Understanding Volcanic Hazard Impacts
Research in this area will mainly be used to improve the ways in which geophysical monitoring data forecasts volcanic activity, and its translation into improved warnings and alerts.
Multiple datasets are produced around volcanoes. These come from instruments that monitor perturbations in the magmatic plumbing system, changes in gas emissions and the deformation of the surrounding crust.
The relationship between these signals and observed volcanic activity is determined through modelling of these data in two ways. The first involves the development of physically-based models that describe cause and effect between changes in the magma source region and volcanic activity. The second uses time-series analysis of multiparametric datasets to identify their important stochastic characteristics and improve the interpretation and automatic detection of these pattern changes.
The detection and interpretation of small volcano-related earthquakes remains the most commonly used monitoring tool. Physically-based models to interpret these data will be further developed by (i) improving the relocation of earthquakes to map out stress changes in the storage region as magma rises; (ii) quantifying the seismic moment of the source to develop models for the dominant trigger mechanisms for volcano-related earthquakes and (iii) determining the relationship between small changes in the seismic velocity field and larger eruptive episodes.
Technical developments in the field of remote sensing of surface deformation have out-paced their application in understanding sub-surface changes and their relation to renewed or changed surface activity. STREVA addresses this by using Satellite Radar Interferometry (InSAR) to map fine-scale deformation fields at both the forensic and trial volcanoes at high spatial and temporal resolution. Joint inversion of InSAR (high spatial resolution) and cGPS (high time resolution and precision) shows remarkable potential for constraining magma fluxes and plumbing system geometries and in particular for constraining pressurisation of the source region.
Both the physically-based and stochastic models focus on the time-dependent characteristics of changes in monitoring data. This provides an essential evaluation of the time-scales over which operational changes in monitoring-based alerts can operate.
This area model development will use the existing datasets from STREVA’s forensic volcanoes and test and validate them for general use in the trial volcano settings and beyond.
This also focuses on improved spatial and temporal forecasting of the footprints of key volcanic hazardous processes, with particular emphasis on lahars (in line with the IRNH call and CARIBRISK outcomes). Existing widely-used models of lahar hazard (e.g. LAHARZ of the US Geological Survey; Iverson et al 1998) lack the detailed physical description to enable prediction of lahar arrival time and inundation and there is an urgent need for improved forecasting tools. Model performance is also limited by DEM resolution (Davila et al 2007) but future developments (e.g. TanDEM-X available in 2014) should lead to a step change in forecasting ability, and this WP will additionally benefit from close links with the CREDIBLE proposal (NERC PURE) in which Phillips is a co-investigator on uncertainty in lahar and avalanche models. Existing and new stochastic modelling approaches will be used to assess pyroclastic flow and ashfall, and characterise their input into lahar hazard.