Evaluation of geophysical mass flow models using the 2006 block-and-ash flows of Merapi Volcano, Java, Indonesia: Towards a short-term hazard assessment tool

Abstract

The dynamics and depositional processes associated with block-and-ash flows (BAFs) are most commonly inferred to be a function of granular or inertial grain flow, similar to debris flows and cold rock avalanches. Existing geophysical mass flow models are either based on frictional (Mohr–Coulomb) behavior (the Titan2D model developed at the State University of New York at Buffalo, USA) or another rheological law (i.e., a constant retarding stress), eventually adding some viscous and turbulent components (the VolcFlow model developed at the Laboratoire Magmas et Volcans, Clermont-Ferrand, France). The 2006 BAFs of Merapi present a rare opportunity to test these two well-established models against a well-constrained field example. Integration of high-resolution field-based data into numerical simulations allows the validity of these models to be tested and rapid quantification of best-fit input parameters. We first show that with the incorporation of spatially varying bed friction angles, Titan2D is capable of reproducing the paths, runout distances, areas covered and deposited volumes of the 2006 Merapi flows over highly complex topography. However, some discrepancies with field data are noted and the velocity and travel time of the flows do not match entirely. Using a single free parameter (a constant retarding stress), simulations obtained with the VolcFlow model also reproduce the morphology and distribution of the natural deposits as well as the time of emplacement and velocities of the flows. The results suggest that the performance of these models in simulating actual events is critically dependent on: (1) the calibration of the model by using extensive field-based data such as deposit distribution, and processes of flow generation, transport and deposition; (2) the incorporation of a suitable numerical topographic dataset (i.e., high-resolution digital elevation model), and (3) the choice of input parameters, such as location and volume of the initial pile of material and source characteristics (single or multiple dome-collapse, dome-collapse duration and total volume of collapsed material). Sensitivity analyses and inundation maps based on the probability of impact were used to produce a suite of potentially inundated areas from future gravitational dome-collapse events affecting the Gendol valley and adjacent areas on the southern flank of the volcano. Our results provide the basis for defining hazard zonations of key areas at risk from BAFs which will be generated during future comparable eruptions at Merapi.