2008-02 – Bukavu/Cyangugu Seismic Crisis

On February 3rd 2008, a seismic sequence started on the Southern shore of Lake Kivu with a Mw 5.9 Earthquake. It caused widespread damages and killed at least 39 people in the cities of Bukavu and Cyangugu along the border between DR of Congo and Rwanda. The main earthquake was followed by a large number of aftershocks among which three were of magnitude above 5 and nine were of magnitude above 3.7. It is the largest event ever recorded in the area after the Mb 6.2 earthquake that occurred on October 24th 2002 at Kalehe, 35 km South West of Goma and that killed two people.

The Bukavu/Cyangugu seismic episode took place in the South Kivu Volcanic Province, a seismically active transfer zone between the Kivu basin and the Rusizi basin.

Local seismicity located by the temporary seismic network and the Goma Volcano Observatory seismic network from 2008 February 8–30. Among the 700 aftershocks detected by these networks, only 37 could be located with an estimated horizontal error (ERH) of less than 10 km (dark grey circles). Green and orange stars, respectively show the location of the 2008 Bukavu-Cyangugu mainshock obtained from the geodetic inversion and the teleseismic data (this study). Inverted yellow triangles are the seismic stations; thin yellow lines are the political borders. Source: d'Oreye et al. 2010)
Local seismicity located by the temporary seismic network and the Goma Volcano Observatory seismic network from 2008 February 8–30. Among the 700 aftershocks detected by these networks, only 37 could be located with an estimated horizontal error (ERH) of less than 10 km (dark grey circles). Green and orange stars, respectively show the location of the 2008 Bukavu Cyangugu mainshock obtained from the geodetic inversion and the teleseismic data (this study). Inverted yellow triangles are the seismic stations; thin yellow lines are the political borders.
Source: d’Oreye et al. (2010)

This earthquake is of particular interest due to its shallow depth, its proximity to active volcanoes in the Virunga Volcanic province (located 100 km to the north), and Lake Kivu, which contains high concentrations of dissolved carbon dioxide and methane.

The possible similarity with dyking events recognized in other parts of East African Rift suggested the potential association of the earthquake with a magmatic intrusion, emphasizing the necessity of accurate source parameter determination.

The 2008 Bukavu/Cyangugu event has been studied in details by d’Oreye et al. (2010). In the absence of a local seismic network, studies using satellite Radar Interferometry (InSAR) and teleseismic waveforms allowed to accurately infer the source parameters. In particular, InSAR allowed locating the earthquake under the southern part of Lake Kivu and not along the Birava peninsula, as previously inferred by other studies using global seismic databases.

Coseismic ground deformation (range displacement – positive away from the satellite, in cm) and models for uniform slip inversion for the 2008 Bukavu-Cyangugu earthquake. A): Ascending track ALOS PALSAR interferogram spanning 29 December 2007 - 30 march 2008. D) Descending track ENVISAT ASAR interferogram spanning 10 January – 14 February 2008. Corresponding best-fit models of a uniform slip elastic dislocation are shown in B) and E). The black solid rectangle indicates the projection of the fault rupture onto the surface and the dashed line shows where the fault, if lengthened, would cut the surface. Residuals are shown in C) and F).
Coseismic ground deformation (range displacement – positive away from the satellite, in cm) and models for uniform slip inversion for the 2008 Bukavu-Cyangugu earthquake. A): Ascending track ALOS PALSAR interferogram spanning 29 December 2007 – 30 march 2008. D) Descending track ENVISAT ASAR interferogram spanning 10 January – 14 February 2008. Corresponding best-fit models of a uniform slip elastic dislocation are shown in B) and E). The black solid rectangle indicates the projection of the fault rupture onto the surface and the dashed line shows where the fault, if lengthened, would cut the surface. Residuals are shown in C) and F).

Geodetic and seismic modeling and inversions lead to highly consistent results. The focal mechanism, strike and dip are consistent with local tectonics. From this, d’Oreye et al. (2010) infer that the rupture was brittle and occurred with little aseismic deformation, discarding the hypothesis of magma involvement, at least at shallow depth.

In the mature eastern branch of the EAR, magmatism is known to play a major role in lowering the seismicity during rift opening. Magma-assisted opening seems also to prevent the occurrence of large magnitude earthquakes in active volcanic provinces of the western branch of the EAR, such as in Virunga Volcanic Province. The mode of extension in the younger, yet extinct, South Kivu Volcanic Province is however poorly understood.

Results from d’Oreye et al. (2010) provide insights into the style of rifting occurring in that part of the EAR and, hence, will aid future studies on seismic risk. It also highlights the possible inter-related hazards (earthquakes, triggered landslides, magmatism, the presence of dissolved carbon dioxide and methane in the Lake Kivu water…), and, as this event occurred along political borders, the importance of a trans-border collaboration and data sharing.

View of a landslide triggered by the Cyangugu-Bukavu earthquake on Ibinja Island (located by the arrow in the inset). The ellipse encircles the top of banana trees now underwater. Viewing directions are marked on both side of the picture. The landslide rupture has replaced the previously gentle slope of the shoreline. Source: d'Oreye et al. (2010)
View of a landslide triggered by the Cyangugu-Bukavu earthquake on Ibinja Island (located by the arrow in the inset). The ellipse encircles the top of banana trees now underwater. Viewing directions are marked on both side of the picture. The landslide rupture has replaced the previously gentle slope of the shoreline.
Source: d’Oreye et al. (2010)

Influence on the Virunga volcanoes

During the three months that followed the 2008 Bukavu/Cyangugu earthquake, unusual NNW-SSE-elongated swarms of long-period (LP) earthquakes in the VIrunga were recorded by GVO. These earthquakes occurred at a shallow depth, between Nyamulagira and the city of Gisenyi, which is located about 15 km SE of Nyiragongo. During inter-eruptive periods, LP earthquakes are mostly centred on the main edifice of Nyamulagira.  The unual swarm observed in February-April 2008 rose the question of a potential influence of the Bukavu/Cyangugu earthquake on volcanic activity.

Long-period earthquakes (yellow dots) recorded by the Goma Volcano Observatory in February, March and April 2008. Source of earthquake locations: Courtesy of GVO.
Long-period earthquakes (yellow dots) recorded by the Goma Volcano Observatory in February, March and April 2008. Source of earthquake locations: Courtesy of GVO.

In early March 2008, fast movements of the Nyiragongo lava lake and fumaroles in the Nyamulagira caldera were observed by a team of Congolese and Belgian scientists. These additional observations suggested that changes in the magmatic pressure of the shallow plumbing systems of Nyiragongo and Nyamulagira occurred. These pressure changes may be induced by the seismic waves of the tectonic earthquakes.

Reference:

d’Oreye N., Gonzalez P., Shuler A., Oth A., Bagalwa M., Ekström G., Kavotha D., Kervyn F., Lucas C., Lukaya F., Osodundu E., Wauthier C., Fernandez J.
 (2010) – Source parameters of the 2008 Bukavu-Cyangugu earthquake estimated from InSAR and teleseismic data.
 Geophysical Journal International, vol. 184, Iss. 2, 934-948