Deformations associated with the El Asnam earthquake of October 10, 1980 (Algeria)

Abstract

The resurvey of both a geodetic network and a levelling net was carried out in June 1981, 8 months after the M s = 7.3 El Asnam earthquake of 10 October 1980. Previous seismological and neotectonic studies indicate that this event results from a north-east-trending overthrust complex fault of about 40-km length, which shows at least three principal segments with slightly different directions. Vertical movements, evaluated by means of a trigonometric levelling method show an uplift of the thrust fault of about 5 m and a depression of the southeastern edge of about 1 m. These movements are progressively attenuated away from the fault trace (see fig. 1). Horizontal movements have been evaluated by a classical first-order triangulation method. (see fig. 2 and table 1). The resulting mean strain tensors, calculated for different triangles of the geodetic network, indicate a shortening of about 2.50 m which is consistent with the SE-NW direction of compression determined from neotectonic evaluations and focal mechanisms (see fig. 3). Dislocation models are used to explain and discuss the observed deformations in the light of the seismological data and the observed ground breakages. Five segments are required to explain both horizontal and vertical deformations. The magnitude of vertical displacement (about 6 m) at the junction between the southwest and the central segments of the fault argues for the breaking of this area during the main shock and for a slip vector of about 8 m, at least in the central segment. In such circumstances where the first field observations show that a strong coseismic movement has taken place, the best methodology to be used seems to be: • - to resurvey as quickly as possible the widest acceptable zone of the old existing geodetic network, without special attention paid to the precision of these measurements. • - to set up as soon as possible a small aperture geodetic network of high accuracy in order to monitor the possible postseismic readjustments. For vertical movements, trigonometric leveling using reciprocal zenith measurements with two high-precision theodolites and EDM appears to be far the best solution. It is much quicker than classical leveling and it can afford the same precision if necessary. In our case, the precision was not critical, so that long ranges (up to 1500 m) between stations were used when necessary.