Mapping lava flows from Nyamuragira volcano (1967–2011) with satellite data and automated classification methods

Abstract

The volume, location and extent of historical lava flows are important when assessing volcanic hazards, as well as the productivity or longevity of a volcanic system. We use a Landsat/Hyperion/ALI dataset and automated classification methods to map lava flows at Nyamuragira volcano (1967–2011) in the Democratic Republic of the Congo. The humid tropical climate of Nyamuragira is advantageous because its lava flows are emplaced onto heavily forested flanks, resulting in strong contrast between lava and vegetation, which contributes to efficient flow mapping. With increasing age, there is an increase in Landsat band-4 reflectance, suggesting lava flow revegetation with time. This results in a distinct spectral contrast to delineate overlapping flows emplaced ∼5 years apart. Areal extents of the flows are combined with published lava flow thicknesses to derive volumes. The Landsat/Hyperion/ALI dataset is advantageous for mapping future flows quickly and inexpensively, particularly for volcano observatories where resources are limited.