Soon after the onset of an eruption, model forecasts of ash dispersal are used to mitigate the hazards to aircraft, infrastructure, and communities downwind. However, it is a significant challenge to constrain the model inputs during an evolving eruption. Here we demonstrate that volcanic lightning may be used in tandem with satellite detection to recognize and quantify changes in eruption style and intensity. Using the eruption of Calbuco volcano in southern Chile on 22 and 23 April 2015, we investigate rates of umbrella cloud expansion from satellite observations, occurrence of lightning, and mapped characteristics of the fall deposits. Our remote sensing analysis gives a total erupted volume that is within uncertainty of the mapped volume (0.56 ± 0.28 km3 bulk). Observations and volcanic plume modeling further suggest that electrical activity was enhanced both by ice formation in the ash clouds >10 km above sea level and development of a low‐level charge layer from ground‐hugging currents.

The 10 day explosive phase of the 2008–2009 eruption of Chaitén volcano, Chile, draped adjacent watersheds with a few cm to >1 m of tephra. Subsequent lava‐dome collapses generated pyroclastic flows that delivered additional sediment. During the waning phase of explosive activity, modest rainfall triggered an extraordinary sediment flush which swiftly aggraded multiple channels by many meters. Ten kilometer from the volcano, Chaitén River channel aggraded 7 m and the river avulsed through a coastal town. That aggradation and delta growth below the abandoned and avulsed channels allow estimates of postdisturbance traction‐load transport rate. On the basis of preeruption bathymetry and remotely sensed measurements of delta‐surface growth, we derived a time series of delta volume. The initial flush from 11 to 14 May 2008 deposited 0.5–1.5 × 106 m3 of sediment at the mouth of Chaitén River. By 26 May, after channel avulsion, a second delta amassed about 2 × 106 m3 of sediment; by late 2011 it amassed about 11 × 106 m3. Accumulated sediment consists of low‐density vesicular pumice and lithic rhyolite sand. Rates of channel aggradation and delta growth, channel width, and an assumed deposit bulk density of 1100–1500 kg m−3 indicate mean traction‐load transport rate just before and shortly after avulsion (∼14–15 May) was very high, possibly as great as several tens of kg s−1 m−1. From October 2008 to December 2011, mean traction‐load transport rate declined from about 7 to 0.4 kg−1 m−1. Despite extraordinary sediment delivery, disturbed channels recovered rapidly (a few years).

High effusion rates of intermediate‐to‐high‐silica lavas seem to be less uncommon than previously thought, in particular during their initial eruptive stages. In this study, we report satellite‐based time‐averaged discharge rates for the 2011–2012 effusive phase at Cordón Caulle, which are well correlated with the evolution of the quasi‐harmonic tremor, the most significant seismic signal after the initial explosive stage. Such a correspondence could become a key method for detection of the onset of effusive phases, especially in remote and/or very cloudy areas, supplying an additional tool for effective warnings and near‐real‐time hazard appraisal.

An innovative 3‐D numerical model for the dynamics of volcanic ballistic projectiles is presented here. The model focuses on ellipsoidal particles and improves previous approaches by considering horizontal wind field, virtual mass forces, and drag forces subjected to variable shape‐dependent drag coefficients. Modeling suggests that the projectile's launch velocity and ejection angle are first‐order parameters influencing ballistic trajectories. The projectile's density and minor radius are second‐order factors, whereas both intermediate and major radii of the projectile are of third order. Comparing output parameters, assuming different input data, highlights the importance of considering a horizontal wind field and variable shape‐dependent drag coefficients in ballistic modeling, which suggests that they should be included in every ballistic model. On the other hand, virtual mass forces should be discarded since they almost do not contribute to ballistic trajectories. Simulation results were used to constrain some crucial input parameters (launch velocity, ejection angle, wind speed, and wind azimuth) of the block that formed the biggest and most distal ballistic impact crater during the 1984–1993 eruptive cycle of Lascar volcano, Northern Chile. Subsequently, up to 106 simulations were performed, whereas nine ejection parameters were defined by a Latin‐hypercube sampling approach. Simulation results were summarized as a quantitative probabilistic hazard map for ballistic projectiles. Transects were also done in order to depict aerial hazard zones based on the same probabilistic procedure. Both maps combined can be used as a hazard prevention tool for ground and aerial transits nearby unresting volcanoes.

This paper introduces an open source computer code to perform an integrated probabilistic spatio-temporal volcanic hazard assessment in distributed volcanic fields. The program, named MatHaz, is a set of Matlab scripts that follows a sequential methodology. After the user has provided a set of input files, this tool first estimates the spatial probability of future volcanic vents, then the temporal probability of future volcanic events, and finally models up to five volcanic phenomena (pyroclastic density currents, ballistic projectiles, lava flows, lahars, and tephra fallout) following a probabilistic approach. These results can be combined and depicted as an integrated quantitative (and/or qualitative) volcanic hazard map, with weightings of hazard factors chosen by the user. We illustrate the use of this tool by applying it to the Carrán-Los Venados Volcanic Field in southern Chile. The opensource, replicable, and user-friendly nature of the code allows its application to any volcanic region of the world, regardless of its extent, type, and amount of volcano-structural data.