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

Abstract

[1] In this work, we study waves generated by the potential collapse of the west flank of the Cumbre Vieja Volcano (CVV; La Palma, Canary Island, Spain) through numerical simulations performed in two stages: (i) the initial slide motion and resulting free surface elevation are first calculated using a 3D Navier-Stokes model; (ii) generated waves are then input into a 2D (horizontal) Boussinesq model to further simulate propagation to the nearby islands. Unlike in earlier work on CVV, besides a similar extreme slide volume scenario of 450 km 3 , in our simulations: (i) we consider several slide scenarios featuring different volumes (i.e., 20, 40, 80 km 3 ), which partly result from a geotechnical slope stability analysis; (ii) we use a more accurate bathymetry; and (iii) an incompressible version of a multiple-fluid/material Navier-Stokes model. We find wave trains for each scenario share common features in terms of wave directivity, frequency, and time evolution, but maximum elevations near CVV significantly differ, ranging from 600 to 1200 m (for increasing slide volume). Additionally, our computations show that significant energy transfer from slide to waves only lasts for a short duration (order 200 s), which justifies concentrating our best modeling efforts on the early slide motion phase. The anticipated consequences of such wave trains on La Palma and other Canary Islands are assessed in detail in the paper.

On the effect of the water body geometry on landslide–tsunamis: Physical insight from laboratory tests and 2D to 3D wave parameter transformation

“…Examples of this are provided by Wiegel et al (1970) with a 2D (x)  x -1/5 , Heller and Hager (2010) with a 2D (x)  x -4/15 or Heller and Spinneken (2013) with a 2D (x)  x -3/10 . The experimentally deduced variation is similarly large for 3D studies namely a 3D (r)  r -19/20 in Davidson and Whalin (1974), H 3D (r)  r -2/3 in Huber and Hager (1997), H 3D (r)  r -0.81 in Panizzo et al (2005) and up to a 3D (r)  r -1.42 in Abadie et al (2012). The decay may also differ for the primary and secondary wave (Panizzo et al, 2005).…”

“…Generic model studies systematically vary parameters (slide properties, hill slope angle, water depth) which may be estimated a priori for real-world events, and express the unknown wave parameters (amplitude, height, period) as a function of these parameters. The resulting empirical equations can be very efficient in predicting future events (Heller et al, 2009), and are often the most straightforward method if time is limited.At the very least, such equations can help to determine whether or not a prototype specific numerical (Løvholt et al, 2008;Abadie et al, 2012) or physical (Davidson and Whalin, 1974;Fuchs et al, 2011) model study is required. These latter methods are both considerably more expensive and time consuming than applying generic empirical equations.…”

“…At the very least, such equations can help to determine whether or not a prototype specific numerical (Løvholt et al, 2008;Abadie et al, 2012) or physical (Davidson and Whalin, 1974;Fuchs et al, 2011) model study is required. These latter methods are both considerably more expensive and time consuming than applying generic empirical equations.…”

Composite modelling of subaerial landslide–tsunamis in different water body geometries and novel insight into slide and wave kinematics

“…In method (ii), a numerical code is validated and calibrated and subsequently applied to a specific case (Ward and Day, 2003;Løvholt et al, 2008;Abadie et al, 2012). This method may result in more accurate predictions than method (i); however, it requires more skills, time (weeks to months) and resources.…”

“…The accurate description of wave kinematics is important for a number of applications, such as evaluating the kinetic wave energy, calibrating and validating the coupling of SPH with a computationally less expensive wave propagation model (Narayanaswamy et al, 2010;Abadie et al, 2012) or to investigate fluid loading on structures. The measured kinematics are compared with a set of non-linear, non-hydrostatic wave theories, aiming to predict theoretically the kinematics through a set of given wave parameters.…”