An unmanned aerial vehicle provided the high-resolution data that allowed scientists to construct their first detailed map of erupting vents at Stromboli, one of the world’s most active volcanoes.

Strombolian activity is generally assumed to be driven by overpressurized gas slugs that rise through the magma‐filled volcanic conduit and burst at the surface. We develop an analytical model for this process that incorporates a generic, depth‐dependent, and non‐Newtonian magma rheology. Our model also describes the film draining after the burst and allows for the computation of the stresses exerted on the conduit walls using a new analytical solution for the thickness of an annular film flow. Using Stromboli volcano, Italy, as reference, it was evaluated with a specifically designed, non‐Newtonian rheological model based on petrochemical data. The results show the importance of using a realistic magma viscosity model when modeling the slug ascent: A 100 kg gas slug attains its maximum overpressure of 2.1 bars about 40 s before it rapidly expands and finally bursts at only 1.4 bars. This preburst pressure drop is not reproduced by constant viscosity models. The normal stresses on the conduit wall are generally dilatational above and adjacent to the slug and constrictive below it and after the burst. In both cases their maximum values exceed 3 bars, whereas the shear stresses stay negligible. Furthermore, it was found that the slug needs to start deeper than 25 m to build up its full burst pressure. At greater depth, the burst pressure decreases because of the adiabatic slug gas. Finally, a viscous plug at the top of the conduit is shown to increase the explosive potential of the volcano significantly.

ABSTRACT:The most commonly used method for satellite cloud top height (CTH) compares brightness temperature of the cloud with the atmospheric temperature profile. Because of the uncertainties of this method, we propose a photogrammetric approach. As clouds can move with high velocities, even instruments with multiple cameras are not appropriate for accurate CTH estimation. Here we present two solutions. The first is based on the parallax between data retrieved from geostationary (SEVIRI, HRV band; 1000 m spatial resolution) and polar orbiting satellites (MODIS, band 1; 250 m spatial resolution). The procedure works well if the data from both satellites are retrieved nearly simultaneously. However, MODIS does not retrieve the data at exactly the same time as SEVIRI. To compensate for advection in the atmosphere we use two sequential SEVIRI images (one before and one after the MODIS retrieval) and interpolate the cloud position from SEVIRI data to the time of MODIS retrieval. CTH is then estimated by intersection of corresponding lines-of-view from MODIS and interpolated SEVIRI data. The second method is based on NASA program Crew Earth observations from the International Space Station (ISS). The ISS has a lower orbit than most operational satellites, resulting in a shorter minimal time between two images, which is needed to produce a suitable parallax. In addition, images made by the ISS crew are taken by a full frame sensor and not a push broom scanner that most operational satellites use. Such data make it possible to observe also short time evolution of clouds.

<p>The geodynamical side of explosive volcanic eruption modelling on the one hand, as well as the petrological one on the other, have reached a high degree of sophistication and maturity independently from each other over the years. Unfortunately, adherents of one discipline often only utilize the other’s tools in a simplified and makeshift way, obscuring the full potential of their synergies. Over the past decade efforts have been made to re-integrate both approaches to the issue into a more holistic view on the sub-surface processes leading to and concurrent with explosive volcanism.<br>One of the difficulties encountered in that effort are conceptual and technical incompatibilities between thermo- and fluid-dynamic modelling toolboxes. While the tools perform well individually, they are often not suitable to work in combination in highly complex numerical models, due to interface problems impeding performance.<br>For an ongoing numerical study on transport processes within a volcanic conduit, it has been deemed necessary to re-implement an established thermodynamic model based on Holland and Powell (2011, and follow-ups) in order to a) attain the required computing performance and b) to gain sufficient petrological insight (starting from a geophysical point of view) to be able to make apt use of the tool then at hand.<br>The path to the intermediate goal of deriving the thermodynamic and transport properties (e.g. density, viscosity, heat capacity and conductivity) in a <em>self-consistent and stable</em> manner suitable for further use in a numerical fluid-dynamics model is illustrated here. The focus is on problems encountered with the petrological modelling, and on the subsequent derivation of the above properties, that are not directly available from the former results. The methods presented are general and applicable to various settings regarding volcanic chemistry and transport processes, however, they will be demonstrated on low-viscosity open-conduit systems typical for strombolian activity.</p>