Tsunamigenic-seismogenic structures, neotectonics, sedimentary processes and slope instability on the southwest Portuguese Margin
Introduction
The 1st November 1755 Lisbon earthquake and tsunami destroyed a large part of Lisbon and caused great damage on the southern coast of Portugal and northern Morocco. The earthquake intensity can be estimated at IX–X in Lisbon (moment magnitude Mw=8.7; Baptista et al., 1998a). Hydrodynamic modeling of the Lisbon 1755 tsunami event was carried out using the tectonic source of the 28th of February 1969 earthquake (Ms=7.9), which was recorded in Lisbon tide gauges and the epicentre of which was located south of the Gorringe Bank in the Horsehoe Abyssal Plain (Fukao, 1973). However, the results obtained by this modelling are not compatible with the historical observations on travel time, wave amplitude and polarity reported from Portugal, southern Spain and northern Morocco. Baptista et al. (1998b) proposed alternative sources for the 1755 Lisbon earthquake, according to which the fault should trend approximately N–S, should be over 200 km long with high displacements on the seafloor (>20 m) and should be located closer to the Portuguese coast than the Gorringe Bank.
During the AR92 survey a multi-channel seismic (MCS) line (AR92-10; see Zitellini et al., 1999) was acquired across the epicentre location of the 1755 Lisbon earthquake proposed by Udias et al. (1976). This seismic profile reveals a landward dipping listric thrust fault, the Marquês de Pombal Fault (MPF), with an offset of the seafloor of about 1000 m along the base of a NNE–SSW trending morphological scarp. The MPF is located offshore the southernmost part of the western Portuguese coast, between the Gorringe Bank and Cape S. Vicente. The MPF was clearly imaged on MCS profiles acquired during the recent BIGSETS project (Zitellini et al., 2001). Although the location, orientation and structural style of the MPF fits in with the seismogenic models proposed by Baptista et al. (1998b) and Ribeiro (1994), the size of the fault – 60 km long parallel to the strike direction – is far too small, according to the same models. This length can be enlarged to almost 100 km if the northern termination (a flexure) is assumed to be associated with a blind thrust. Nevertheless, the size of the structure remains too small to account for the intensity of the 1755 Lisbon earthquake.
The relocation of seismic events recorded in the area by the National Seismic Networks of Portugal and Spain shows that earthquake hypocentre depths can be larger than 30 km but shallower than 65 km (Cabral, 1995, fig. 51). The 26th March 2000 event (M=4.7) was located at 30 km depth and, further west, near the Gorringe Bank, large earthquakes were recorded at a depth of 70 km (Grimison and Chen, 1986). Based on these data, a thickness of 60 km for the schizosphere was used for the computation of the rupture areas of the 1755 Lisbon earthquake. This value should be envisaged as an approximate/conservative value since only few hypocentres have been determined in the study area. Considering that the 1755 event was an intra-plate earthquake with high stress-drop, an average 20 m slip could be attained for a Mw=8.7 event; based on this, the minimum length of the fault that caused the 1755 Lisbon earthquake should be 150 km. This value suggests that additional rupture areas have to be associated with the MPF to generate such a destructive earthquake.
The pattern of seismic activity off southwestern Iberia is diffuse with a few patches of concentrated seismicity in the Gorringe Bank (particularly its southern flank), the Guadalquivir Bank and the S. Vicente Canyon (see for example Buforn et al., 1988). The focal mechanisms obtained offshore southwestern Iberia from P-wave polarity and the study of small and large magnitude events show a consistently NW–SE direction for the maximum inferred compressive stress (Zitellini et al., 2001). Most of the mechanisms are strike-slip but a few of the events are dip-slip. Near the MPF, all events are strike-slip, including the preliminary solution for the 26th March 2000 earthquake (−9.731, 36.444°, source: F. Carrilho, pers. commun.). If this latter event is related to the activity in the MPF, then it can be interpreted in terms of regional tectonics, considering seismic activity at a transfer fault located at the southern termination of the MPF. The right-lateral strike-slip component of this event suggests that the hypocentre was located in the MPF foot-wall and that this is undergoing downward movement.
Recent analysis of onshore cores recovered from alluvial deposits near the coast has revealed the existence of tsunami deposits between fine-grained sediments. These deposits were dated and are supposed to have been deposited by the 1st November 1755 Lisbon tsunami in areas located ca. 2000 km apart, such as on the southern coast of Portugal (Boca do Rio, Algarve) and Scilly Islands (southwest England) (Dawson et al., 1995).
Various explanations are possible for this. Either the palaeogeographic conditions that allowed the formation of these deposits were unique, or the previous ‘1755-like Lisbon earthquake’ occurred more than several thousand years ago. If the latter explanation is correct, then the seismic cycle is very long and the return period for such an extreme event is very large (Ribeiro (1994) suggests a return period between 300 and 1500 years). It may therefore happen that after 245 years of the main shock we are still in a post-seismic gap of activity. Nevertheless, the possibility of disruption of a previous deposit by the 1755 Lisbon event should not be discarded. These data are compatible with the age of the youngest turbidites from one gravity core sampled during the BIGSETS campaign in front of the MPF, which yielded an age of <349 yr BP (A. Correggiari, pers. commun.). The seismicity pattern and the relative paucity of large magnitude earthquakes probably means that the active tectonic structures that currently show little activity may be the ones that ruptured during the 1755 event. This could be the case for the MPF and also for other structures identified north and northwest of it.
Since the MPF is not large enough to generate an earthquake with the magnitude of the 1755 Lisbon earthquake and there is not a single structure in the area that could account for the generation of such an event, it has been suggested that the MPF could be connected to other faults by means of transfer faults or relay deformation zones (Ribeiro et al., 2000). Ribeiro (1994) suggests that the 11th November 1858 Sines earthquake (see Fig. 1 for location) with an estimated moment magnitude 7.1 (Johnston and Kanter, 1990) and the 1755 Lisbon earthquake were generated by the same seismogenic structure. Thus, it was decided to investigate whether the 65-km-long Pereira de Sousa Fault (PSF; Fig. 2) constituted the northward prolongation of the MPF, offset to the east across a transfer fault zone.
The main objectives of the Training Through Research-10 (TTR-10) cruise off the southwest Portuguese coast were: (1) to investigate the MPF and its possible prolongation to the north along the PSF; (2) to study the deformation along the active MPF and PSF zones; and (3) to assess slope instability structures associated with these fault zones. For this purpose, an OKEAN mosaic, single channel seismic profiles and one ORETECH line were collected in the study area (Fig. 1).
The continental shelf between the Setúbal and the Cape S. Vicente canyons displays a uniform gradient. The shoreline is mainly made up of living and uplifted cliffs with pocket beaches related to small river valleys. The inner shelf is a narrow wave-cut terrace with a thin package of recent sediments draping a regional unconformity cutting through the Mesozoic sediments or Palaeozoic basement (down to 120 m deep). The external shelf is overlain by prograding sediments of probable Pliocene–Quaternary age depicting a moderately marked convex slope. The area between 200 and 700 m in depth is a large accumulation of sediments of moderate dip, with no pronounced outer edge except at the northern and southern limits where the Prı́ncipes de Aviz and the Descobridores Mountains are bounded by steep ‘cuestas’ (Mougenot, 1989; Fig. 1).
In contrast with the lateral continuity of the continental shelf, from a morphological point of view the continental slope between the Setúbal and the Cape S. Vicente canyons can be subdivided into three different sectors. In the northern sector the main morphological feature is the Setúbal Canyon, which together with the Tagus Canyon forms the main sedimentary feeding conduit into the Tagus Abyssal Plain. A central sector extends from 37°50′N to 38°20′N. It consists of a westward gently dipping, slightly convex depositional surface of sediment aggradation linking the inner shelf to the Tagus Abyssal Plain. The southern sector extends from the Cape S. Vicente Canyon to 37°50′N and displays a very complex morphology comprising active fault scarps, exhumed fault scarps and an interior deep basin (Rincão do Lebre; see Fig. 1). This basin is separated from the Tagus and Horseshoe Abyssal Plains by the eastern end of the Gorringe Bank and by the Marquês de Pombal uplifted block southern flank of the central sector of the study area which was uplifted during Miocene times.
This work documents that the morphological differentiation of the continental slope with respect to the continental shelf is intimately associated with its structural compartmentalisation and the differential tectonic evolution of the study area from Middle Miocene through Present times.
Section snippets
The dataset
The dataset acquired during the first part of Leg 1 of the TTR-10 cruise off the southwestern Portuguese Margin consists of: (1) 470 km of single channel seismic data and 3.5-kHz high-resolution sub-bottom profiles; (2) ca. 5500 km2 of 10-kHz OKEAN long range side-scan sonar (SSS) lines; (3) 90 km2 of 30-kHz ORETECH deep-towed SSS lines; and (4) 45 km of deep-towed high-resolution sub-bottom profiling (Fig. 1).
The source of the seismic profiles consisted of one 3.5-l airgun on Lines PSAT
Seismogenic and tsunamigenic structures, active tectonics
In this work faults are considered to be active if they record movements of their relative blocks during Pliocene–Quaternary times. Without a very precise stratigraphic calibration of the seismic profiles, in which only the position of the Middle Miocene horizon is known, other criteria for assessing the Quaternary activity on faults have been chosen, such as: (1) deformation of the most recent sediments on seismic sections displaying continuous stratigraphy above the Middle Miocene horizon;
Conclusions
It is concluded from interpretation of the data acquired during the TTR-10 cruise that the PSF and the TTR-10 fault are active faults, i.e. they have been active during Pliocene–Quaternary times. The present data and interpretation confirm that the MPF is an active fault (as already demonstrated by Zitellini et al., 2001) and it is also proposed that the PSF and TTR-10 faults (Fig. 2) are also active or are experiencing passive uplift and control the mass transport sedimentation in the study
Acknowledgments
This work and the TTR-10 cruise were funded jointly by the UNESCO/IOC, the University of Moscow and the Portuguese Geological and Mining Institute (through the INGMAR Project, financed by the Portuguese Foundation for Science and Technology). The BS MCS were acquired by the BIGSETS project, Contract No. EMV4-CT97-0547 EC project of the Environment Climate Program 1994–1998, Technologies to Forecast, Prevent and Reduce Natural Risk. The authors also thank Captain and crew of the R/V Professor
References (21)
- Banda, E., Torné, M., IAM group, 1995. IAM group investigates deep structure of ocean margins. EOS Transactions,...
- et al.
Constraints on the source of the 1755 Lisbon tsunami inferred from numerical modelling of historical data
J. Geodyn.
(1998) - et al.
The 1755 Lisbon tsunami; evaluation of the tsunami parameters
J. Geodyn.
(1998) - et al.
Seismicity, source mechanisms and tectonics of the Azores–Gibraltar plate boundary
Tectonophysics
(1988) - Cabral, J., 1995. Neotectónica em Portugal Continental, Memórias do Instituto Geológio e Mineiro 31, Lisbon, p....
- et al.
Tsunami sedimentation associated with the Lisbon earthquake of 1st November AD 1755: Boca do Rio, Algarve, Portugal
Holocene
(1995) Thrust faulting at a lithospheric plate boundary. The Portugal earthquake of 1969
Earth Planet. Sci. Lett.
(1973)- Gracia, E., Dañobeitia, J.J., and HITS cruise party (2001). High-Resolution Imaging of Tsunamigenic Structures in the...
- Gracia, E., Danobeitia, J.J., Vergés J., Cordoba, D., Parsifal cruise party. Mapping active faults at the SW Iberia...
- et al.
The Azores–Gibraltar plate boundary: focal mechanisms, depths of earthquakes and their tectonic implications
J. Geophys. Res.
(1986)
Cited by (132)
A numerical methodology for estimating site-specific cascading earthquake and tsunami dynamic loading on critical infrastructure
2024, International Journal of Disaster Risk ReductionSurface water dynamics of the last 20 kyr documented by coccolithophores in the Gulf of Cadiz
2023, Palaeogeography, Palaeoclimatology, PalaeoecologyCascade earthquake and tsunami hazard assessment: A deterministic perspective for engineering purposes
2022, International Journal of Disaster Risk ReductionThe sines contourite depositional system along the SW Portuguese margin: Onset, evolution and conceptual implications
2020, Marine GeologyCitation Excerpt :These collision events caused the closure of the Rifian and Betic oceanic gateways between the Atlantic and the Mediterranean from ~7 to 6.1 Ma ago (Hsü et al., 1973; Sierro et al., 2008; Flecker et al., 2015; Capella et al., 2018), which led to the progressive isolation of the Mediterranean region (Hsu et al., 1978; Duggen et al., 2003) and the Messinian Salinity Crisis between 5.96 and 5.33 Ma (Cita, 2001; Bache et al., 2012). In the Late Miocene, the extensional regime changed to an oblique NW-SE convergence (Dewey et al., 1989; Ribeiro et al., 1990; Terrinha et al., 2003) while the collapse of the Betic-Rif orogenic front occurred, probably associated with westward roll-back subduction of an oceanic lithosphere slab beneath the Gibraltar Arc and development of its accretionary wedge (Gutscher et al., 2002, 2012; Duarte et al., 2013). During this last period, a transtensional regime induced the reopening of the connection between the Atlantic and the Mediterranean through the Strait of Gibraltar, putting an end to the Messinian Salinity Crisis (Maldonado et al., 1999; Duggen et al., 2003; Garcia-Castellanos et al., 2009).