Advanced analysis of satellite data reveals ground deformation precursors to the Brumadinho Tailings Dam collapse

Abstract

Catastrophic failure of a tailings dam at an iron ore mine complex in Brumadinho, Brazil, on 25th January 2019 released 11.7 million m3 of tailings downstream. Although reportedly monitored using an array of geotechnical techniques, the collapse occurred without any apparent warning. It claimed more than 200 lives and caused considerable environmental damage. Here we present the Intermittent Small Baseline Subset (ISBAS) technique on satellite-based interferometric synthetic aperture radar (InSAR) data to assess the course of events. We find that parts of the dam wall and tailings were experiencing deformation not consistent with consolidation settlement preceding the collapse. Furthermore, we show that the timing of the dam collapse would have been foreseeable based on this observed precursory deformation. We conclude that satellite-based monitoring techniques may help mitigate similar catastrophes in the future.

The challenge to use multi-temporal InSAR for landslide early warning

“…By exploiting the information contained in the radar phase differences of multiple complex SAR images, surface deformations can be derived with millimetric accuracy (Crosetto et al 2016). Application areas include detection and monitoring of landslides (Carlà et al 2019) and subsidence (Khorrami et al 2020) as well as structural health monitoring of critical infrastructure (Schlögl et al 2021;Grebby et al 2021).…”

Ground motion baseline analysis of the Cheshire UK GeoEnergy Observatory

“…Satellite radar data were downloaded from the Copernicus Open Access Hub ( https://scihub.copernicus.eu/dhus/#/home ) of the European Space Agency and processed by Terra Motion Limited using a technique known as ISBAS 23 . This method was chosen due to its ability to provide a higher density of measurements over rural areas 24 . Retrieving ground motion information for non-urban and semi-vegetated areas using standard InSAR approaches with C-band (~ 5.6 cm wavelength) satellite radar imagery is challenging due to sources of noise like temporal decorrelation 25 .…”

Remote Monitoring of Ground Motion Hazards in High Mountain Terrain Using InSAR: A Case Study of the Lake Sarez Area, Tajikistan

“…Another advantage of the ISBAS algorithm, and a consequence of the high density of measurements that can be achieved, is the ability to map ground motion without the use of ground control, such as GPS receivers and artificial corner reflectors. This has been demonstrated for a variety of disaster risk reduction applications associated with slope failures, such as landslide inventory mapping [45,46] and tailings dam monitoring [47]. The ability to monitor ground motion without any requirements for ground control or repeated field visits is particularly advantageous in inaccessible mountainous terrain.…”