14 July 2017
Understanding the La Palma mega-landslide hypothesis: part 2
Earlier this week I explored the main structural features of La Palma that have led some to propose that there is the potential for a mega-landslide there. This proposed volcanic flank collapse would be immense – the proposed volume is up to about 500 cubic kilometres. The idea that gained some popular traction is that this landslide could generate a tsunami that would devastate a large part of the coastline on both sides of the Atlantic.
Of course La Palma has undergone a previous flank collapse event, and there have been similar collapses elsewhere in the Canary Islands. Interestingly, none of these appear to have generated widespread tsunami deposits around the Atlantic basin. The key to the idea such an event developing again is the events of the major eruption in 1949. During this event, a fault structure was observed to develop along a part of the Cumbre Vieja ridge. This has been interpreted as indicating movement of the flank of the volcano towards the west, and thus the development of a potential flank collapse landslide on the southern part of La Palma. I spent a day up on Cumbre Vieja, with the main aim of taking a look at this fault scarp. Armed with a map from one of the key papers I roamed up and down the mountains on a most beautiful day. The feature that I found is remarkably unremarkable in landslide terms. Based upon the maps, the most obvious feature that I could find is the scarp shown below:-
This scarp is about 3 m high at the maximum – if you look carefully there is a person walking down the path for scale. It is reasonably clear in the landscape for some hundreds of metres before merging into the flank of one of the volcanic craters:
I have to admit that I struggle to believe that this feature is a key component of the mega-landslide hypothesis given its small size, but I could not find any more convincing alternative. Perhaps I missed the correct feature. The scarp suggests limited movement on the landslide, even if it is a correct interpretation, suggesting in turn that an actual failure would be unlikely in the foreseeable future.
This fault scarp did not reactivate in the smaller 1971 eruption. Monitoring of the flanks suggest that there is no sign of rapid current movement on this scale, but that there may be some signs from satellite data of very low velocity creep on this slope, which is not a surprise (this may well be true of all large slopes in weak materials). Thus, the mega-landslide hypothesis is that this volcano would only become unstable in a future eruption, and that in such circumstances the flank could collapse in a single coherent block to generate the tsunami. The paper that modeled the tsunami, Ward and Day (2001) modeled a landslide of about 450 cubic kilometres – i.e. they took the very largest volume that is imaginable. This seems a little odd to me – the rear scarp of their mega-landslide appears not to mobilise the scarp shown above, but one considerably to the east, creating a much larger block. I am not sure that I understand the reasoning for this. The model then assumes a series of extreme scenarios:
- The landslide occurs as a single coherent mass along the entirety of the ridge (thus over a distance of 25 kilometres);
- The landslide occurs as a single coherent mass through the cross-section (i.e. there is a single failure event over the 15 kilometre cross-section of the slope, rather than a series of retrogressive slip blocks);
- This huge block remains intact over a travel distance of 15 km before fragmenting;
- The landslide mass rapidly reaches a peak velocity of 100 m/sec (360 km/h)
A change (reduction) in any of these parameters would yield a much smaller tsunami. For example, subsequent work (Abadie et al. 2012) has taken the ““credible worst case scenario” (derived using slope stability analysis), to have a volume of 80 cubic kilometres (but note the factor of safety of the slope was found to be considerably higher than one, indicating that the slope is not particularly unstable). Modelling of the tsunami generated by such a landslide, using a more refined tsunami simulation, does generate a very significant wave close to La Palma. This wave would be significant as it crossed the continental shelf off the east coast of North America, but would lose a great deal of energy due to frictional effects in this region. Thus, for a 80 cubic kilometre “credible worst case scenario” flank collapse on La Palma, wave heights on the east coast of the USA were found by Tehranirad et al. (2015) to be less than 2 metres along the coastline.
This is not the disastrous scenario that the newspapers have so enjoyed featuring.
Abadie S., Harris J.C., Grilli S.T. and R. Fabre, 2012. Numerical modeling of tsunami waves generated by the flank collapse of the Cumbre Vieja Volcano (La Palma, Canary Islands) : tsunami source and near field effects. Journal of Geophysical Research, 117: C05030.
Tehranirad, B., Harris, J.C., Grilli, A.R. et al. 2015. Far-Field Tsunami Impact in the North Atlantic Basin from Large Scale Flank Collapses of the Cumbre Vieja Volcano, La Palma. Pure and Applied Geophysics 172: 3589.
Ward S. N. and S. Day, 2001. Cumbre Vieja Volcano potential collapse at La Palma, Canary Islands. Geophysical Research Letters, 28: 397–400.