Spaceborne Support to the Prediction of Lava Distance to Run: a Discussion

The prediction of lava flow distance-to-run is a key-activity in support to risk mapping and/or planning the emergency response to an eruptive (effusive) volcano events. Distance-to-run is driven by several chemical and physical parameters, and can be approached by estimating the mass eruption rates through their radiant flux proxy.

Radiant fluxes mainly depend on spectral emissivity, which is defined as the efficiency of physical bodies in radiating thermal energy at a specific wavelength: it is seldom measured and mostly assumed to be ɛ = 1 or 20% less (ɛ = 0.8), and not to change as a function of temperature and wavelength. However, true (absolute) emissivity has substantial importance in passive remote sensing, due to its close relationship with wavelength depending Land Surface Temperatures, and the inherent impact on the estimate of mass eruption rates.

To fill this gap in knowledge involving spectral emissivity, we set up a programme of systematic in-lab measurement of spectral emissivity on rock samples by three independent methods, using Fourier Transform Infra-Red (FTIR) spectroscopy between 0.6÷15.0 μm wavelength range and at temperatures up to 400÷900°K. Rock samples representative of recent (2001-2017) Mount Etna (Italy) lava flows were collected in sampling grids scaled to the spatial resolution of decametric multispectral payloads equipped with Thermal Infra-Red channels on-board Landsat-5, Landsat-7 and ASTER (eruptions 2001 to 2003), and Landsat-8 (eruptions 2015-2017).

The aim of this approach is to estimate the lateral spatial heterogeneity of spectral emissivity on ground at known volcanic targets, and to assess whether spaceborne estimates are reliable enough for incorporating experimental emissivity laws into the techniques of automated eruption detection and quantitative assessment and monitoring of lava flows. The refinement of emissivity contributions to the accuracy of actual temperatures, is expected to improve the remote-sensed contributions to mass flux estimates - through the radiant flux proxy, by an appropriate cooling model - whereas relevant rheological parameters (latent heat, vesicularity, and percentage of crystals grown) still cannot be estimated from remote, and rely upon known spatial/temporal distributions on the ground.