Gravity structure of Piton de la Fournaise volcano and inferred mass transfer during the 2007 crisis

Abstract

The structure and mass transfers at Piton de la Fournaise volcano are investigated using new on land and offshore measurements. The structures are defined using 2 and 3D modelling. Short-wavelength positive anomalies to the W and NW of the central area are interpreted as shallow piles of thick, dense lava flows which have filled volcano-tectonic depressions, several hundreds of meters deep. This finding provides a new insight into the evolution of the volcano. Conversely, the currently active central zone is associated with a negative short-wavelength anomaly, explained by the presence of thin, highly vesiculated and fractured lava flows building the Central Cone. Low density formations underlying the Rivières des Remparts and Langevin areas are tentatively interpreted as breccias related to erosion and/or to southward directed mass-wasting. Topographic offshore highs in the continuation of the SE and NE rift zones of Piton de la Fournaise are characterized by negative gravity anomalies and reversed magnetic anomalies. Their interpretation as hyaloclastites and pillow lavas provides a coherent explanation for both types of signals if we admit that they were primarily constructed during the Alizés stage. Two main deeper, dense structures are also found: the Grand Brûlé complex, and a complex beneath the Plaine des Sables and the Enclos Fouqué. Both are identified as hypovolcanic intrusive complexes. The Grand Brûlé complex (10 × 18 km in plan view) shows a marked NS-elongated geometry, unlike the other comparable complexes on the island. The Plaine des Sables-Enclos Fouqué complex, a more or less circular structure 5 km in diameter, is attributed to the Ancient Shield stage. It is centered on the Enclos fault, suggesting that the paleo-center of the Ancient Shield was located about 1.5 km eastward of the previously proposed location. An apparent contradiction was noted between the seismic and gravity images of the internal structure. In particular, a high velocity cylinder beneath the summit was not detected in the gravity models. This contradiction was resolved by considering the relative sensitivity of each method. Finally, the gravity dataset has allowed us to explore the gravity changes that accompanied the major eruptive and volcano-tectonic crisis of April 2007 (350 m collapse of the summit and emission of ∼ 120 × 10 6 m 3 of lava). The change in the mass distribution of the edifice was large enough to be detected in the survey data collected in the central area before and after the crisis. However, the accuracy was too low for the precise assessment of the mass transfers.