About SCEC Research Resources Learn & Prepare

Turkey, Izmit Earthquake

SCEC Geologic Investigations of the August 17, 1999 Izmit, Turkey Earthquake


The Mw 7.4 Izmit, Turkey earthquake produced about 110 kilometers of surface rupture, with up to 5 m of dextral slip, along a western portion of the North Anatolian fault zone (NAFZ). The August 17, 1999 earthquake is the latest in a sequence of large events that have occurred this century, and which have ruptured a nearly 1000-km-long section of the NAFZ. This sequence began with the 1939 Erzincan earthquake, followed by a generally westward progression of earthquakes in 1942, 1943, 1944, 1951, 1957, 1967, and finally the 1999 Izmit earthquake. This apparent progression of earthquakes has led many scientists to use the NAFZ as a natural laboratory in order to better understand fault interaction and behavior, and thereby apply lessons learned there to other high slip-rate strike-slip faults, such as the San Andreas fault in California.

The similarities between the NAFZ and the San Andreas fault has lead to several collaborations between SCEC scientists and scientists in Turkey. Tom Rockwell has been working with Aykut Barka and Serdar Akyuz of ITU for the past 4 years. Along with Tim Dawson (also of SDSU), they have completed initial paleoseismic studies along the 1912 rupture on the Saros segment in the Galipoli peninsula (1999, in review) and recently began paleoseismic investigations in the Izmit to Sapanca area, the site of the 1999 rupture. One of the purposes of this new study was to assess the possibility of using an Ottoman period canal berm as a piercing point for offset caused by the A.D. 1754 earthquake. The August 17, 1999 earthquake ruptured through this area, producing between 3-4 meters of offset across the fault.

For the past two years, James Dolan and his doctoral student Ross Hartleb (both at USC), together with Barka and Akyuz of ITU, have also conducted paleoseismic studies on the NAF farther to the east along the 1939, 1943, and 1944 ruptures. Their work has produced tantalizing new evidence of the past earthquake history on those important strands that will hopefully elucidate our understanding on the repeatability of large surface rupturing earthquakes along the NAF. Specifically, their preliminary data suggest that short-lived earthquake sequences such as the1939-1999 sequence may be typical of strain release along this important fault zone.


SCEC Response to the 1999 Earthquake

Within 24 hours of the earthquake, discussions began to organize a SCEC response team to assist our Turkish colleagues with the mapping of the surface rupture. On August 20, 1999 the SCEC Group C response team, consisting of James Dolan, Rob Langridge, Tim Dawson, Allan Tucker, and Sheri Christofferson left southern California and arrived in Istanbul where they met with Ross Hartleb (already working in Turkey) and the team sent by the USGS Earthquake Hazards group. SCEC Group C leader Tom Rockwell followed a week later after returning from his honeymoon in French Polynesia.

The following morning the SCEC and USGS teams met with Aykut Barka, the coordinator of the mapping effort from Istanbul Technical University (ITU). It was decided that the best way to proceed was to spend the first 2-3 days on a reconnaissance mission in order to assess the general characteristics of the rupture, including length and magnitude of slip. This was to be followed by splitting into mapping teams, which would start the detailed mapping of the rupture and collection of slip measurements along the fault.

Izmit Photos

Mapping Effort

By August 22, the eastward limit of the surface rupture had been defined as ending in the Eften Golu area, just east of Golyaka. On August 24, a joint meeting between the ITU, Institut du Physique du Globe de Paris (IPGP), SCEC, and USGS groups was held in order to delegate mapping responsibilities. The IPGP group was given responsibility for mapping the western reach of the fault, including the Gölcük area. The ITU group mapped between Izmit and Lake Sapanca, whereas the SCEC/USGS group focussed on the area from Lake Sapanca eastward to the eastern terminus of the rupture near Gölyaka, with the SCEC group paying particular attention to the fault segment east of Akyazi.

The logistics of the mapping effort proved daunting as torrential rain, a lack of aerial photography and outdated maps (replete with the names of towns that were changed decades ago) slowed the effort to map the rupture. Indispensable to the mapping effort were our Turkish colleagues, who also served as translators, and handheld GPS units, which helped to locate the team members on 1:25,000 scale topographic maps. Particularly important to the mapping effort were the local inhabitants, who were happy to tell us where the rupture was, often over a cup of tea. All of the team members were greatly impressed with the knowledge of the location and amount of slip along the surface by the locals. Without their help, many of the details of the surface rupture would have been overlooked.

Initial Observations

The mapping groups have defined the surface extent of the rupture as extending from at least Gölcük to the Eften Golu area, just east of Gölyaki. Three primary segments have been identified. The east­west trending western segment extends nearly 40 km from Gölcük to Lake Sapanca. A wall on the west side of the military base in Gölcük is offset between 4 and 5 meters, as are the adjacent two streets to the west. East of Izmit Bay, slip measurements cluster around 2.5-3.5 meters, although there is a suggestion in some areas of distributed slip. In the area of the Ottoman canal, built in 1591 just west of Lake Sapanca, offset tree lines indicate about 3.5 m of dextral slip, consistent with much of this section of the rupture.

East of the ~3-km-wide right step-over at Lake Sapanca, the rupture continues with high slip along the E-ESE trending Sakarya segment, which extends some 26 km from Lake Sapanca to the town of Akyazi. Dextral slip reaches a maximum of 5.10 ± 0.25 m near Arifiye, with many measurements clustering around 4-4.5 meters. Slip values gradually decrease to ~2 m over a distance of 16 km near Kazanci, as measured eastward from the area of high slip. The rupture then steps left a couple of kilometers and continues with up to 2.5 m of slip, which then decreases to zero northeast of the town of Akyazi.

East of the town of Akyazi, the E-NE trending Düzce segment extends the 1999 rupture an additional 35 km, albeit with considerably less slip. We measured a maximum of 1.5 m of dextral slip along the central portion of the rupture, although the average slip is more on the order of 1 m. This segment dies out eastward towards the city of Gölyaka in the vicinity of Eften Golu. The overall mapped length of the 1999 rupture, along with the measured amounts of strike slip, are consistent with initial estimates of moment.

Izmit Photos



USGS Earthquake Hazards Program Report


Created in the SCEC system
© 2014 Southern California Earthquake Center @
Privacy Policy and Accessibility Policy