Estimation of flow duration of basaltic magma in fissures

Abstract

We suggest a relationship between the thickness of chilled margins often observed in basaltic dikes and magma flow duration. For this purpose, a new thermal model of cooling basaltic dikes is proposed which takes into account flow duration in a fissure before solidification of the magma. The thermal problem is solved by adding a boundary condition (serving as a substitute for magma flow) to the classical initial conditions for dikes of Fourier's equation. We assume a bound temperature T o at the center of the dike as long as magma flows. At the beginning of the cooling process, the new model gives the same results as the classical one and, consequently, the same chilled margin thicknesses. But later on, when the classical model enforces temperature decreases, our temperatures increase due to the heat brought by magma flow. The chilled margin is reheated and partially or fully destroyed. Petrological observations on the chilled margins of dikes from Salagou and Bédarieux (South of France) and their country rock support this theoretical study. The chilled margins are always thinner than 10 mm; glass includes rare true phenocrysts (mainly olivine and Fe-Ti oxides) and branching quench crystallites of clinopyroxene. The conditions of formation and evolution of the chilled margins have been specified from experimental data: cooling rate = 800° C h −1; glass transition temperature ( T M) = 675° C and glass deformation temperature ( T D) = 705° C. The initial chilled margins predicted by the two models are always thicker than those actually observed; there is evidence for reduction due to reheating by devitrification and softening. In our new model, the chilled margin thickness which remains is directly dependent on the flow duration. For example, for a dike of Salagou (1-m wide and a chilled margin of 5 mm), a flow duration of six days has been computed. At Bédarieux, for a 1-m-wide dike and a thicker chilled margin (8.5 mm), flow duration was 4.6 days. These values must be regarded as upper limits because the effects of the latent heat on the temperature distribution during reheating have not been taken into account. The metamorphism of the country rock (pelites and bauxites) has been investigated; temperatures estimated from classical geothermometers are in agreement with those computed by our model and support its validity.