EGU2020-11244, updated on 10 Mar 2020
https://doi.org/10.5194/egusphere-egu2020-11244
EGU General Assembly 2020
© Author(s) 2020. This work is distributed under
the Creative Commons Attribution 4.0 License.

Back-analysis of rockfalls for the definition of an empirical vulnerability function for buildings

Sandra Melzner1, Paolo Frattini2, Federico Agliardi2, and Giovanni Battista Crosta2
Sandra Melzner et al.
  • 1Geoconsult ZT GmbH, Wals/Salzburg, Austria (sandra.melzner@geoconsult.com)
  • 2Universita degli Studi di Milano-Bicocca, Earth and Environmental Science, Milano, Italy

The vulnerability of buildings to the impact of rockfalls is a key component of Quantitative Risk Assessment for rockfall phenomena. Only a few attempts to quantitatively assess vulnerability have been presented in the literature due to the lack of high-quality rockfall and damage data. For processes such as debris flows, snow avalanches or earthquakes, well-established methods for the estimation of physical vulnerability are already available.

The present work aims to develop an empirical rockfall vulnerability function by coupling rockfall back-analysis modelling of several damaging events occurred in different lithological and geomorphological settings. A sound database of damage to specific categories of structures impacted by rockfalls is build up by archive research of historical events and high-quality field observations of recent events. Damages are classified according to four damage types: superficial (degree of loss: 0.1-0.2) to structural (degree of loss: 1.0). The back-analysis of rockfalls and the interaction with element at risk is performed with the 3D numerical model Hy- STONE. The code uses a hybrid modelling approach and random sampling of input parameters from different probability density distributions (uniform, normal, exponential) to account for the complexity of the rockfall process and influencing factors. The elements at risk are integrated as lines to the model, impact points being able to be displayed and extracted as point vector data. This enables a precise analysis of simulated energies and observed damage for each building impacted in the past to define an empirical vulnerability function. The empirical vulnerability function is established by fitting damage-energy data through a sigmoidal function. This empirical vulnerability function for buildings is fundamental to compute the expected degree of loss for each element of risk, especially in areas where no detailed rockfall or damage data is available.