In a year that has so far seen a below-average seven M7+ earthquakes globally, the Mw 7.3 earthquake that struck the mountainous Iran-Iraq border region late evening local time on Sunday November 12, ranks as the most destructive and deadly earthquake of 2017. It was the eighth strongest to hit the region since 1900, and the strongest to hit within 150 miles (250 kilometers) in the last century. The epicenter was located 20 miles (32 kilometers) south of the city of Halabja in Iraq. Reports state that over 500 people were killed and over 8,000 injured, according to the United States Geological Survey (USGS).
In Iraq, at least nine people were killed and 535 injured, with the most severe shaking affecting a smaller, more sparsely populated area. The worst-affected settlements are reported to be the city of Khanaqin and the town of Darbandikhan, both in Sulaymaniyah Governate.
Iran suffered the biggest impact — Iran’s National Disaster Management Organization (NDMO) stated that 123 settlements have been damaged. An estimated 11,000 rural dwellings were destroyed, alongside thousands of homes in taller multifamily dwellings in the town and cities. The total estimate of damaged buildings ranges up to 30,000 and leaves tens of thousands of people homeless.
Many towns and cities in the Iranian province of Kermanshah, close to the Iraq border, were affected. The town of Sarpol-e Zahab, with a population of 45,000 and some 20 miles (32 kilometers) from the border saw about 90 percent of buildings destroyed. The hospital, which is Kermanshah Province’s primary hospital, is reported to be severely damaged. More than 140 fatalities are reported in this town alone.
Across Kermanshah Province, damage was also reported in the province capital Kermanshah with a population of around 850,000, and ranged from Eslamabad-e-Gharb in the south of the province and Paveh in the north. The earthquake affected the city of Ilam in Ilam Province south of Kermanshah, and Khorramabad and Borujerd in Lorestan Province, which are about 170 miles (280 kilometers) from the epicenter. In Borujerd alone, 354 buildings are reported damaged.
A “Slip Over a Larger Fault Area”
The Zagros Mountains, spanning from southeastern Turkey and through the Iranian western border with Iraq, contains the 1,100 mile-long (1,800 kilometer) Zagros fold and thrust belt which marks the boundary between the Arabia and Eurasia plates, and is a known zone of active crustal deformation and seismicity.
This earthquake happened due to oblique-thrust faulting where the plates converge. The size and mechanism of the quake mean that it is better thought of as a “slip over a larger fault area” than a point event, according to the USGS. Events of this magnitude may have a length of about 40 miles (65 kilometers), and damage will therefore occur at some distance from the designated epicenter.
This region is sparsely populated, otherwise the number of deaths may have been in the tens of thousands, similar to the 2003 earthquake that hit the ancient city of Bam in southeast Iran.
Although the earthquake risk is well recognized, there is not enough consideration of the impacts that large earthquakes could have on the buildings in this region, which is one reason the event lead to so many collapsed buildings and deaths.
Damage has been particularly severe to state-subsidized buildings constructed under the Mehr housing program. The program, started by President Ahmadinejad in 2007, had the aim to build two million affordable homes within five years. By 2014, some 1.18 million residences had been built, with developers offering free government land to build affordable housing units for first-time owners. Reports suggest that 90 percent of the Mehr housing units in Kermanshah Province were damaged. It is unclear whether construction regulations were followed, and many of the failed or partially destroyed apartment blocks appear to be made with unreinforced masonry infill walls that have crumpled.
Improving Risk Insight; Improving Resilience
RMS is currently developing earthquake models for this region to aid global reinsurers in estimating the earthquake risk, and to help development of reinsurance programs to cover such devastating events. These models need to evaluate the potential correlation of risk posed by “slip over a larger fault area” leading to correlated damages along the strike orientation of the fault plane. If earthquakes are modeled based on epicenters or point events, footprints fail to capture the correlation appropriately.
The RMS approach of using fault rupture planes to release seismic energy allow appropriate calculation of the spatial correlations of hazard and loss, which is why in our event descriptions we do not consider earthquake epicenters.
Similar to many other earthquake-prone regions, the core principles of building resilience need to be robustly pursued, such as risk avoidance by stricter building codes; risk control through education programs to reduce potential impacts; risk acceptance by budgeting for costs of future events, and potential risk transfer by the development of insurance programs. As we build new models of earthquake risk for the region we will quantify the seismic risk and the potential losses any future events may have, and we hope these enable better organization to develop risk strategies to build a more resilient future.