ALBERT

Notes: The study of the mechanics of caldera formation can yield valuable insight into the behaviour of near-surface magma chambers and their mechanical interaction with the crust. We have addressed the problem of physical processes responsible of long-term ground deformation from the emplacement of magma in the crust to a final resurgent stage. A mechanical model based on a thermally coupled elastovisco-plastic rheology and a finite deformation formulation has been used, and solved via a finite element approximation. The history of an ideal caldera is modelled in the following two stages.〈list xml:id="l1" style="custom"〉1A doming phase corresponding to the growth of the magma chamber. This stage is constrained by values of amplitude, half-width and time constant of regional doming and by a range of realistic over-pressures. This uplift is linked to the plastic strain field within the crust, which strongly depends upon the pressure and temperature fields.1A resurgence phase which results from a collapse stage, with an instantaneous change in shape of the topography. The magma chamber is passive and this stage is basically similar to a crater relaxation process. The rebound amplitude and half-width are controlled by the length-scale of the collapse. A gravitational rebound is a very likely mechanism explaining large caldera uplift without over-pressure mechanisms.For most intracrustal loading processes, the crustal deformation is critically influenced by the relative importance of the upper crustal rheology (pressure-dependent plasticity) versus the lower crustal rheology (temperature-dependent viscosity). Because the upper part of the crust can deform without time-dependent dissipation, it may be responsible for large uplift variations that can occur for moderate loading variations. By contrast the creep of the middle crust acts as a regulating medium.

Notes: We have conducted a palaeomagnetic study of a Permian red sandstone layer intruded by a 70 cm wide Plio-Pleistocene basaltic dike, with the ultimate goal of estimating the flow duration of the magma before its solidification. 31 samples were collected at various distances from the dike/sandstone interface and were thermally demagnetized in order to separate the overprinted Plio-Pleistocene magnetization component from the primary Permian one and to determine for each sample the unblocking temperature, Tub, necessary to erase totally the overprint. This temperature can be used to determine the maximum palaeotemperature reached at any distance from the contact, provided the secondary magnetization is a (P)TRM and the thermal fluctuation effects and actual cooling rates are taken into account.Examination of various magnetic parameters (NRM intensity, 15 days VRM intensity, initial magnetic susceptibility and hysteresis curves) allowed us to establish the nature of the overprinted magnetization and consequently to exclude samples carrying a CRM overprint. An apparent palaeotemperature profile and two corrected ones were thus deduced from the samples in which the dike emplacement produced no crystallization of new magnetic minerals.These profiles were compared with the calculations of a new conductive thermal model which includes the basalt's latent heat of crystallization, the flow duration of the dike, the width of the circulating magma, and the temperature dependence of the thermal conductivity, k, in the basalt and in the sandstone, or alternatively convection simulated by an increase of the k values.Good fits with the experimental data are obtained for a flowing width consistent with field observations and when the latent heat is taken into account, in the two following cases. First, when k is temperature dependent in the basalt and in the sandstone, which implies a magma flow duration of 5 days. Second, when a limited convection process is assumed in the sandstone only; here, a flowing time of only 1 day is required.The corrected palaeotemperature profiles argue in favour of the second hypothesis, suggesting that when a dike is emplaced, heat transfer by fluids may play an important role in the surrounding rock.