Highlighting the buoyancy effects on surface deformation above an inflating magma chamber from analogue experiments

At active volcanoes, observed surface deformation results from the complex interaction between the magma and the host rock in the magmatic plumbing system. One common source of deformation is magmatic recharge, resulting in pressurisation of the magma chamber. All active systems, from the most basaltic to the most silicic magmatic composition, are subject to magma chamber replenishment. The majority of numerical models used to invert surface deformation focus on the effect of over-pressure and neglect the effect of the fluid buoyancy. In the case of silicic magma, the buoyancy is positive and can further be increased by the gas exsolution that occurs at shallow depths. Basaltic magma, on the other hand, has a negative buoyancy which is measurable by gravimetry. We present here experiments investigating the effect of buoyancy on surface deformation. Surface deformation, the shear strain pattern and the chamber overpressure are measured throughout the injection of liquid at constant volumetric flux. Then, we use the McTigue (1987) model to predict the surface displacement from the measured overpressure in the chamber, and conversely. We show that predictions are 7% below the observation when the liquid buoyancy is positive (ẟρ =-81 kg⋅m^-3) and 9% above it when the liquid buoyancy is negative (ẟρ =-157 kg⋅m^-3). Even if the effect of buoyancy is small, this highlights the possible error made on source overpressure when inverting surface deformation. This call to a careful consideration of the geological context in unrest period at active systems when volume change needs to be precisely estimated.