Hazard assessment at Mount Etna using a hybrid lava flow inundation model and satellite-based land classification

Using a lava flow emplacement model and a satellite-based land cover classification, we produce a map to allow assessment of the type and quantity of natural, agricultural and urban land cover at risk from lava flow invasion. The first step is to produce lava effusion rate contours, i.e., lines linking distances down a volcano’s flank that a lava flow will likely extend if fed at a given effusion rate from a predetermined vent zone. This involves first identifying a vent mask and then running a downhill flow path model from the edge of every pixel around the vent mask perimeter to the edge of the DEM. To do this, we run a stochastic model whereby the flow path is projected 1,000 times from every pixel around the vent mask perimeter with random noise being added to the DEM with each run so that a slightly different flow path is generated with each run. The FLOWGO lava flow model is then run down each path, at a series of effusion rates, to determine likely run-out distance for channel-fed flow extending down each path. These results are used to plot effusion rate contours. Finally, effusion rate contours are projected onto a land classification map (produced from an ASTER image of Etna) to assess the type and amount of each land cover class falling within each contour. The resulting maps are designed to provide a quick look-up capability to assess the type of land at risk from lava extending from any location at a range of likely effusion rates. For our first (2,000 m) vent zone case used for Etna, we find a total of area of ~680 km2 is at risk from flows fed at 40 m3 s−1, of which ~6 km2 is urban, ~150 km2 is agriculture and ~270 km2 is grass/woodland. The model can also be run for specific cases, where we find that Etna’s 1669 vent location, if active today, would likely inundate almost 11 km2 of urban land, as well as 15.6 km2 of agricultural land, including 9.5 km2 of olive groves and 5.2 km2 of vineyards and fruit/nut orchards.