Digital photogrammetry as a tool in analogue modelling: applications to volcano instability

Abstract

Three techniques of digital photogrammetry have been applied successfully to laboratory analogue models to study surface displacements caused by various volcano deformation types. Firstly, side-perspective videos are used to differentiate profile displacements between cryptodome intrusion models and models deforming by ductile inner-core viscous flow. Both models show similar morphologic features including a bulged flank and an asymmetric upper graben. However, differences in displacement trajectories of the bulge crest reflect upward intrusion push contrasting with essentially downward displacement vectors of weak core models. The other two techniques use vertical views correlated automatically either as time-sequence monoscopic views or as coeval stereoscopic pairs. This exploits to a maximum the method’s potential by imaging surface displacements over the whole model. Successive monoscopic photograms, because they suffer only moderate numerical processing for topographic effect removal, can detect very small displacements occurring early in deformation processes. As illustrated by analysis of intrusion models, the monoscopic method allows prediction of fault locations and main displacement locations. It can also anticipate the principal strain directions, and separate different deformation stages. On the other hand, the stereo-photogrammetry technique, although more complicated, provides topography and volume changes, as well as pictures of surface displacements in three dimensions. Results are presented for the spreading of volcano models on a ductile substratum and viscous cored cones. We have found digital photogrammetry to be a useful tool for analogue modelling, because it provides quantitative data on surface displacements, including movement invisible to the eye, as well as topographic changes. It is a good method for investigating and comparing different deformation mechanisms. It is especially useful for interpretation of displacement patterns obtained from monitoring of natural active volcanoes. In fact, results of the methods used in the laboratory can be directly compared with field data from geodetic or photogrammetric surveys, as at Mount St. Helens in 1980.