“Know, that mountain -
it will not stand still.
Believe me it´s a troublemaker -
you can ask whoever you will.”
“La ballata di Longarone“, by Beppe Chierici 1969

The valley of Vajont (also Vaiont) was characterized in its upper part by a large catchment area, smoothed by ancient glacier activity, and a narrow gorge at the valley mouth, eroded by the river Vajont.
This geomorphologic situation and the nearby industrial areas of the Po plain made this valley a perfect site for a dam and a hydroelectric power station.
Construction began in 1956 and was completed in 1960, at this time it was the highest doubly curved arch dam in the world – rising 261.6 meters above the valley floor it was 190m broad across the top, with a storage capacity of 150 to 168 million cubic meters of water.

Video.1. A movie was produced to document the construction of the Vajont, considered a wonder of architecture and engineering at the time.

Filling of the reservoir began in February 1960, in October of the same year the lake was already 170m deep. Soon afterwards fissures were noted on the slopes of Mount Toc.
November 4, with a lake 180m deep, a landslide with 700.000 cubic meters of material plunged into the lake.
Since this landslide and the observed movements at Mt. Toc had occurred during the infill of the reservoir, it was assumed that the rising water table and water pore pressure in the slope of the mountains were the key destabilising factors. By slowly elevating the level of the reservoir it was hoped that these instabilities would decrease and the movements of the slope could be controlled. This strategy appeared to have been successful until mid-1963 when, between April and May, the depth of the reservoir was rapidly increased from about 195 to 230m. Rates of slope movement increased slightly but never exceeded 0,3cm per day. By mid-July, the depth had reached 240m, and some of the control points indicated small increases in displacement to 0,5cm per day, increasing until mid-August to 0,8cm day.
In early September the depth of the lake had reached 245m and the movements had accelerated until 3,5cm day. In late September the water level was slowly reduced in an attempt to slowdown the movement -  October 9, the reservoir’s depth had been lowered to 235m. Even so the slope movements continued to accelerate reaching more than 20cm per day.

October 9, 22.39 a part of the slope of Mt. Toc collapsed. Within 30 to 40 seconds estimated 240 – 270 million cubic meters of material plunged into the reservoir, filling completely the 400m deep gorge behind the dam. The wave generated by the impact of the landslide travelled 140m up on the opposite shore, reaching some buildings of the village of Erto. At that time, the reservoir contained 115 million cubic meters of water. The landslide pushed part of the water out of the lake, producing a wave with a maximal height of 230-240m. A 100 to 150m high wave overleaped the dam into the gorge of the Vajont, in direction of the larger and densely inhabited Piave valley.
The wave destroyed the villages of Longarone, Pirago, Villanova, Rivalta and Fae, in less than 15 minutes more than 2.000 people were killed – many bodies never to be found again.

Fig.1. Aerial photo of the valley of Vajont after October 9, 1963 (image in public domain), compare also with an photo taken previously to the landslide.

Video.2. Showing the movement of the final landslide and the directions of the generated wave.

Many geological factors caused the landslide of the Vajont, the world’s worst dam disaster, despite the fact that the dam is still standing today.

Geological surveys at the construction site had started already in the late 1920s. The area is characterized by a succession of Jurassic/Cretaceous to Eocene marls and limestone, folded into a large syncline with the valley following the fold axis. Between 1956 and 1960 it was realized that the slopes of Mt. Toc were more prone to instability than expected, since the mountain’s outer flanks consisted of an ancient landslide deposit and not of bedrock, as initially inferred. The Vajont river subsequently excavated the gorge into the landslide deposits and underlying bedrock. Research carried out after the event confirmed the presence of interbedded layers of green claystone (5-10 cm thick) in the limestone of the Fonzaso Fm. (an only 10-40 m thick formation), which could act as sliding planes. The harder rocks were also fissured by tectonic movements and the rebound of the Alps after the removal of the weight of the overlying glaciers at the end of the last glacial maximum.
This metastable situation was heavily weakened by the increase of the water level inside the mountain due the rising lake level and strong rainfalls in August, September and October 1963.

Fig.2. Summary of events recorded at the Vajont, modified after MÜLLER 1964 and BELLONI & STEFANI 1992 – geological investigations and proposed models, precipitation, water levels in the reservoir and groundwater levels (measured with piezometers) and rate of movements. The last rise of the reservoir level was accompanied also by stronger earthquakes coming from the slopes of Mt. Toc. Note also how the groundwater level became after 1961 synchronous with the reservoir level, indicating that the previously isolated aquifers in the mountain became (trough the sliding planes?) connected to the lake.

After the catastrophe a legal investigation was initiated to clarify if the landslide was predictable.
For more than three years the movements were monitored and various geologists studied the slope, drill cores intersected brittle zones and shear zones with broken rocks were discovered during the construction deep inside the mountain. Some geologists warned of a deep-seated landslide, like Austrian engineer Leopold Müller in 1960 ad later Italian geologists Eduardo Semenza and Franco Giudici, but the most pessimistic views were ignored and a model with more superficial sliding planes, proposed in 1960 by Italian geologist Pietro Caloi, who however changed opinion a year later, adopted. The local authorities also feared minor landslide events that could threat the local villages on the shores of the reservoir.
Minor landslides, as happened in 1960, were always expected, in 1961 even the construction of a by-pass tunnel was started, just in case the reservoir would became partially obstructed by a mass wasting. Also in 1961 even hydrologic simulations (Ghetti & Marzolo 1961/62) of the impacts of a landslide (of minor dimensions with a generated wave of 30 meters) on the lake were carried out. As results a security water level of 700m a.s.l. was recommended (surpassed in 1963 by 10 meters).

Also important factors transforming a natural event in a catastrophe was the unclear and chaotic situation of responsibilities and various conflicting interests in the project.

The continuous rejection of the worst case scenario – a gigantic single landslide – by the authorities of the company and the government was in part results of the lack of research on such events. In the first half of the 20th century the interests of geologists and engineers focused mainly towards slow mass movements, like slope deformations, better studied and understand at these times. After the tragedy of the Vajont the interests shifted back to the research on rock mechanics and fast mass movements.
There were also a political and financial conflict of interest – a gigantic investment and many politicians directly involved in the promotion and support of the project.
Finally also the situation of responsibilities was unclear – the dam was initiated by a private company, but in March 1963 the project was nationalized. The increased movements were observed by engineers of the company and the public consultants, but in the end nobody felt responsible for risk communication to the public. When in the last weeks of September and early October 1963 it was realized that the worst case scenario – a mass wasting with over 200 million cubic metres of volume moving as a single block – was occurring, it was already too late.

Bibliography:

ABBOTT, P.L. (2009): Natural Disasters. 8th ed. McGraw Hill Publisher, New York: 541
BELLONI, L.G. & STEFANI, R.F. (1992): Natural and induced seismicity at the Vajont slide. In: Semenza, E., Melidoro, G. (Eds.), Proc. Meeting 1963 Vaiont Landslide, Ferrara 1986. Univ. of Ferrara, Ferrara: 115- 132
BORGATTI, L. & SOLDATI, M. (eds.): Geomorphology and slope instability in the Dolomites (Northern Italy): from glacial to recent geomorphological evidence and engineering geological applications. Field trip Guide book P22, 32nd International Geological Congress, Florence August 20-28, 2004: 53
HYNDMAN, D. & HYNDMAN, D. (2010): Natural Hazards and Disasters. 3th ed. Brooks/Cole Publisher, Belmont: 571
KILBURN, R.J.C. & PETLEY, D.N. (2003): Forecasting giant, catastrophic slope collapse: lessons from Vajont, Northern Italy. Geomorphology 54: 21-32
MÜLLER, L. (1964): The Rock Slide in the Vaiont Valley. Felsmechanik und Ingenieurgeologie – Rock Mechanics and Engineering Geology Vol. 2(3/4): 10-16
ROSSI, D. & SEMENZA, E. (1965): Carte geologiche del versante settentrionale del M. Toc e zone limitrofe, prima e dopo il fenomeno di scivolamento del 9 ottobre 1963, Scala 1:5000. Istituto di Geologia dell´Universitá di Ferrara.
SEMENZA E. (2001): La Storia del Vaiont – raccontata dal geologo che ha scoperto la frana. K-flash edizioni, Ferrara: 279
SUPERCHI, L.; FLORIS, M.; GHIROTTI, M.; GENEVOIS, R.; JABOYEDOFF, M. & STEAD, D. (2010): Technical Note: Implementation of a geodatabase of published and unpublished data on the catastrophic Vaiont landslide. Nat. Hazards Earth Syst. Sci., 10: 865-873