A new analysis of a fatal landslide that occurred on 13 February 2024 at theÇöpler Gold Mine in Türkiye reveals that the site of the landslide had been slowly moving for at least four years prior to the failure.
"Additionally, our analyses detected deformation anomalies in other sectors of the mining operation, which could potentially lead to similar catastrophes," said Pınar Büyükakpınar of the GFZ German Research Centre For Geosciences, who published the study in The Seismic Record with her colleagues.
The Çöpler Gold Mine is one of the largest gold producers in the country, accounting for 20% of total gold production in Türkiye. The massive landslide buried nine miners.
The study uses seismic and remote sensing data, along with site investigations, to analyze the disaster. Processing of spaceborne radar data through the interferometric synthetic aperture radar (InSAR) technique showed continuous slow deformation at the site over the past four years, at an average rate of 60 millimeters per year.
Although it has not failed so far, a cyanide leach pond on the site also exhibits an average deformation of 85 millimeters per year. "This calls for urgent action, as our observations suggest an increasing risk the pond could fail, potentially releasing toxic waste into the nearby Euphrates River," said Büyükakpınar.
"Our findings underline the need for seismic detection systems that go beyond earthquakes to include secondary hazards like landslides," she said. "Long term monitoring and rapid detection of such events are critical for disaster response and mitigation."
The mine sits in a tectonically active area near the North and East Anatolian faults, Büyükakpınar said. "This region has a robust seismic monitoring network due to its seismic history, which allowed for high-resolution recording of the landslide."
The researchers analyzed data from seismic stations up to 400 kilometers away from the site, and found two distinct pulses 48 seconds apart indicating the source effect, corresponding to two mass sources. Their analysis indicates that the landslide included a westward detachment of debris on a steep slope and north-northeast movement on a gentler slope. Büyükakpınar and colleagues confirmed this pattern with photos and visits to the site.
The data provided an excellent opportunity to test real-time landslide detection algorithms that have been developed for Alaskan landslides "in a very different tectonic setting," Büyükakpınar noted.
"Thanks to the dense seismic network in the eastern part of the country, the algorithm successfully detected and located the event," she said. "In the future, it will be interesting to test whether the approach can be extended and applied to detect submarine landslides, as they are of high relevance and have the potential to induce tsunamis."
The study also described a combination of conditions that could lead to landslides in mining regions, including the long-term slow deformation caused by mining itself, increased pore pressure in rock due to rainfall or snowmelt, and steep slopes.
"The conditions are even stronger in tectonically active areas, which further amplifies the likelihood of failure, for example by dynamic triggering. We emphasize the importance of continuously monitoring slow movements and assessing the overall cumulative risk," Büyükakpınar said.
"Our findings suggest that integrating infrastructure management with continuous monitoring, rapid response mechanisms, and cooperation between regulators and mine operators can avoid future disasters," she added.
