Traversa P., Pinel Virginie, Grasso J. R. (2010). A constant influx model for dike propagation : implications for magma reservoir dynamics. Journal of Geophysical Research-Solid Earth, 115, p. B01201. ISSN 0148-0227.
Titre du document
A constant influx model for dike propagation : implications for magma reservoir dynamics
Journal of Geophysical Research-Solid Earth, 2010,
115, p. B01201 ISSN 0148-0227
Most observations of seismicity rate during dike propagation on basaltic volcanoes show (1) rate stationarity despite possible variations of the dike tip velocity, (2) frequent lack of clear and monotonic hypocenter migration following dike propagation, and (3) event occurrences located backward with respect to the dike tip position. On these bases, the origin of the seismicity contemporary to dike intrusion within basaltic volcanoes cannot be solely related to the crack tip propagation. Seismicity rather appears to be the response of the edifice itself to the volumetric deformation induced by the magma intruding the solid matrix. The volume change induced into the volcano edifice over time by the intruding magma is equal to the magma flux injected into the dike from the reservoir. The consequence of this is that the stationary seismicity rate observed during the intrusion is a proxy for the magma flux withdrawn from the reservoir. We consider a two-phase dike propagation model, including a first vertical propagation followed by a lateral migration along a lithological discontinuity. We explore (1) under which geophysical conditions the vertical dike is fed at constant flow rate of magma and (2) dike propagation patterns. Implications entailed by constant volumetric flux on the Piton de la Fournaise volcano case study suggest a minimum size for the magma reservoir of about 1 km(3) and a maximum value for the initial magma reservoir overpressure of about 2.2 MPa. Considering similar magma inflow rates during vertical and lateral dike propagation phases, we reproduce independent estimates of propagation velocities, rise times, and injected volumes when applying the model to the August 2003 Piton de la Fournaise eruption.
