Constraining explosive volcanic eruption parameters and environmental impacts using remote sensing observations of forest disturbance and recovery.

Explosive volcanic eruptions can damage or destroy surrounding forests, with the potential to alter their characteristics over different timescales. The damage can range from minor/temporary damage to total destruction and burial of vegetated landscapes. While eruptions resulting in vegetation damage over 100s of kilometres are rare, some volcanoes regularly impact and damage local vegetation, with frequencies of months to decades. The intensity and mechanism of the driving volcanic process influences the extent and style of the forest damage, with resulting timescales and patterns of regrowth reflecting the nature of initial impacts and local floral, climatic and environmental parameters. As such, vegetation damage holds potential as a novel proxy for the magnitude and nature of volcanic eruptions. Particularly in the case of volcanic eruptions without recorded observations, vegetation damage could potentially constrain parameters such as tephra-fall deposit thickness, dispersal and pyroclastic density currents (PDC) distribution. Additionally, mapping the initial impact and trajectories of recovery are key to understanding the long-term environmental consequences of volcanic eruptions. Here, we aim to constrain eruption magnitudes and deposit volumes, particularly in remote environments, using optical (Landsat 8, Sentinel-2) and radar (Sentinel-1) satellite data to study forest disturbance and recovery following an explosive volcanic eruption in Southern Chile: the 2015 eruption of Calbuco volcano. The 2015 eruption of Calbuco consisted of three explosive episodes between the 22nd-23rd of April resulting in buoyant ash plumes depositing tephra over 100s km2, pyroclastic flows reaching over 6km and lahars reaching over 15km. Thus, different intensities and spatial extents of damage were experienced by the surrounding temperate broadleaf forests. We identify areas impacted by the different eruptive deposits using the disturbance and recovery signatures from the optical and radar satellite data. We observe a decrease in vegetation coverage and health immediately following the eruption in the areas impacted by PDCs, lahars and tephra. PDC impacted regions exhibit the greatest decrease in vegetation coverage and health, and consequently a much slower vegetation recovery rate. We develop a satellite-based methodology through time series analysis and cluster analysis to understand the impact of explosive volcanic eruption on vegetation properties. This allows us to assess the extent and severity of forest disturbance caused by the eruption and to map the rates of post-eruption vegetation recovery. We hope to expand this methodology to be applied to different ecosystems and for different styles of eruption using freely available satellite data. With the eventual aim of developing a toolkit for identifying the footprint of past volcanic eruptions on forest environments.