This work presents results from inverse and
forward modelling in order to assess tsunami characteristics onshore.
The study site is located on the Portuguese southern coast, at Salgados
lowland where inverse modelling was performed using TsuSedMod (Jaffe and
Gelfenbaum, 2007, Sedimentary Geology) based on four Livingstone
sediment cores. Forward modelling including tsunami generation and
propagation was performed respectively using the methods of Okada (1985,
Bulletin of the Seismological Society of America) and Delft3D-FLOW.
Onshore topography was corrected for the 1755 scenario based on
extensive deposit thickness data. The tsunami source was chosen based on
recent results from the same authors that pointed to a good correlation
between modeled and field tsunami data considering Marques de Pombal
fault. Results from inverse model show tsunami onshore average speed
varying from 7.3 up to 9.3 m/s and shear velocities from 0.52 up to 0.66
m/s. Varying the bottom roughness results in the forward model result in
average flow velocities between 7.0 and 8.0 m/s, induced by a 3-meter
high tsunami at 50 m depth. The good agreement between forward and
inverse model estimates of tsunami velocity highlights the potential of
numerical modelling (coupled with geological records) to improve the
understanding of historical events. Additional research on correlating
modelling and geological data is needed and will likely lead to a better
understanding of the effects of similar events and contribute to the
ability to assess tsunami hazard and coastal vulnerability.
Acknowledgements: Work supported by Instituto Dom Luiz and by project
OnOff ‑ PTDC/CTA‑GEO/28941/2017 – financed by FCT