Forecasting of flood basalt eruptions: lessons from Bárðarbunga

Abstract:

The 2014-15 eruption from Bárðarbunga volcano produced ~1.6 km³ of lava and was the largest Icelandic eruption since the Laki eruption in 1783-84. Nevertheless the recent eruption was still an order of magnitude smaller than Laki and some previous eruptions of Bárðarbunga, which raises the question of whether Bárðarbunga currently has the potential for a larger eruption and, more generally, can we predict how large this type of eruption is going to be once it has started? The eruption was accompanied by a gradual caldera collapse, which provides a plausible mechanism for producing the largest flood basalts that have occurred in Iceland. Here we investigate the potential for larger eruptions through coupled modeling of the caldera collapse and eruption. Our model provides predictive capacity in terms of the end date and final volume of eruption; based on a simple model of piston collapse constrained by subsidence data from a continuously operating GPS station within the caldera, we made a prediction of the end date that was off by only 9 days. We have extended our modeling to consider the effects of episodic pressure change due to earthquakes on faults beneath the caldera, constrained by GPS and InSAR. Our model constrains the aerial extent and volume of the magma body and thus its potential to develop into a larger eruption. The model can be applied to future eruptions that are accompanied by caldera collapse, to forecast the duration and eventual erupted volume, which is critical information for civil authorities.