F. Carrilho, P.M. Alves, D. Vales, J.A. Pena, I. Abreu, S. Cortês

Instituto de Meteorologia, I.P., Rua C ao aeroporto, 1749-077 Lisboa, Portugal

Introduction

On the 2007.02.12 (10:35 UTC) an earthquake of magnitude 5.9ML occurred approximately 160 km SW of San Vincent Cape, Portugal mainland. The earthquake was felt almost throughout mainland Portugal, the Madeira archipelago, and parts of Spain and Morocco, with felt reports up to 1000 km away from the epicentre. The macroseismic field extended for 3x106 km2, with a maximum intensity of 5 (EMS98) assigned to the Algarve region (Southern part of Portugal mainland) and part of Lisbon. This earthquake and its aftershocks were recorded by the Portuguese seismic network, and it can be considered as the best ever instrumentally recorded at the Portuguese national seismic network (Figure 1). A preliminary fault plane solution was computed from P-wave polarity inversion, pointing to an oblique reverse mechanism, which is in good agreement with the regional tectonic setting.

Figure 1. - Some records of the 2007.02.12 10:35 (UTC) earthquake recorded by the Portuguese national seismic network

1. Tectonic setting

Portugal is located at the western part of Iberian Peninsula. West of Portugal, the N-S oriented continental margin, related to the Atlantic opening and considered passive, crosses a tectonic boundary oriented E-W, between the Africa and Eurasia plates. Focal mechanisms and plate kinematics studies show that the Gorringe Bank (GB) and the Cadis Gulf (GC) are consistently characterised by compression (McKenzie, 1972; Auzende et al., 1978; Grimison & Chen, 1986, Cabral, 1993). However, it is hard to follow the precise location of the plate boundary close to Iberia. This is usually interpreted as the result of the relatively low inter-plate motion given by most kinematic plate models (e.g., De Mets et al., 1994; Sella et al. 2002; Fernandes et al., 2003), reaching about 4 mm/year along the NW-SE to WNW-ESE oriented eastern segment of the Azores-Gibraltar tectonic boundary.

The Portuguese Instituto de Meteorologia (IM) network located the 2007.02.12 earthquake on the Horseshoe plane NW of the major Horseshoe thrust (HSF) trace (see Figure 2). This is an area where the major earthquake of 1969.02.28 (8.0MS [USGS]) took place and where the 1755 “Lisbon” earthquake (Magnitude 8¾ [Abe, 1979]) might have been originated. The dominant active structures in this region are the Gorringe Bank (GB), the Marquês de Pombal Fault (MPF), the Horseshoe Fault (HSF) and the Guadalquibir Fault (GQ) [Figure 2], which have been studied by several authors (Baptista et al., 2003; Terrinha et al., 2004; Zittelini et al., 2004). The 2007.02.12 (10:35 UTC) event is an expression of this interplate domain.

Figure 2. - Comparison between the fault plane solution obtained in this work by polarity inversion of 133 polarities from regional and global data (IM), and the fault plane solutions obtained from moment tensor solutions from regional and global data which were rapidly disseminated (ETHZ - Swiss Seismological Service, Zurich; INGV - MEDNET network, Instituto Nazionale di Geofysica e Vulcanologia, Italy, IGN - Instituto Geografico Nacional, Spain; USGS - United States Geological Survey; HARV - Department of Earth and Planetary Sciences, Harvard University). Tectonic structures adapted from Terrinha et al. (2003), Matias et al. (2004) and Rovere (2002) [HAP - Horseshoe Abissal Plain; HSF - Horseshoe Fault; GB - Gorringe Bank; MPF - Marquês de Pombal Fault; PSF - Pereira de Sousa Fault; GQ - Guadalquivir Fault]

2. Seismic monitoring

IM has recently introduced significant improvements in the Portuguese seismic network through a modernization project of the National Seismic Network (MODSISNAC). Since July 2006, seven new broadband stations have been installed on the mainland (Figure 3), connected in real-time by VSAT to the IM headquarters, in Lisbon. The data coming from two of them (PESTR and PVAQ) is sent to ORFEUS and IRIS/DMC in real-time. All of the new stations are equipped with very broadband sensors (3 Guralp 3T , 3 Guralp 3ESPCompact and 1 STS-2) and 24 bit digitisers (6 Guralp DM24 and 1 Quanterra Q330). Also, a short-period (SP) network is still in operation, with 15 stations installed on mainland (12) and Madeira archipelago (3), transmitting data by VSAT, public telephone lines and radio (UHF).

Figure 3. - Portugal mainland seismic network operated by the Instituto de Meteorologia

IM has adopted the SEISCOMP system (GEOFON, Potsdam) as the standard for data acquisition. Some of the older SP stations have already been integrated into the new acquisition system, and efforts are being made in order to integrate the remaining ones. Up to now, data from 17 seismic stations are concentrated at a central computer at IM, including available regional stations operating in real-time (RT) (MTE, from GEOFON with support from IM; EVO, from Évora University; MORF from Lisbon University, with support from IM) and part of the SP network. Also, data from other regional stations is being obtained from the ORFEUS Data Centre, using AutoDRM protocol. The data is analysed using SEISAN (Havskov & Ottemöller, 2005).

With the availability of near real-time data, automatic processing is now possible. Two minutes after the 2007.02.12 event, 10:35:25 UTC, an automatic location was released by the AUTOLOC procedure (SEISCOMP) running on a test-bed at IM head-quarters. The manual solution was released 14 minutes after the earthquake occurred.

3. Hypocentre and Fault plane solution

The local magnitude computed for the 2007.02.12 earthquake was 5.9ML. The hypocentre location (35.933ºN; 10.496W; 37 km depth) [Figure 4] was computed using 102 arrivals from 83 regional stations. The confidence ellipse (90%) has a major semi-axis of 4.3 km, oriented according to azimuth 135º, and a minor semi-axis of 2.5 km. The estimated hypocentre depth of 37 km is poorly constrained (±12km, computed at 90% confidence level) mostly due to the lack of nearby stations. However, centroid moment tensor inversions disseminated by several agencies pointed to a depth of about 42 km (e.g. HARV).

Figure 4. - Seismic activity in the 2007.02.12 epicentre area, between January 1961 and February 2007 (IM seismic catalogue 1961-2000 [Carrilho et al., 2004] and seismic bulletins). Red circles represent the epicentres within 150 km of the 2007.02.12 epicentre, and grey circles represents epicentres outside of 150 km distance from the referred epicentre.

Several source mechanisms based on rapid moment tensor inversion were released within a few hours after the event (IGN, HARV, ETHZ, USGS, INGV), where oblique reverse mechanisms type are determined but with small differences between them (Figure 2). Most of the moment magnitude determinations points to 5.9MW (HARV, ETHZ, USGS and INGV), which is in good agreement with the 5.9ML evaluation from IM. The exception is IGN, which computed 6.2MW.

A preliminary fault plane solution was obtained using a method based on the inversion of the P-wave polarities (EXAUST algorithm, modified by Matias [pers. comm.] from Borges [1991]). A total of 133 polarities from regional and global stations were used. The solution points to a reverse mechanism with a slight strike-slip component (Figure 2). The fault planes have the following parameters: NP1- Strike 100º/Dip 40º/Slip 130º; NP2- Strike 232º/Dip 60º/Slip 62º. The resulting pressure axis has an orientation~N-S (Azimuth 342º).

The preliminary focal mechanism computed in this work, showing oblique reverse faulting, is similar to what was obtained for the 1969 (7.5MS; Fukao, 1973) and 2003 (5.3MW; Stich et al., 2005) events that occurred nearby, but distinct from the one computed for the 2004.12.13 (5.4ML; Carrilho, 2005; 4.8MW; Stich et al., 2005), which corresponds to an event occurred some km to north-east. When considering the best epicentre solutions computed by the regional networks (IM and IGN), the correlation with significant major structural features, such as the Horseshoe thrust (HSF) is not straightforward. In fact, one of the nodal planes has an orientation compatible with HSF surface trace and dip orientation, but the epicentre is located NNW of that structure.

4. Seismic activity

The seismic activity observed in the Horseshoe abyssal plain is the result of the interaction of the Eurasia and Africa plates. Although most of the earthquakes are low to moderate magnitude, the offshore area located south-west of Portugal mainland coast is occasionally struck by strong earthquakes (M>6.0). Some of those larger earthquakes generated tsunamis. Examples of this activity are the 382 DC and the great 1755 “Lisbon” earthquakes. In more recent years, important instrumental events occurred here, like the mentioned 1969.02.28, which was the largest before the present event.

Figure 4 shows significant seismic activity in the surrounding area, with 3 clusters of epicentres well identified. Figure 5 shows the distribution of earthquakes over time, where the developments of the seismic network explain the changes in detected activity. As it can be seen, the present earthquake was the largest in the last 38 years.


Figure 5. - Seismicity distribution over time, inside a region of 150 km around the epicentre of the 2007.02.12 earthquake

In the 10 days following the earthquake, only 6 aftershocks were recorded, with magnitudes ranging from 2.0 to 3.7ML, significantly lower than the main shock. This pattern is consistent with what was observed for the last larger earthquakes, like the July 29th, 2003, but very different from what happened in 1969, where significant aftershock activity was observed (Figure 5), which could be related with the large differences in magnitude of the main shocks.

5. Macroseismic effects

The earthquake caused no damage, but it was widely felt with its macroseismic effects reaching large distances such as Madeira archipelago (~700km away from the epicentre region), Morocco and Spain, with felt effects at Santiago de Compostela (~800km) and Zaragoza (~1000km).
A Web inquiry (http://www.meteo.pt/pt/sismologia/inquerito/sism_inq.jsp) was used by IM to collect macroseismic information, which proved to be a success with approximately 600 reports received.

The estimated macroseismic field extends over an area of 3x106 km2. The maximum intensity of 5 (EMS-98) was reported from the Algarve region, southern part of Portugal mainland. Intensity 5 was also consistently assigned for one part of Lisbon, Campo Grande (Figure 6 and Figure 7). At Madeira archipelago, the earthquake was felt in city of Funchal with intensity 3. The punctual intensities felt over mainland are shown in Figure 6. It is very interesting to note that the intensities are extremely spread. According to EMS guidelines, reports concerning more than 5th floor were not considered. Nevertheless it was common that people in the same building or nearby zones felt the earthquake with a difference of one or even two degrees of intensity. Direct causes of these differences can be related to the site geology, building construction, human activity (usually high for that hour of the day) and human perception. The geographical pattern of responses follows the demographic distribution and Internet accessibility. It is noticeable that the region of Lisbon is consistently more affected (Figure 7), partially due to its higher vulnerability. But other causes can be considered present and should be studied in detail because this pattern is very common over available macroseismic history.

Figure 6. - Intensity distribution (EMS98) on Portugal Mainland - Effects of the 2007.02.12 earthquake

Figure 7. - Lisbon macroseismic effects (EMS98), from a preliminary assessment

6. Conclusions

The 2007.02.12 (10:35 UTC) earthquake, while causing no damage, generated a large macroseismic field and a maximum intensity of 5 (EMS98) in the Portuguese territory. Its epicentre area is close to the large - 7.5MS earthquake on 1969.02.28. Few aftershocks were recorded during the following days. The fault planes computed from an inversion of a set of regional and global polarities, shows a reverse type mechanism with a component of strike-slip, similar to the solutions provided by other agencies from moment tensor inversion.

The event was recorded by the new broadband stations which have been recently installed by IM in Portugal mainland, and it can be considered as the best ever instrumentally recorded earthquake in Portugal.

The availability of extensive reports allowed the reformulation of all macroseismic data acquisition and analysis. New software tools were developed to gather and display macroseismic data based on the smallest administrative divisions of the Portuguese territory. Current outputs are GMT (Wessel and Smith, 1998) maps or layers for GoogleEarth® program. We intend to continue this strategy, aiming to compile automatically incoming e-mails with questionnaires in order to show them in a graphic mode in the operational seismic centre to be used as preliminary macroseismic data in the alerts emitted by IM in the first minutes after the event. Validation seems still an important point and we noticed that questionnaires arriving in the first minutes are the most reliable of all.

Acknowledgements

The authors wish to thank Dr. Jose Morales, from Instituto Andaluz de Geofísica, Universidad de Granada, Spain, for providing part of the data. The authors also appreciate valuable comments given by Dr. Miguel Miranda. This work was partially supported by the Portuguese Science and Technology Foundation (FCT), project MODSISNAC.

References

Abe, K., (1979), Size of Great Earthquakes of 1837-1974 Inferred from Tsunami Data, J.Geophys. Res. 84 (NB4): 1561-1568.

Auzende, J.M., Olivet, J., Charvet, J., LeLann, A., Le Pichon, X., Monteiro, J., Nicolas, A., Ribeiro, A., (1978), Sampling and Observation of Oceanic Mantle and Crust on Gorringe Bank, Nature, 273, pp45-48.

Baptista, M.A., Miranda, J.M., Chierici, F., Zitellini, N., (2003), New Study of the 1755 Earthquake Source Based on Multi-Channel Seismic Survey Data and Tsunami Modeling, Natur. Haz. and Earth System Science, 3, 330-340.

Borges, J.F., (1991), Métodos Automáticos na Determinação de Mecanismos Focais, Relatório de Licenciatura, Fac. Ciências, Univ. Lisboa, pp 93.

Buforn E., M. Bezzeghoud, A. Udías, and C. Pro (2004), Seismic sources on the Iberia-African plate boundary and their tectonic implications, Pure Appl. Geophys. 161, doi: 10.1007/s00024-003-2466-1.

Cabral, J., (1993), Neotectónica de Portugal Continental, tese de doutoramento, Fac. Ciências, Univ.Lisboa, Portugal.

Carrilho, F., Nunes, J.C., Pena, J., Senos, M.L., (2004), Catálogo Sísmico de Portugal Continental e Região Adjacente para o período 1970-2000, Instituto de Meteorologia, ISBN 972-9083-12-6.

Carrilho, F., (2005), Estudo da Sismicidade da Zona Sudoeste de Portugal Continental, M.S. Thesis, Fac. Ciências, Univ. Lisboa, pp 172.

DeMets, C., Gordon, R.G., Argus, D.F., and Stein, S. (1994), Effect of Recent Revisions to the Geomagnetic Reversal Time Scale on Estimates of Current Plate Motions, Geophys. Res. Lett., 26, 1921-1924.

Fukao, Y., (1973), Thrust Faulting at a Lithosphere Plate Boundary: The Portugal earthquake of 1969, Eart Plan. Sci. Lett. 18, 205-216.

Fernandes, R. M. S., Ambrosius, B. A. C., Noomen, R., Bastos, L., Wortel, M., Spakman, W. & Govers, R., (2003), The relative motion between Africa and Eurasia as derived from ITRF2000 and GPS data, Geophys. Res. Lett., 30(16), 1828, doi:10.1029/2003GL017089.

Grimison, N.L. and Chen, W.P. (1986), The Azores Gibraltar Plate Boundary: Focal Mechanisms, Depth of Earthquakes and their Tectonic Implications, Journal of Geophyical Research, 91, 2029-2047.

Havskov, J., Ottemöller, L., (2005), SEISAN: The Earthquake Analysis Software, Version 8.1, Univ. Bergen, Norway.

Matias, L., Terrinha, P., Mendes-Victor, L.M., MATESPRO team, (2004), The MATESPRO Multibeam Cruise Reveals Unknown Recent Tectonics in SW Ibéria.

McKenzie, D.P. (1972), Active Tectonics of the Mediterranean Region, Royal Astronomic Society Geophysics Journal 30, 109-185.

Rovere, M., (2002), Strutturazione del margine atlântico ibérico ed inversione miocenica in prossimità del limite di placca Eurásia-Africa, Tesi di Dottorato, Università degli Studi di Bologna.

Sella, G., Dixon, T. H., & Mao, A.,( 2002), REVEL: a model for recent plate velocities from space geodesy, J. Geophys. Res., 107(B4), 2081, doi:10.1029/2000JB000033.

Stich, D., Macilla, F., and Morales, J. (2005), Crust-Mantle Coupling in the Gulf of Cadiz (SW-Iberia), Geophys. Res. Lett., 32, doi:10.1029/2005GL023098.

Terrinha, P., Pinheiro, L.M., Henriet, J.P., Matias, L., Ivanov, M., Monteiro, J., Akhmetzhanov; A., Volkinskaya, A., Cunha, T., Shaskin, P., Rovere, M., (2003), Tsunamigenic-seismogenic structures, neotectonics, sedimentary processes and slope instability on the southwest Portuguese Margin, Marine Geology, 195, pp 55-73.

Wessel, P., Smith, W., (1998), New, improved version of the Generic Mapping Tools Released, EOS Trans. AGU, 79, 579.

Zitellini, N., Rovere, M., Terrinha, P., Chierici, F., Matias, L., and BIGSETS Team, (2004), Neogene Through Quaternary Tectonic Reactivation of SW Iberian Passive Margin, Pure and Apl. Geophis., 161, pp 567-587.