1. Comparison of the different proposed sources for the 1755 Lisbon tsunami: modeling in the West Antilles. Jean ROGER, M.A. BAPTISTA, H. HEBERT ITS 2009, Novosibirsk, Russia

2. Introduction 1755: reminder concerning earthquake and associated tsunami  estimated magnitude 8.5-9.0  lot of casualties (60000 dead people) and destructions  waves reaching more than 15 m locally (Cadiz Gulf, Morocco)  reports from Morocco to England of tsunami arrival  Reports of waves causing destructions and loss of life in the Açores (Andrade, 2006) and in the West Indies (Lander et al., 2002)  Source(s) not well established Gutscher et al., 2002 Source ????

3. Contribution of far field effects study

4. Historical data in the West Indies  Several coeval documents from 1755 in about 10 places in the Caribbean Islands.  Reports of important inundation in some places / with a lot of details  Be careful concerning the inhabited areas at this time compared to today’s -1745: 80000 inhab. Incl. 65000 slaves -Today: around 380000 inhab. 1759 Document-based maregram

5. Modelling vs. Historical reports  Negligible effect on Lesser Antilles 10 cm vs. 2-3 m

6. Barkan et al. (2008)’s source n°5  But this source has NO geological evidence  Shows a wave amplification in the West Indies Use of a multigrids’ model  Increasing resolution until 40 m Construction of high resolution grids

7. Zitellini et al., 1999; Baptista et al., 2003 2 segments: Marques de Pombal + Guadalquivir Bank  Look at the strike of the 2 segments Mw=8.5 / geological evidence

8. Gutscher et al., 2006 Cadiz Wedge / Mw=8.3  Nothing relevant in the West Indies

9. The resonance effect What ?  phenomenon of trapping and amplification of wave energy (Woo et al, 2004) Where ?  semi-enclosed water body: harbor, lagoon, bay, fjord, etc. When ?  the period of arriving waves ≈ the eigenperiod of the water surface of the considered water body (Bellotti, 2007) Harbor resonance: every harbor has a natural oscillation mode with eigenperiod depending on physical characteristics of the water body (Jansa et al., 2007) i.e. its geometry and depth (Monso de Prat and Escartin Garcia, 1994) Some submarine features can induce resonance effects as shelf resonance Supposed resonance How can we explain such differences between modelled amplitudes ?

10. Bathymetric grids Palma Harbor, Spain 24/11/2015 High resolution grids (10 m)  made from georeferenced and digitized nautical maps N N  With respect of harbor structures  All the grids are adapted to the imbrication used by our modelling code during calculation Jijel Harbor, Algeria Roger and Hébert, 2008 Yelles et al., 2009 Ste Anne’s Bay, Guadeloupe

11. First step  look at the maximum sea elevation (Hmax); use of numerical tide gages If it does not explain the reported amplifications in specific areas, The most resonant site for tsunami arrivals in Balearic Palma Harbor, May 2003

12. 2 nd step: frequency analysis Accurate study of recorded signal  FFT (Fast Fourier Transform)  reveals frequency peaks Sahal et al., submitted Signal near the Source Harbor Arrival on Coast

13. Use of synthetic signals (Ampl. and period known)  Determine harbor eigenperiod  spectral analysis of tide gage records and eigenvalues analysis  Thus we know what kind of signal is able to make the water body to react

14. 3rd step Go back to the geometric parameters of the source...  Inverse problem: eigenperiod of harbor (natural and harmonics) Knowledge of arriving signal What kind of deformation is able to generate such signal Be careful to other type of resonance Comeback over source’s parameters

15. Conclusion Considering far-field effects could be a good contribution to the estimation of the source’s parameters Strike angle not seems to be a major factor to wave amplification in the West Indies (  In each tested case: No main energy paths towards the West Indies) Underlines the need of high resolution bathymetric datasets Frequency analysis: this method could explain the great sea elevation values in Antilles during 1755 event and allows to select one or several possible sources We could protect harbors using these eigenvalues knowledge (Monso de Prat and Garcia, 1994) Today we are not able to reproduce amplification exactly

16. Perspectives Try to find source’s parameters able to generate far field effect in agreement with historical reports and compare them to proposed source’s parameters, taking into account the eigenperiods of studied harbors Look at other places concerning by far-field effects of this 1755 tsunami and do inverse modelling

17. спасибо ! project MAREMOTI Work funded by the French ANR project MAREMOTI MAREograph and field tsunaMi observations, mOdeling and vulnerabiliTy studIes for Northeast Atlantic and western Mediterranean