Best-fit results from application of a thermo-rheological model for channelized lava flow to high spatial resolution morphological data

The FLOWGO thermo-rheological model links heat loss, core cooling, crystallization, rheology and flow dynamics for lava flowing in a channel. We fit this model to laser altimeter (LIDAR) derived channel width data, as well as effusion rate and flow velocity measurements, to produce a best-fit prediction of thermal and rheological conditions for lava flowing in a similar to 1.6 km long channel active on Mt. Etna ( Italy) on 16th September 2004. Using, as a starting condition for the model, the mean channel width over the first 100 m ( 6 m) and a depth of 1 m we obtain an initial velocity and instantaneous effusion rate of 0.3 - 0.6 m/s and similar to 3 m(3)/s, respectively. This compares with field- and LIDAR-derived values of 0.4 m/s and 1 - 4 m(3)/s. The best fit between model-output and LIDAR-measured channel widths comes from a hybrid run in which the proximal section of the channel is characterised by poorly insulated flow and the medial-distal section by well-insulated flow. This best-fit model implies that flow conditions evolve down-channel, where hot crusts on a free flowing channel maximise heat losses across the proximal section, whereas thick, stable, mature crusts of 'a'a clinker reduce heat losses across the medial-distal section. This results in core cooling per unit distance that decreases from similar to 0.02 - 0.015 degrees C m(-1) across the proximal section, to similar to 0.005 degrees C m(-1) across the medial-distal section. This produces an increase in core viscosity from similar to 3800 Pa s at the vent to similar to 8000 Pa s across the distal section.