The Düzce segment of the North Anatolian Fault Zone (Turkey): Understanding its seismogenic behavior through earthquake geology,

Abstract

The area struck by the November, 12, 1999, Mw 7.1 earthquake that ruptured the Düzce segment of the North Anatolian Fault Zone (NAFZ) was investigated. In order to document the Düzce seismogenic fault characteristics, segment of the North Anatolian fault, systematic geological, geomorphological and paleoseismological analyses were integrated in this thesis. A detailed mapping and study of the 1999 earthquake coseismic ruptures and of the short- (Holocene) long-term (Pliocene-Pleistocene) tectonic landforms, first, in a key area, then, along the whole Düzce fault was carried out. The major objective were to compare the detailed coseismic surface expressions with the short/long-term morphology and structural architecture of the Düzce fault zone. This was accomplished to explore the persistency or evolution through time of the active fault setting, at the surface, that could highlight characteristics of the seismic source, at depth. Along the key area was possible to zoom in a scale-independent en-échelon arrangement of the coseismic surface ruptures and to evidence by the comparison with the short/long-term geomorphic expression of the Düzce Fault near the 1999 ruptures, that: 1) the principal slip zone at depth accommodates the bulk of the displacement during an individual rupture event and 2) may stay localized through many rupture episodes with persistent geometry and kinematics. At the same time, an old and complex fault arrangement has been mapped, partially coinciding with the 1999 rupturing fault, whose relationships with the coseismic fault systems suggest an evolution of the fault pattern trough time, with a tendency to simplify a geometric complexity into a straighter, mature trace. Along the whole area, also, the older complex fault system, which involves a wider zone of deformation, was identified and the structural pattern of the simple 1999 coseismic fault trace was analyzed at the different scales of observation. Overall, two different sections of the Düzce segment were recognized: a western section, where the coseismic fault trace has a staircase trajectory and reactivated part of the older fault system; an eastern section, where the coseismic fault trace shows a straight trajectory and cross-cuts the older and complex fault system. The Düzce fault sections may represent different stages of the segment evolutionary tendency towards a simpler mature trace, as a mechanically more favorable setting. The western section of the Düzce fault segment splays out from a restraining bend of the Izmit (Karadere) fault segment of the NAFZ, and forms a releasing fault wedge. By comparing the coseismic surface deformation field with the observed long-term morphology it is clear that the present landforms and setting are the result of 1999-type coseismic deformation repeating through several seismic cycles. Because of the mechanical interaction of the faults in the release junction, the western section of the Düzce fault undergoes a transtensional strain field that may justify and cause its complexities to be a steady state of the structural arrangement. The boundary at the surface between the two portions of the Düzce fault is not only a surface characteristic but it separates at depth a portion of fault plane characterized by a big single asperity, to the east, from a portion of plane with lower slip, to the west. Thus the peculiar arrangement of the Izmit (Karadere) and Düzce fault segments may permanently control the difference in behavior of the two portions of Düzce fault and furthermore control rupture propagation and fault loading. Under this light, the Izmit/Düzce release fault junction (1) may produce an unfavorable setting for the build up of asperities in the western part of the Düzce segment also in the future and (2) could have delayed the propagation of the 1999 August Izmit rupture on the Düzce segment that ruptured on November 1999 along the asperity of its eastern section. These results highlight that the surface geological data contain the potential for integrating and completing the information for imaging structures also at a seismogenic depth. The integrated investigation of short/long-term tectonic morphologies and structural pattern offers a noteworthy frame to interpret the coseismic rupture kinematics and clarifies their complexities. Moreover, for a full understanding of the principal slip zone at depth, this case study shows the importance to define the strain distribution pattern and evolution of surface rupturing faults. The geological and geomorphological map along the fault trace permitted to analyze the spectacular tectonically driven cumulative landforms and the drainage pattern settings, in order to provide new estimates on the Quaternary slip rate of this part of the active transform margin of North Anatolia. As offset geomorphic markers, right-hand stream deflections and remnant of an old alluvial fan modeled by fluvial terraces were reconstructed and described. The streams are deflected for a total of about 100 m and the onset of the offset was radiocarbon dated about 7000 yr BP. The two documented and correlated Late Pleistocene, terrace risers are offset of about 300 and 890 m, respectively. These terrace risers were dated by means of Optically Stimulated Luminescence (OSL) method about 21 000 yr BP and 60 000 yr BP. These ages and offsets translate to a constant rate of deformation of the Düzce Fault, at different time scales, of 14.0 ± 1.8 mm/yr and disproves a time-variable model at least for the last 60 000 yr. On this light, considering the GPS-measured strain accumulation due to the plate motion along this part of the North Anatolian Fault Zone, the Düzce Fault importantly participates to the North Anatolian margin deformation and assumes a relevant role in the seismic hazard of the area. To learn about recurrence of large earthquakes on this fault, paleoseismological trench investigations were undertaken. On the basis of sedimentary and structural relations observed in the trench walls, evidence for repeated surface faulting paleoearthquakes pre-dating the 1999 event were found. By merging information obtained from all the trenches it is possible to reconstruct the seismic history of the Düzce fault for the past millennium. Coeval events between different trench sites were correlated under the assumption that, similarly to the 1999 event, paleoearthquakes ruptured the whole Düzce fault. Besides the 1999 earthquake, prior surface faulting earthquakes are dated as follows: penultimate event, possibly at the end of 19th century; third event, possibly close to AD 1700; fourth event, AD 1185-1640; fifth event, possibly AD800-1000. According to the above results, the AD1719, AD1878 and AD 1894 historical earthquakes, may have ruptured the Düzce fault and not the faults they are usually associated to or, alternatively, a cascade of events occurred on the Düzce and nearby faults (similarly to the Izmit and Düzce 1999 earthquakes). On this basis can be inferred an average recurrence time of ~300 yrs for large surface faulting events on the Düzce fault. Moreover, assuming that the slip produced by the 1999 earthquake is characteristic, the Düzce fault presents a strain release model, with a not perfectly periodic interseismic interval, and an average strain accumulation of 13.3 mm/yr, comparable with the slip rate results obtained by the geomorphic marker analysis.