The death toll from Sunday's magnitude 7.2 earthquake in the Van Province in eastern Turkey has now risen to over 500 people, and will undoubtedly continue to rise as rescuers continue to search the hundreds of buildings that collapsed during the shaking. The tectonic forces ultimately responsible are quite straightforward to explain, but as is often the case, the picture becomes more complicated when we take a closer look - a fact that has consequences for the people caught up in this disaster.
The big tectonic picture
The most well-known seismic hazard in Turkey is probably the North Anatolian Fault, a major strike-slip fault which runs along the south of the Black Sea and underneath Istanbul, and has a long history of damaging earthquakes. The East Anatolian Fault, another strike slip fault that runs southwest from eastern Turkey to the Mediterranean, is another well-known source of large earthquakes.
But Sunday's earthquake occurred some way to the east of these structures, in a high region known as the East Anatolian Plateau that has also seen its fair share of large earthquakes in the past. The focal mechanism shows that the latest rupture was due to movement on a thrust fault, a response to the crust around it being compressed in a north-south direction. As the figure to the left illustrates, this region is on the boundary between two regions with very different tectonics. To the east, in Iran, there is a great deal of convergence and mountain building as the Arabian plate to the south moves northwards into the Eurasian plate, as part of a belt of mountain building that runs all the way from the Alps to the Himalayas. To the west is the strike-slip faulting concentrated on the North and East Anatolian Faults, which is, surprisingly, in response to the exact same continental collision. The bit of crust that makes up Turkey is apparently strong enough that rather than internally deforming in response to the Arabian plate encroaching on its territory, it is getting pushed westwards out of the way, with the motion largely occurring on the North Anatolian and East Anatolian Faults.
Caught in the midst of this fundamental shift in the way the crust is deforming, the geology and the geological history of the Van Region is very complicated. Rather than large throughgoing structures, it is a mish-mash of lots of smaller faults.
Geological evidence suggests that up until a few million years ago, thrust faulting was quite common in this area, but more recently strike-slip faulting has taken over. Recent small earthquakes in this region have definitely been largely strike-slip: in the figure below, most of the beachball-like focal mechanisms are split into four quarters, which tells us those earthquakes were on strike-slip faults.
However, this figure also tells us that seismometers have recorded a few compressional thrust fault ruptures (where the focal mechanisms are split into three slices, with a red/blue central slice), just like Sunday's event. And one of the areas that has experienced some of these thrust earthquakes is along the eastern shore of Lake Van. It appears that some peculiarities of the structure and local crustal stresses in this region mean that some thrust faulting is still happening here.
Complicated areas like this are very challenging for geologists to interpret, and make detailed assessments of seismic risk extremely difficult too. There is a lot of deformation that must take place in response to the encroaching plates to the north and south, and the escaping microplate to the west, so the seismic hazard is high; but it is also diffuse, being distributed over many faults that may in aggregate produce significant earthquakes every few decades, but individually may not rupture more than once every few thousand years. In contrast, in western Turkey the seismic risk is also high, but mainly focused on the North and East Anatolian faults, which take up the lion's share of the deformation required for Turkey's exit from the continental collision zone. Geologists might not be able to predict the 'when', but at least they have a better handle on the 'where'.
Damage and building codes
Of course, as always when an earthquake occurs in a region with lots of people living in it, the casualties are as much determined by how resilient the buildings in the area are as the size and location of the earthquake. In the wake of the magnitude 7.6 Izmet earthquake that ruptured a 150 kilometre section of the North Anatolian Fault in 1999, killing 20-30,000 people, the Turkish government has stepped up efforts to enforce building codes designed to stop things like catastrophic collapses of structurally weak multi-story buildings. However, lack of compliance with these regulations has long been an issue, and it is clear from images from the town of Ercis, which appears to have borne the brunt of the damage in the region, that many buildings have collapsed that perhaps would not have if they had been built up to spec.
Assessments from the earthquake zone of "...concrete thinned with gravel, insufficient steel girders and supports knocked out to create more space" certainly reinforce this impression: in a dramatic TV demonstration, one emergency worker apparently crumbled some concrete from a collapsed building to dust with his hand. However, since many of these buildings probably predate the last decade, it is hopefully more a symptom of past neglect, and the challenges of undoing that neglect when your entire country is earthquake prone, rather than an indictment of more recent practices. The complicated tectonic picture in this region discussed above may also have complicated matters by making it difficult to identify, and prioritise for strengthening, the towns and buildings most at risk in the area.
Aftershocks and future seismic risks
The global seismometer network has recorded almost 50 aftershocks of magnitude 4 or greater in the past 5 days, the largest of which was a magnitude 6.0 ten hours after the main shock on Sunday.
The aftershocks are already dying down - only 6 of the reported aftershocks occurred yesterday - but the stress changes in the region due to this earthquake may interact in complicated and hard-to-predict ways with other faults in the area, and may lead to a heightened chance of further large earthquakes in the months and years ahead. Unfortunately, we'll just have to wait and see.