The SAR track image was plotted on a simplified 10 m interval contoured bathymetric map. The E-W trending SAR lines follow the deepest scarps of this basin.

The lowermost scarps are affected by N130° to N140° active faults. Although minor deformation is observable on the central N-S trending track, an E-W sparker profile (not shown here) reveals clearly vertical fault planes in the central part of the  abyssal plain (see section 63 in plate 10).

Seismic reflection data were collected by conventional means as sketched in Fig.1. The data were acquired over 1850 km total length of profiles. The energy source and receiver were placed near the sea surface. The source is a sparker towed almost at the surface of the sea to minimize the ghost generation. However, the signal-to-noise ratio in the collected data was degraded when the sea and weather conditions were not good. The data were recorded in digital form with 0.5 ms sampling interval. Most of the profiles were recorded with 5 s shot interval and 3 s record length while sailing with a speed of about 5 knots. Howerver, some of the surface sparker data were recorded simultaneously with the Pasisar seismic data with a sailing speed of about 2 knots ; therefore, those surface sparker line recorded simultaneously with Pasisar lines have about 2.5 times more shots per unit distance than the surface sparker records that were shot alone. The sections s79, s86, s108, s111, s114 and s120 in this poster were recorded simultaneously with the Pasisar lines. Conventional seismic data processing methods were applied to the data as summarized in Fig. 2. Sections displayed in this poster show typical active structures within the basins of the sea of Marmara. The vertical exaggeration is about 23. The numbers on top of the sections are the shot numbers.

Pasisar seismic data were collected on profiles with total length of 700 km. The seismic sections in this poster are the results of unconventional data collection and processing methods developed by Ifremer. In the conventional seismic, the source and the receiver are towed near the sea surface. In this experiment, the receiver is towed at about 75 meters above the sea bottom. This way of data collection geometry introduces some advantages over the conventional seismic data collection at sea. In these data, the attenuation of the higher frequencies of the seismic signal is relatively less since the water column is crossed only once. This results in data with higher resolution both in vertical and horizontal direction. Furthermore, diffractions and out-of-plane reflections are less due to the geometrical considerations of the wave propagation. Therefore the collected data is more useful for active tectonic interpretation with respect to its surface collected counterpart. However, this type of reflection seismic data must be corrected for geometrical considerations before interpretation. Notice in Fig. 1 above that the trajectory of the reflected points on the interfaces follows a nonlinear path. Another difficulty arises due to the fact that the receiver has to be adjusted with respect to the bathymetry during operations. This causes the distance between the receiver and the shot continuously change during recording. Therefore the data must be unconventionally processed as shown in Fig. 2 for correction of the geometrical considerations as well as the signal quality and display. Sometimes the change of the receiver with respect to the sea bottom may be erratic, in this case the correction may not be well accomplished. This may result in artificial structures to appear in the section. Each section must be carefully checked against its surface collected counterpart where there are erratic changes in the depth of the receiver. Those anomalies marked with exclamation (!) signs in the sections are examples of such artificial anomalies.The vertical exaggeration is about 20. The numbers on top of the sections are the trace sequential numbers. Refer to the Atlas booklet for interpretational guide lines.