No one can be sanguine that any array will capture its intended information immediately after being established. Forecasting of earthquakes is rife with much uncertainty, and an early maxim in strong motion seismology stated that earthquakes tended to cease occurring in areas that had been instrumented to record them! In spite of this seeming initial setback many arrays have been created in many different countries, and useful data recorded for the benefit of the world community of scientists and engineers. There must be a reasonable expectation that an earthquake can be recorded within a meaningful length of time, and that the records will be representative of the seismic environment that produced them.

Some current research indicates that there exist plausible grounds to expect heightened odds for a major earthquake to affect Istanbul, the largest city in Turkey, and the principal seat of the country’s financial institutions.14 This forecast considers earthquakes on the North Anatolian fault system in the Sea of Marmara during the past 5000 years, and tests the resulting catalog against the frequency of damage in the city during the preceding millennium. If the time-dependent effect of stress transferred by the 1999 Mw = 7.4 Izmit earthquake to faults nearer to the city are considered, then there appears to be a 32 percent probability of strong shaking to occur within 50 km from Istanbul during the next decade. The ground rupture during the 17 August 1999 earthquake appears to have terminated beneath the Marmara Sea to the east of Yalova, thus increasing the stress and the likelihood of a triggering there. Yalova is 40 km to the south-east of Istanbul. There are no seismic faults within the immediate vicinity of Istanbul. It is therefore prudent to set up the first array in the area of Yalova, containing the urban boundaries from east to west, but extending beyond them transversely and longitudinally.