Seismic behaviour of the Dead Sea fault along Araba valley, Jordan

The Dead Sea fault zone is a major left‐lateral strike‐slip fault. South of the Dead Sea basin, the Wadi Araba fault extends over 160 km to the Gulf of Aqaba. The Dead Sea fault zone is known to have produced several relatively large historical earthquakes. However, the historical events are unequally distributed along the fault and only four events have been reported in the Araba valley over the last few thousands of years. Magnitudes estimated from the historical record are probably slightly smaller than that’of the Mw  ∼ 7.3 earthquake that struck the Gulf of Aqaba in 1995. The fault cuts straight across Pleistocene to Holocene alluvium and shows morphologic evidence for essentially pure strike‐slip motion. Regional seismic monitoring reveals little microseismicity along the fault except around the Dead Sea and Gulf of Aqaba, where the fault splays into complex pull‐apart basin fault systems. We have investigated the fault zone at several sites selected from SPOT images and the study of aerial photography. At the site of the now destroyed Tilah Castle, a well‐preserved wall, dated to be about 1200 yr BP ( 14 C age on charcoal), is cut by the fault and offset by 2.2 m. Comparison with offset gullies at a nearby site 3 km to the north and at three other sites, respectively 25, 50 and 65 km to the south, reveals that this specific fault displacement is probably related to the last seismic event that ruptured that fault segment, possibly in AD 1458. Moreover, the offset gullies suggest a characteristic slip behaviour with recurring slip of about 1.5 m on average. Given the 4 ± 2 mm yr −1 slip rate derived for this fault segment, we infer that the fault should produce Mw  ∼ 7 earthquakes along some segment in the Araba valley about every 200 years. The historical period, with only four well‐documented large earthquakes in AD 1068, AD 1212, AD 1293 and AD 1458, thus appears to have been relatively quiescent, with a 20 per cent deficit of Mw  ∼ 7 earthquakes. However, our data do not exclude the possibility of larger Mw  ∼ 7.6 earthquakes or time clustering of earthquakes over longer timespans. An alternative seismic behaviour involves Mw ∼  7.6 earthquakes about every 6000 years and Mw ∼  7 earthquakes about every 250 years. The historical catalogue would then appear to be complete for Mw  ∼ 7 earthquakes.