Late Quaternary evolution of the Çanakkale Strait region (Dardanelles, NW Turkey): implications of a major erosional event for the postglacial Mediterranean-Marmara Sea connection

Seismic and bathymetric data from the Çanakkale Strait and its extensions onto the shelves of the Marmara and Aegean seas indicate that the strait was formed mainly by an erosional event. Four seismic units are observed on seismic profiles. The lower two of these (units 4 and 3) constitute the basement of a regionally widespread erosional unconformity (ravinement), which developed during marine isotope stage 2 (MIS 2). The two upper units (units 2 and 1), which overlie the ravinement surface, form a higher-order sequence. Sequence stratigraphic analysis indicates that units 2 and 1 deposited as lowstand and highstand systems tracts respectively, since the end of MIS 2. The transgressive systems tract is represented by a major erosional event which occurred throughout the Çanakkale sill area when the Mediterranean-Marmara Sea connection and, hence, the Çanakkale Strait was formed. The existence of the erosive ?arköy Canyon along the shelf edge of the southern Marmara Sea demonstrates that the flow direction causing the erosion was from south to north, thus proving that it was produced by Mediterranean water flowing over the sill into the Marmara Sea basin.     

Strong ground motion during the largest aftershock (Mw=5.8) of the 1999 Izmit earthquake, Turkey

We present a preliminary study of strong ground motion during the largest aftershock (M_w 5.8) of the 1999 Izmit earthquake (M_w 7.4), Turkey, at 11:55 on 13 September 1999. The peak ground acceleration observed near the epicentre of this aftershock was in agreement with that predicted by standard empirical prediction equations. Its spectral source parameters of the largest aftershock are also typical for a M_w 5.8 earthquake. At greater epicentral distances, there is an order-of-magnitude in scatter in peak ground acceleration values for this aftershock, which is attributed to site effects. The presence of thick layers of low-velocity sediments caused significant amplification of S-waves in the Avcılar district of Istanbul, at frequencies of 1 Hz, explaining the observed concentration of damage there as a result of the Izmit mainshock.     

Location Performance and Detection Threshold of the Spanish National Seismic Network

Spain is a low-to-moderate seismicity area with relatively low seismic hazard. However, several strong shallow earthquakes have shaken the country causing casualties and extensive damage. Regional seismicity is monitored and surveyed by means of the Spanish National Seismic Network, maintenance and control of which are entrusted to the Instituto Geográfico Nacional. This array currently comprises 120 seismic stations distributed throughout Spanish territory (mainland and islands). Basically, we are interested in checking the noise conditions, reliability, and seismic detection capability of the Spanish network by analyzing the background noise level affecting the array stations, errors in hypocentral location, and detection threshold, which provides knowledge about network performance. It also enables testing of the suitability of the velocity model used in the routine process of earthquake location. To perform this study we use a method that relies on P and S wave travel times, which are computed by simulation of seismic rays from virtual seismic sources placed at the nodes of a regular grid covering the study area. Given the characteristics of the seismicity of Spain, we drew maps for M _L magnitudes 2.0, 2.5, and 3.0, at a focal depth of 10 km and a confidence level 95 %. The results relate to the number of stations involved in the hypocentral location process, how these stations are distributed spatially, and the uncertainties of focal data (errors in origin time, longitude, latitude, and depth). To assess the extent to which principal seismogenic areas are well monitored by the network, we estimated the average error in the location of a seismic source from the semiaxes of the ellipsoid of confidence by calculating the radius of the equivalent sphere. Finally, the detection threshold was determined as the magnitude of the smallest seismic event detected at least by four stations. The northwest of the peninsula, the Pyrenees, especially the westernmost segment, the Betic Cordillera, and Tenerife Island are the best-monitored zones. Origin time and focal depth are data that are far from being constrained by regional events. The two Iberian areas with moderate seismicity and the highest seismic hazard, the Pyrenees and Betic Cordillera, and the northwestern quadrant of the peninsula, are the areas wherein the focus of an earthquake is determined with an approximate error of 3 km. For M _L 2.5 and M _L 3.0 this error is common for almost the whole peninsula and the Canary Islands. In general, errors in epicenter latitude and longitude are small for near-surface earthquakes, increasing gradually as the depth increases, but remaining close to 5 km even at a depth of 60 km. The hypocentral depth seems to be well constrained to a depth of 40 km beneath the zones with the highest density of stations, with an error of less than 5 km. The M _L magnitude detection threshold of the network is approximately 2.0 for most of Spain and still less, almost 1.0, for the western sector of the Pyrenean region and the Canary Islands.