Aurelie Germa

Relationship between dike and volcanic conduit distribution in a highly eroded monogenetic volcanic field: San Rafael, Utah, USA

We mapped 63 conduits, ∼2000 dike segments, and 12 sills in the San Rafael subvolcanic field, Utah (United States), where this Pliocene magmatic system is eroded to a depth of ∼0.8 km and is exceptionally well exposed. Although the number of mapped conduits, dikes, and sills might represent minimums, depending on the level of erosion and exposure, mapped dikes are more numerous around the areally extensive sills and interact with sills and conduits in complex ways. We analyze conduit distribution using kernel density methods and compare results with dike and sill distribution. We find that the distribution of conduits matches the major features of dike distribution, including development of clusters and distribution of outliers. These statistical models are then applied to the distributions of volcanoes in several recently active volcanic fields, where intrusion distributions must be inferred from very sparse data, and compared with San Rafael conduit distribution. This comparison supports the use of statistical models in probabilistic hazard assessment for distributed volcanism. Specifically, renewed dike intrusion and potential eruptions in active basaltic systems can be assessed probabilistically from the distribution of older volcanoes in distributed volcanic systems.

Construction and destruction of Mont Pelée volcano: Volumes and rates constrained from a geomorphological model of evolution

This study presents long-term volumes and construction rates for the Mont Conil-Mont Pelee volcano and rate estimates at which volcanic activity creates relief. An algorithm, ShapeVolc, is used to numerically model topographic surfaces. Volcano morphology is analyzed using current digital elevation model in combination with mapped geology to produce 10 paleotopographies at the end of four constructional stages and three destructional events. Volumes of each constructional stage were estimated at about 35.2 km3, 26.2 km3, 8.3 km3, and 2.5 km3 for a total cumulative erupted volume of 72.2 km3. We estimate that Mont Pelee accounted for about 10% of the Lesser Antilles arc production in the last 100 kyr. The volcano has been built at an average rate of 0.13 km3/kyr during the last 550 kyr. During that time, construction rates varied by a factor of 15, from 0.04 km3/kyr in early stages up to 0.52 km3/kyr after the second flank collapse. Volumes displaced by each flank collapse were estimated at 14.7 km3, 8.8 km3, and 3.5 km3, thus about 37% of the total constructed volume. Integrated over the volcanos lifetime, the rate at which flank collapses removed material off the island is 0.15 km3/kyr. In contrast, long-term erosion rates outside collapsed areas are estimated at about 0.05 ± 0.7 km3/kyr, or ~11 km3 of material removed. This latter rate is not negligible, which strengthens the importance of taking into account recurrent small erosional events on the geomorphological evolution of a volcanic island in a tropical context.

A Geophysical Model for the Origin of Volcano Vent Clusters in a Colorado Plateau Volcanic Field

Variation in spatial density of Quaternary volcanic vents, and the occurrence of vent clusters, correlate with boundaries in Proterozoic crust in the Springerville volcanic field (SVF), Arizona, USA. Inverse modeling using 538 gravity measurements shows that vent clusters correlate with gradients in the gravity field due to lateral variation in crustal density. These lateral discontinuities in the crustal density can be explained by boundaries in the North American crust formed during Proterozoic accretion. Spatial density of volcanic vents is low in regions of high-density Proterozoic crust, high in areas of relatively low-density Proterozoic crust, and is greatest adjacent to crustal boundaries. Vent alignments parallel these boundaries. We have developed 2D and 3D numerical models of magma ascent through the crust to simulate long-term, average magma migration that led to the development of vent clusters in the SVF, assuming that a viscous fluid flow through a porous media is statistically equivalent to magma migration averaged over geological time in the full field scale. The location and flux from the uniform magma source region are boundary conditions of the model. Changes in model diffusivity, associated with changes in the bulk properties of the lithosphere, can simulate preferential magma migration paths and alter estimated magma flux at the surface, implying that large-scale crustal structures, such as inherited tectonic block boundaries, influence magma ascent and clustering of volcanic vents. Probabilistic models of volcanic hazard for distributed volcanic fields can be improved by identifying crustal structures and assessing their impact on volcano distribution with the use of numerical models.