Locating Incipient Volcanic Vents Using Multidisciplinary Remote Sensing Data and Source Modeling Information

Knowing the location of an impending volcanic eruption would aid civil authorities in preparing the proper response to a developing crisis. Deterministic predictions of an eruption's location are not possible due to the highly complex nature of volcanic processes. Therefore, probabilistic forecasting techniques are required for estimating the location of an impending eruption. In this letter, we present a process for locating incipient volcanic vents using a spatial probability density function (PDF) constructed from a combination of multidisciplinary remote sensing data and source modeling information. Our process is flexible and designed to adapt to changing data availability and geologic conditions that typically occur during a developing volcanic crisis. The utility of our process is evaluated through the analysis of PDFs that forecast probable vent locations prior to Okmok's 2008 eruption. Results show our multidisciplinary approach produces a robust PDF that is spatially constrained to areas exhibiting various types of volcanic unrest and adapts to evolving conditions at the site.