James, Mike R. and Applegarth, Louisa and Pinkerton, Harry and Fryer, Tom (2011) Estimating rheological properties of lava flows using high-resolution time lapse imaging. In: AGU Fall Meeting 2011, 2011-12-05.
Abstract
During effusive eruptions, property and infrastructure can be threatened by lava flow inundation. In order to maximise the effectiveness of the response to such an event, it is necessary to be able to reliably forecast the area that will be affected. One of the major controls on the advance of a lava flow is its rheology, which is spatially and temporally variable, and depends on many underlying factors. Estimating the rheological properties of a lava flow, and the change in these over space and time is therefore of the utmost importance. Here we report estimates of rheological properties made from geometric and velocity measurements on integrated topographic and image data using the method of Ellis et al. (2004) (Ellis B, Wilson L & Pinkerton H (2004) Estimating the rheology of basaltic lava flows. Lunar & Planetary Science XXXV Abst. 1550). These are then compared to the viscosity predicted from composition and temperature by the GRD model (Giordano D, Russell JK, & Dingwell DB (2008) Viscosity of Magmatic Liquids: A Model. Earth & Planetary Science Letters, 271, 123-134). During the 13 May 2008 - 6 July 2009 eruption of Mt Etna, Sicily, lava flows were emplaced into the Valle del Bove, reaching a maximum length of >6 km. Towards the end of the eruption, multiple channelized aa flows were active simultaneously, reaching tens to hundreds of metres in length. Flow lifetimes were of the order hours to days. In the last month of the eruption, we installed a Canon EOS 450D camera at Pizzi Deneri, on the north side of the Valle del Bove, to collect visible images at 15-minute intervals. On one day, topographic data (using a Riegl LPM-321 terrestrial laser scanner) and thermal images (using a FLIR Thermacam S40) were also collected from this location. The fronts of some of the larger flows were tracked through the time lapse image sequence. Using knowledge of the camera imaging geometry, the pixel tracks were reprojected onto the topographic surface to determine flow advance in 3-D geographic coordinates. Integrating the tracking results with the topographic data allows flow lengths and velocities to be extracted. Using these parameters together with estimates of the flow width and thickness, we estimate effective yield strengths, apparent viscosities and Gratz numbers for the tracked flows. We then evaluate the success of this method using predicted viscosities from the GRD model of Giordano et al. (2008).