Albania earthquake strikes highest-hazard zone in the Balkans, devastating nearby towns

by Ross S. Stein, Ph.D., and Volkan Sevilgen, M.Sc., Temblor, Inc.
 
As of today, some 24 people are dead and 650 injured, and many more are unaccounted for in collapsed concrete buildings on the coastal plains surrounding the epicenter. The quake likely struck on a ‘blind thrust fault’ that does not reach the Earth’s surface but had nevertheless been previously identified by geologists.
 
CITATION: Ross S. Stein and Volkan Sevilgen (2019), Albania earthquake strikes highest-hazard zone in the Balkans, devastating nearby towns, Temblor, http://doi.org/10.32858/temblor.057
 
It could have been worse

On 26 Jul 1963, some 165 km (100 mi) east of the 26 Nov 2019 Mag. 6.4 Albania shock, a Mag. 6.1 earthquake struck Skopje, capital of Northern Macedonia destroying 80% of the city, killing over 1,070 people, injuring 3,000-4,000, and leaving more than 200,000 people homeless.

It would appear that the 26 Nov 2019 quake, although 5 times larger than the Skopje event, did much less damage, perhaps because it struck more than 35 km from the 375,000 people living in the Albanian capital of Tirana.

 

Map showing Temblor’s globally consistent PUSH (Probabilistic Uniform Seismic Hazard) model, together with earthquakes principally from the EMSC catalog. The region is one in which residents should expect powerful shaking in their lifetimes. The aftershock distribution is roughly perpendicular to the presumed NW-SE oriented thrust fault.
Map showing Temblor’s globally consistent PUSH (Probabilistic Uniform Seismic Hazard) model, together with earthquakes principally from the EMSC catalog. The region is one in which residents should expect powerful shaking in their lifetimes. The aftershock distribution is roughly perpendicular to the presumed NW-SE oriented thrust fault.

 
Immediate foreshocks and remote aftershocks

The M 6.4 quake was preceded during the previous 6 hours by four M3 shocks in the epicentral region, the largest of which was Mag. 4.4, striking 1 hour before the mainshock. While uncommon, this kind of activity does not make earthquake prediction any easier, because the foreshocks do not show any features that would distinguish them from typical shocks. Perhaps more surprising, there was a burst of seismicity 230 km (140 mi) to the northwest of the M 6.4 a little over 6 hours later near the city of Mostar in Bosnia and Herzegovina, capped by a Mag 5.4 shock. This might be a coincidence, but if these two distant events are indeed related, it would have to be by the stresses carried by the seismic waves of the M 6.4, which dissipate within several minutes.
 

Could the Mag. 5.4 shock be a remotely triggered aftershock of the M 6.4? There is likely no smoking gun, but the possibility is nevertheless tantalizing.
Could the Mag. 5.4 shock be a remotely triggered aftershock of the M 6.4? There is likely no smoking gun, but the possibility is nevertheless tantalizing.

 

A long history of large earthquakes

The high earthquake hazard of coastal Albania stems from tectonic compression of the crust that extends from Croatia south to Greece. The compression is evident in the contraction of the Earth’s surface measured by GPS, and by the long history of large earthquakes in the region. In fact, the southern Balkans are more seismically active than Italy.

 

Figure from Sevilgen et al. 2014 showing the ‘EMEC’ instrumental and historical earthquake catalog (Grünthal and Wahlström, 2012). The earthquake catalog extends to AD 1000; however, it is incomplete until about the past century, meaning that not all M≥5.5 quakes were recorded due to lack of seismometer coverage. But even when restricted to that time period (red shocks), the site of the 25 Nov 2019 quake is among the most active anywhere in the Balkans or Italy.
Figure from Sevilgen et al. 2014 showing the ‘EMEC’ instrumental and historical earthquake catalog (Grünthal and Wahlström, 2012). The earthquake catalog extends to AD 1000; however, it is incomplete until about the past century, meaning that not all M≥5.5 quakes were recorded due to lack of seismometer coverage. But even when restricted to that time period (red shocks), the site of the 25 Nov 2019 quake is among the most active anywhere in the Balkans or Italy.

 
Blind thrust fault a likely culprit

The contraction has produced a series of thrust faults, only some of which come to the surface. Those that do not cut the Earth’s surface instead fold the overlying strata. Croatia’s coastal islands are one example of these folds, and one of these folds lies at the epicenter of the 26 Nov 2019 quake.

 

Koci et al. (2011) and Kastelic et al. (2016) identified active faults in the region for the European SHARE project using satellite imagery and field mapping (with inferred fault slip rates in black numerals), and they subsequently remapped the region (revised fault slip rates in red numerals; Sevilgen et al., 2014). One can see that thrust faults near the 26 Nov 2019 epicenter are about 75 km long with slip rates of ~1 mm/yr. That would mean that M~6.4 quakes on these faults would have mean inter-event times of roughly 500-1,000 years.
Koci et al. (2011) and Kastelic et al. (2016) identified active faults in the region for the European SHARE project using satellite imagery and field mapping (with inferred fault slip rates in black numerals), and they subsequently remapped the region (revised fault slip rates in red numerals; Sevilgen et al., 2014). One can see that thrust faults near the 26 Nov 2019 epicenter are about 75 km long with slip rates of ~1 mm/yr. That would mean that M~6.4 quakes on these faults would have mean inter-event times of roughly 500-1,000 years.

 
Amplified shaking on the coastal plains

The shaking produced by the Mag. 6.4 shock was almost certainly amplified in the weak, unconsolidated basins and coastal estuaries surrounding the epicenter. Temblor’s STAMP model shows amplification factors of 4-5 over the shaking that was experienced at bedrock sites, such as at the epicenter itself. Compounding the weak soil are water-saturated coastal plains, which are susceptible to liquefaction. This is where the soil turns to quicksand, causing buildings to sink or tilt. Sandblows and eruptions of artesian water often accompany shaking in such regions, compounding the damage.
 

Temblor’s STAMP high resolution (200 m) model of site amplification reveals that in Thumanë, Durres, and Lezhe, the shaking could have been severely amplified, contributing to the damage of weak buildings. Areas in black likely shook four times higher than those in yellow. The black areas are water-saturated coastal estuaries and plains that might also liquify when shaken violently, which can cause buildings to sink and tilt, rendering them a total loss. The fault on which the Mag. 6.4 quake struck is probably concealed by a growing fold.
Temblor’s STAMP high resolution (200 m) model of site amplification reveals that in Thumanë, Durres, and Lezhe, the shaking could have been severely amplified, contributing to the damage of weak buildings. Areas in black likely shook four times higher than those in yellow. The black areas are water-saturated coastal estuaries and plains that might also liquify when shaken violently, which can cause buildings to sink and tilt, rendering them a total loss. The fault on which the Mag. 6.4 quake struck is probably concealed by a growing fold.

Undoubtedly, emergency responders, geologists, and seismologists will learn more about this quake in the days to come. But what we can say now is that the hazard was not unforeseen.

 

Acknowledgements. We gratefully acknowledge funding from the Office of Foreign Disaster Assistance of USAID for the ‘Balkans-OQ’ project, a collaboration of eight Balkans scientists, and for assistance from the GEM Foundation.

References

Grünthal, G. and Wahlström, R. (2012): The European‐Mediterranean Earthquake Catalogue (EMEC) for the last millennium, Journal of Seismology, 16, 3, 535‐570, doi: 10.1007/s10950‐012‐9302‐y

Vanja Kastelic, Michele M.C. Carafa, and Francesco Visini (2016), Neotectonic deformation models for probabilistic seismic hazard: a study in the External Dinarides, Geophysical Journal International, 205, 1694–1709, https://doi.org/10.1093/gji/ggw106

R. Koci, N. Kuka, V. Kuk, K. Kuk, J. Mihaljevic, B. Glavatovic, H. Hrvatovic, I. Brlek, S. Kovacevic, S. Radovanovic, Z. Milutinovic, R. Salic (2011), Seismotectonic model as the OHAZ input, in Harmonization of Seismic Hazard Maps for the Western Balkan Countries, Closing Meeting Zagreb, 12-13 May 2011 (mihaljevic@seismo.co.me)

Volkan Sevilgen , R.A. Bennett , I. Brlek , Laurentiu Danciu , V. Kastelic , S. Kovacevic , C. Kreemer , K., Kuk , N. Kuka , Z. Milutinovic , S. Mustafa , B. Sket-Motnikar , Ross S. Stein , and L. Vucic (2014), BALKANS-OQ: A collaborative seismic hazard assessment of the Balkan countries using the OpenQuake software and the GEM Strain Rate Model, Second European Conference on Earthquake Engineering and Seismology, Istanbul, http://www.eaee.org/Media/Default/2ECCES/2ecces_esc/3276.pdf