You are here: Home » Research & Monitoring » EQ Geology & Paleoseismology » 1999 Izmit, Turkey Earthquake » August 25, 1999

August 25, 1999

Summary

So far about 110 km of fault rupture has been found with fault slips mostly around 2.5 to 3 m with the maximum so far of 4.9m. Several interesting features have been found and many questions still remain. A map of the USGS observations shows a complex pattern of faulting. Some of the most interesting features include a 5 km stepover in the fault at Sapanca Lake. This reminds us of the 5 km stepover that separates the Hayward and Rodger's Creek faults under San Pablo Bay in the eastern San Francisco Bay area. The faulting also moves from the North Anatolian fault through a broad zone of deformation onto the Druzce fault near Akyazi. This is similar to the intersection of the San Andreas and San Gregorio faults offshore of the Golden Gate. Understanding how the Izmit earthquake linked together these various fault segments may provide a new understanding of the range of possible earthquakes that can occur near San Francisco and will ultimately help us more accurately estimate hazards in many regions. Another question arising from these observations involve the location of aftershocks well west of Golcuk while the rupture goes offshore there on faults that Akyut Barka thinks may end before reaching the area of these aftershocks. How are these aftershocks related to the mainshock and the faulting?

Field Report

We have spot-checked the length of the fault and believe the rupture to be approximately 110km (rough estimate because we cannot get to the west end due to road closures and the small possibility that the rupture continues farther to the east of where we have observed it). So far, it looks as if the rupture is only between Golcuk to the west and Eften Lake (called Melengolu on the 1:250000 topo map), to the east just southwest of Duzce.

We are concentrating on measuring offsets and so far our largest offset is 4.9m at roughly N40 42.569', E030 22.050' (just east of Sapanca Lake). There continues to be large offsets of up to 4.6m (+/-0.1m) for approximately 2km to the east of the 4.9m offset. Near the far eastern extent of the rupture the offset is more on the order of 1.2-1.5m.

We have established an alignment array at roughly the lat./long. listed above.

Damage is reportedly quite bad in Golcuk, but we have not been there. The worst area damaged that we have seen is at the far eastern extent to the west of Eften Lake. People are camping in tents to protect themselves against building collapse from aftershocks and also because many are homeless. Disease has begun to spread and a large rain storm yesterday did not help.

Attached should be 8 photos taken 8/22/99 and 8/23/99 from various points along the North Anatolian fault. We tried to send the photos at a decent resolution, but it would have taken over 45 minutes, so please bear with the low-resolution for now.

USGS Damage Photos from Izmit Earthquake, USGS

Tom Fumal, USGS

  • Line of metal poles offset 2.2 m.

    This picture shows a line of metal poles offset 2.2m adjacent to the east side of Izmit Bay.

  • Offset and moletrack.

    This picture shows an offset 3m and is in a large field with a well expressed moletrack. A moletrack is a descriptive term for how strike slip faults often look when seen at the surface.

  • Fault through a school.

    This picture is on the far east end of the rupture. The fault went directly through the school - the chalkboard reads something like: "We're all happy because school is out".

  • Flattened apartment buildings.

    This is where 6 large apartments buildings all were completely flattened. Luckily they were uninhabitated. Residents were supposed to have moved in during June, but were delayed. The fault is seen in the foreground. Aykut Barka has a photo taken of this area from a helicopter.

  • Slumping near Izmit Bay.

    This is adjacent to Izmit Bay where the fault leaves the Bay on its eastern side. There is a local slope down to the Bay that allowed slumping along the cracks to form the small scarps apparent.

  • Moletrack and sag pond.

    This shows another well expressed moletrack and marshy area around a sag pond. Sag ponds are another common feature in strike slip fault zones which occur where the fault is not straight.

  • Northern strand of a 100-m-wide graben.

    This is actually the northern strand of a 100-m-wide graben. A graben is an area bounded by multiple faults that has dropped down.

  • Moletrack near Izmit.

    Izmit moletrack is near the town of Izmit.