Rockfall Runout Simulation Fine-Tuning in Christchurch, New Zealand

A calibration of rockfall propagation model based on a full-scale test performed in the Richmond Hill area after the 2011 Christchurch earthquake is presented. Two quantitative accuracy measures have been used to guide the calibration process: (1) the mean distance between the location of arrest of actual blocks and the location of all simulated blocks; (2) the difference between the velocity of simulated blocks and the velocity of actual rockfalls along the propagation path. While the first measure did not provide a sensitive response between the different simulations, the second allowed calibrating the model through a number of trial and error repetitions with progressively improved parameters. This measure is objective, repeatable, and could be integrated within an optimization procedure for an automated calibration of rockfall model.