Seismotectonics of the Canary Islands

208 447 447- 452 Elsevier Science Publishers B.V., Amsterdam Short Note Seismotectonics of the Canary Islands J. Mezcua a, E. Buforn b, A. Udias b and J. Rueda a a Institute Geogrcifico National, M adrid, Spain h Dept. de c;esfiica, ~ni~ersidad ~ompiute~e, M adr~, Spain (Received 2 July 1991; revised version accepted 28 November 1991) ABSTRACT Mezcua, J., Buforn, E., Udias, A. and Rueda, J., 1992. Seismotectonics of the Canary Islands. Tectcmophy sics, 208: 447- 452. A revision of the seismicity, both historical and instrumental for the Canary Islands is presented. The occurrence on May 9, 1989 of an earthquake of magnitude 5.2 between the islands Gran Canaria and Tenerife, followed by a large number of aftershocks have been interpreted on seismotectonic grounds. The main conclusion is that a horizontal compressional stress regime in NW-SE direction is present in the region which is compatible with the tectonics in the northwestern part of the African continent. The Canary Islands are located at the border of the continental passive margin of northwestern Africa in one of the provinces of Tertiary volcanism present in this part of the Atlantic Ocean. Their origin is not clear and different theories have been proposed among others by Morgan (19711, Wilson (19731, Schimincke (1973), Anguita and Hernlin (19751, Fuster (19751, Carracedo (1984) and Arafia and Ortiz (1986). In the past, the main argument about the continental or oceanic character of the islands, was considered to be solved after a series of seismic refraction lines were carried out in the archipelago, which showed their oceanic structure (Banda et al., 1981; Surifiach, 1986). The Correspondence to: J. Mezcua, Institute Geogrefico National, Subdireccidn General de Astronomia y Geofisica, General Iblnez de Ibero 3, 28003 Madrid, Spain. 0040-1951/92/$05.00 tectonic stress field in the region, however, has been difficult to obtain due to the low seismicity level, and the lack of seismological instrumentation. Only recently, after 1975, a local network of seismographic stations has been installed and from that time to 1990, only very little activity has been recorded. Historically, however, important earthquakes have occurred related with volcanic eruptions. On the island of Tenerife two moderate earthquakes (I z VII) took place in 1706 and 1909 and a large one (I = X) on Lanzarote in 1790, associated to volcanic eruptions (Galbis, 1932, 1940). In Figure 1, the historical and instrumental seismicity are shown, as given by Monge (1980) and the Seismological Data File of the fnstituto Geogrifico National, Madrid (Mezcua and Martinez Solares, 1983). The largest concentration of epicenters are located on the island of Tenerife and between this island and that of Gran Canaria. Small events are located on the islands of La Palma and Lanzarote and disperse shocks are located in the area between Lanzarote Q 1992 - Elsevier Science Publishers B.V. All rights reserved 44x I \ I I \i J Ml /( and archipelago. islands the rest the m,, (Lg) magnitude earthquake took place at sea between the islands of Tenerife and Gran Canaria and was felt in most of the islands. Earthquake parameters are given in Table 1. In Figure 2, the epicenters of the main shock and well recorded aftershocks are shown. They present a clear lineament oriented NE-SW (N37”E). Since the seismic network is elongated in the E-W direction, poor control can be established in the N-S direction in which more or less the epicenters of the aftershock sequence are distributed. However, as clear onsets of P and S waves were used, for events with a larger number of readings (N r 6), enough control is present and we believe that the general trend may be considered valid. This trend agrees with that found also in the seismlcity ot the area between the islands of Tenerifc and Gran Canaria (Fig. 1). The main shock was followed by a sequence of aftershocks which lasted over a month with most shocks occurring in a few hours after the main event. The distribution of the number of aftcrshocks with magnitude and time has been determined obtaining a h value of 0.61 + O.lY (log N = u - bA4) and a p value of 1.40 A 0.16 [log N(t) = k -p logtt + cl] (Fig. 3). These values are representative of a tectonic origin of the sequence and cannot be associated with a volcanic origin (Magi, 1967). The focal mechanism of the main shock has been obtained using 37 polarities. The solution is shown in Figure 2 and its parameters given in Table 2. The solution corresponds to a mechanism with strike-slip and reverse faulting. P and T axes are nearly horizontal in NNW-SSE and E-W direction, respectively. The seismic moment obtained by spectral analysis of the Rayleigh waves as recorded in two standard stations (TOL =iLiizz enot assigned ’ (2.0 2.0s: I <3.0 1 a.o<: Q X4.0 4.0s: cl <s.rJ 5.0~5IJ <6.0 28-N 26’N 16 A st at ion II bw Fig. 1. Historical zyxwvutsrqponmlk Seiw 8pbic b 16 W and instrumental seismicity 14’W of the Canary Islands. (I.G.N. Seismological Data File). 449 RxaI parameters of the May 9, 1989 Catlarian earthquake Origin time CG.M.T.1 Latitude Longitude Depth Magnitude fMb r-81 02t? 3&n 37.4s 27” S6.8’N 16” 12”O’W 25 Km 5.2 Fig. 2. Epicenter and focal mechanism of the May 9, 1989 ea~bquakc and its afte~h~ks. difatations. 2 a.6 Intensity (max.1 V (M.S.K.) Dots are com~~~ssio~s white triangles are a zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA 6 aa *.a (I aa 6 r&&c&WE Fig, 3. Dist~~~ti#fl of cwmulat~vc nurn~~~ of afte~h~~ TtME as a function of magnitude and the activity constant. TABLE 2 Focal mechanism parameters of the May 9, 19X9 Canariao earthquake Seismic Axes T and P -I- moment (Nm) KG- Nl?X 4.cI x lflgh 4.1 x lo’* Nodal planes R i”: 23”i P: 4” * Qt - 13” 755”:~ 6” X” i4h” + 1.3” 18 A 11” 7i”&. Ii” If” wj- 1hV * 1J” -37 wt II” 6 A: H: 33”* ?w+ 7‘: hO” YI” A: 143”’ 67” P: Iti” 214” R: 371“ .w Y” Nunuta. --...- -I-.-LII. see,re --_--... 1) 7.x IIh” 351’ ^..._“” tr = plunge; CD= trend; C#P = strike; 8 = dip; A = slip. and MALI in the Spanish rnajn~a~d is 4. I x IO’” Nm. Moment tensor inversion solution for this earthquakes is given by PDE (Preliminary Determination of Epicenters U.S.G.S., NEXC, May 1989; Table 2). The solutions are quite different, although they agree in the trend of plane B (about 290’1 and in the nearly horizontal position of the P axis. We are in favour of our solution since the first-motion inconsistencies are few and it fits the distribution of aftershocks and the regional seismicity trend. Plane A of our soIution has been selected as the fault plane on the basis of its agreement with the alignment of aftershocks (Fig. 2). Seismotectonic interpretation As mentioned before, the level of seismicity of the area before this earthquake was so low and diffused that it was not possible to relate it t.o fractures deduced from geophysical and geological studies. However, the occurrence of this shock and several microseismicity surveys carried out in Tenerife during 1980, 1981 and 1982 (Mezcua et al., 19901 makes it possible to interpret seismic activity on seismotectonic grounds. Main fractures in the islands and ocean ffoor detected by geophysical, petrographic and geochemical methods may be cfassified in two types or families: Atlantic and African, depending on their relationship with the opening of the Atlantic or the tectonics of the Atlas range in the African continent (Anguita and t-fern&t, 1975: Fuster, 1975; Carracedo, 1984; Emery and Uchupi, 1984; Daiiobeitia, 19881. For the African family of fractures the orientations are ENE-WSW (coincident with the Hierro-Gomera and Tenerife island axis) and NNE-SSW (coincident with the Fuerteventura-Lanzarote axis); and for the Atlantic (the one marked by the line La PalmaGomera-Tenerife and Gran Canaria) WNWESE. This last alignment is called Atlantic, because it is oriented parallel to the transform faults of the mid-Atlantic ridge, in particular the Atlantic fracture zone (Rona, 19801. If such families of fractures are correlated with the seismicity presented in Figures I and 2, the most significant seismic trend present corresponds to the fracture of the African type detected by Bosshard and McFarlane (19701, located between Tenerife and Gran Canaria and considered responsible for the origin of the Canary Islands by Anguita and Hernan (19751. As described above, plane A of the focal mechanism of the main shock oriented NE-SW agrees with the distribution of aftershocks and marks the trend of the fat.& between Tenerife and Gran Canaria. According to this mechanism, we can conclude that the rn~~tion on this fault is left-lateral strike-slip with a reverse component of motion resulting in underthrusting of the west block (Tenerife). This type of motion present in the Fault of the African type between Tenerife and Cran Canaria corresponds to a pressure axis almost horizontal in NNW-SSE direction. This direction agrees with that found by Scheidegger (1978) for the predominant orientation of stresses derived from the observations of joints in Tenerife and Gomera. This coincidence may indicate a 1 2 :w SiW Fig. 4. ~eismotect~nic framework of the southern part of the horizontaI pressure pattern in NNW-SSE direction which on reverse faufts produces vertical movements with underthrusting of the west block and strike-slip motion with southwest relative displacement of Tenerife with respect to Gran Canaria and Africa. A simplified tectonic picture of the region is shown in Figure 4. The border between the Eurasian and African plates is located to the north of the archipefago 03ufor-n et al., 1990). The trans-Agadir-Nekor fautt system to the northeast separates a block including the high Atlas and possibly an area of oceanic crust (Buforn et al., 19901. A horizontal compressional stress regime with NW-SE orientations is present in the whofe region. The Canary islands are located south of this btock, and it is likely that the fracture in NE-SW direction may be related to this important fault system (Sanz de Galdeano, 199OXThe orientation of the fault between Tenerife and Gran Canaria agrees also with other features in the archipelago such as the alignment of volcanoes in Tenerife and Lanzarote and is parailei to other faults, in the same direction, c* Iberia-Africacontact. existing in the African continent such as the Zemmour fault (Anguita and Iiernin, 19751. The results presented above, in the absence of more focal mechanisms data and better definition of the seismic&y in the area, must be interpreted in a very provisional way. They point to the fact that main faults a presently active in the Canary archipelago are of the African type with orientation almost parallel to the Trans-Agadir and Zemmour faults, subject to NW-SE horizontal pressure, suggcstingtectonics in the isfands related to that of the African Continent. zyxwvutsrqponmlkjih Acknowiedgments This work has been supported in part by the Direction General de ~nvestigaci~n Cientifica y Tecnologica, project PB-89-0097. Publication 225, Sub. Gen. Astronomia y Geoffsica, Institute Geografico National, and Publication 333, Department of Geophysics, Universidad Complutense de Madrid. 452 References Galbis, J., 1940. Catalog0 entre Anguita, F. and model versus Earth Planet. Arana, Hernin, F., 1975. A propagating a hot spot origin fracture for the Canary E., Daiiobeitia, Buforn, Planet. del volcan- Suriiiach mechanisms E., Udias, M.A., and tectonics A. and J.. estudio del terremoto Is- rkplicas. Inst. Geogr. 1988. Seismicity, Monge, 18. y Rev. Geofis., the Western Gravity Canary D.J., 1970. Crustal Islands Data. J. Geophys. Carracedo, from Seismic structure Refraction of and Res. 75: 4901-4918. J.C., 1984. El origen de las islas. Geogr. turas J.J., C‘anario. Emery Canarias, situadas K.O. and Uchupi, Springer de estruc- y sur del Archipielago E., 1984. The Geology Verlag. y temporal of the At- de1 vulcanismo un ejemplo de evolucibn oceinico. Estud. Geol., entre J., 1932. Catilogo 10s meridianos Inst. Geogr. Sismico de la zona 5”E y 2o”W y paralelos Cat. Madrid, I, 807 pp. Solares J.M. and de las Islas Madrid, Canarias. Pub. Tee. 73. 24 pp. and fractures. Tectonophysics. comprendida 45”N y 25”N. del archipiilago Universidad Canario. Complutense, plumes 1980. The central continental Washington, 5: T&is Madrid. in the lower mantle. margins. North Atlantic N.O.A.A. Ocean Atlas basin 3. U.S.D.C.. DC. C., 1990. La prolongaci6n las fosas y desgarres del norte de interpretacibn. y centro hacia et Sur de de Europa: Una Rev. Sot. Geol. Esp. 3: 231- 241. Scheidegger, A.E., Schmincke, 1978. The tectonic Islands. H.U., regime Rock Mech. 1973. in the Canary, significance Magmatic Madeira of joints in 11: 69-85. evolution and Azores and tectonic Islands groups. Geol. Sot. Am. Bull., 84: 634-648. Surifiach, E., Canario. 31: 439-463. Galbis, P.A., and the Canary New York, NY, 1020 pp. J.M., 1975. Las Islas Canarias espacial geofisico al norte Rev. Sot. Geol. Esp. 1: 143-155. lantic Ocean. Fhster, 1988. Reconocimiento Del 2Y9 pp. 230: 42-43. propuesta submarinas 1983. Sismicidad Nat. Madrid, del Y de mayo de 1989 y su sPrie de W.J., 1971. Convenction Nature, Sanz de Galdeano, pp. 55-63. Daiiobeitia, Morgan, Rona, E. and Macfarlane, J.M., J., Martinez Nat., F., 1980. Sismicidad de Licenciatura. 1990. Sismicidad Ibero-Mogrebi. 45”N y 25% 35-55. of the Azores-Gibraltar J.. Solares, J., Rueda, Ansorge. comprrndid;l II, 277 pp. E., 1990. Sismotectcinica the Canary 46: 173-182. Bosshard, J.. Galin, de la zona Inst. Geogr. Mogi, K., 1967. Earthquakes 152: 89-l Mezcua, de la regi6n Cat. Madrid, Ibero-Mogrebi. Mezcua, Sci. Lett., 55: 1 l-24. Tectonophysics, sismotect6nica E. and under Sismico 5”E y 2o”W y paralelos J. and Martinez Buforn, structure A. and ColombBs, plate boundary. Buforn, J.J., of crustal E., Udias, source Mezcua, An. Fis., 82: 202-231. 1081. Features lands. Earth Islands. Area V. and Ortiz, R., 1986. Marco geodinimico Banda, Inst. Geogr. Sci. Lett., 27: 11-19. ismo canario. 10s meridianos sismicos Wilson, 1986. La estructura Resultados profundos. W.I., Tectonophysics, 1973. de la cortical interpretacibn de1 archipielago de perfiles An. Fis., 82: 62-77. Mantle plumes 19: 149- 164. and plate motions.