Final Activity Report Summary - SEISMIC CYCLES (Understanding the irregularity of seismic cycles: A case study in Turkey)
The Marie Curie Excellence Grant project focused on establishing the seismic history over several thousands of years of a main strike-slip fault system in Turkey to get deeper insights in the fault seismic behaviour. The targeted North Anatolian plate-boundary was particularly suitable because of the type of faulting, its rapid deformation rate (up to 24 mm/yr), its relative structural simplicity and its particularly simple seismic behaviour characterised by cascading sequences of M7 earthquakes. The team used a diverse array of complementary field techniques involving trenching across the fault combined with subsurface geophysics, dating of displaced geomorphic features and coring of lake sediments during three main field campaigns (summers 2006, 2007, 2008). During the project we worked on four different paleo-seismological sites and on six lakes covering the different fault segments.
The different paleo-seismological trenches contained records of ground-rupturing earthquakes of 3000-year record in the Destek Trench (eight earthquakes), 3000-year in the Gunulan Trench (six earthquakes), of 4000-year in Resadiye Trench (eight earthquakes), a relict record of seven earthquakes between 600 AD and 2000 BC in Elmacik site. The data obtained from this project was combined with all paleo-seismic studies published on the North Anatolian Fault to build a database of paleo-earthquakes. The database was analysed to answer some essentials questions about recurrence behaviour of large earthquakes and to critically test variability in recurrence intervals.
The statistical methods used show that:
(1) the western, central and eastern sections of the NAF have different recurrence intervals that we interpreted to be caused by changes in the fault-normal stress; and that
(2) the 20th century earthquake sequence that migrated along the NAF is not typical of this fault, but may only occur around every 1200 years.
The use of lake sediments to track past earthquakes is a new approach that provides a continuous sediment record that lacks in most of the commonly used paleo-seismic trenches. Continuous recording allows sediments disturbed by earthquakes to be dated, theoretically, with much greater accuracy than the layers disrupted by faulting in trenches, yielding great improvement in the chronology of past earthquakes.
In five shallow lakes along the North Anatolian Fault, the sedimentological study of short 1 m gravity cores allowed us to characterise the signature of the last sequence M7 of earthquake that occurred during the 20th century. The characteristics of the last earthquake sequence were used to identify earthquake disturbed sedimentary layers or seismites in 3 to 5 m long cores. The seismites were correlated to known historical earthquakes or events identified in nearby paleo-seismological trenches by extrapolating the sedimentation rate obtained from the study of short cores. However, in some cases, an independent reliable chronology of the lacustrine sequence could not be obtained using radiocarbon dating, because of gas emission along the fault.
The study of the large Hazar Lake on the East Anatolian Fault encompassed a seismic survey and a coring survey. The seismic survey revealed that a continuous fault crosses the lake linking the two faults located in the northeast and southwest ends of the lake, and allowed a better understanding of the seismic cycle on this fault segment. The 5 m long cores collected in the Hazar Lake contained a sedimentary record of about 4000 years with clear indication of earthquakes. The Hazar Lake is a 200 m deep lake very different from the 5 to 20 m deep shallow lakes along the North Anatolian Fault, and a reliable radiocarbon age model was achieved. So the long-term earthquake recurrence of the East Anatolian Fault has been constrained and compared it with the obtained record on the North Anatolian Fault.
The different paleo-seismological trenches contained records of ground-rupturing earthquakes of 3000-year record in the Destek Trench (eight earthquakes), 3000-year in the Gunulan Trench (six earthquakes), of 4000-year in Resadiye Trench (eight earthquakes), a relict record of seven earthquakes between 600 AD and 2000 BC in Elmacik site. The data obtained from this project was combined with all paleo-seismic studies published on the North Anatolian Fault to build a database of paleo-earthquakes. The database was analysed to answer some essentials questions about recurrence behaviour of large earthquakes and to critically test variability in recurrence intervals.
The statistical methods used show that:
(1) the western, central and eastern sections of the NAF have different recurrence intervals that we interpreted to be caused by changes in the fault-normal stress; and that
(2) the 20th century earthquake sequence that migrated along the NAF is not typical of this fault, but may only occur around every 1200 years.
The use of lake sediments to track past earthquakes is a new approach that provides a continuous sediment record that lacks in most of the commonly used paleo-seismic trenches. Continuous recording allows sediments disturbed by earthquakes to be dated, theoretically, with much greater accuracy than the layers disrupted by faulting in trenches, yielding great improvement in the chronology of past earthquakes.
In five shallow lakes along the North Anatolian Fault, the sedimentological study of short 1 m gravity cores allowed us to characterise the signature of the last sequence M7 of earthquake that occurred during the 20th century. The characteristics of the last earthquake sequence were used to identify earthquake disturbed sedimentary layers or seismites in 3 to 5 m long cores. The seismites were correlated to known historical earthquakes or events identified in nearby paleo-seismological trenches by extrapolating the sedimentation rate obtained from the study of short cores. However, in some cases, an independent reliable chronology of the lacustrine sequence could not be obtained using radiocarbon dating, because of gas emission along the fault.
The study of the large Hazar Lake on the East Anatolian Fault encompassed a seismic survey and a coring survey. The seismic survey revealed that a continuous fault crosses the lake linking the two faults located in the northeast and southwest ends of the lake, and allowed a better understanding of the seismic cycle on this fault segment. The 5 m long cores collected in the Hazar Lake contained a sedimentary record of about 4000 years with clear indication of earthquakes. The Hazar Lake is a 200 m deep lake very different from the 5 to 20 m deep shallow lakes along the North Anatolian Fault, and a reliable radiocarbon age model was achieved. So the long-term earthquake recurrence of the East Anatolian Fault has been constrained and compared it with the obtained record on the North Anatolian Fault.