A Fluid-filled Fracture as Possible Mechanism of Ground Deformation at Phlegraean Fields, Italy V. BONASIA F. PINGUE R. SCARPA Dipartimento di Geofisica e Vulcanologia, Universit& di Napoli, Italy Osservatorio Vesuviano, Ercolano, Napoli, Italy ABSTRACT Two uplift episodes have recently been recorded at Phlegraean Fields, a volcanic region near Naples (south-central Italy). The first episode occurred in 1970 and lasted up to 1972; the second has begun at the end of 1982 and is still in progress. An attempt to model the ground deforma- tions which occurred during the 1970-1972 event is made in this paper. The source has been assumed to be a two-dimensional fluid- filled fracture, similar to a sill. A good fit with experimental data has been obtained for a short (1-2 km long) shallow (3 km deep) source and a driving pressure ranging from 60 to 210 bars. Rigidity values have been fixed at 3.5- 4.0 × 10 I° c.g.s., with Poisson ratio equal to 0.3. This solution which differs from previous approaches by MOGI (1958) and WALSH and DECKER (1971) is non-unique, but the present results are in good agreement with observed horizontal and vertical displacements. Notwithstanding the oversimplification made in assuming a homogeneous elastic semi-infm- ite medium, the predicted stress field seems to be in agreement with the fault-plane solutions and the pattern of epicenters determined for the uplift-seismic swarm episode that is still in progress. INTRODUCTION Many models have been proposed to explain observed ground deformations in volcanic areas and geothermal systems Bull. VolcanoI., Vol. 47-2, 1984 (e.g. MOGI, 1958; SUN, 1969; WALSH and DECKER, 1971; DIETERICH and DECKER, 1975; POLLARD and HOLZHAUSEN, 1979; DAVIS, 1983, RYAN et at., 1983). The problem is highly non unique and some constraints based on other geophysical parameters and field observations are necessary to fmd aceptable solutions. In fact, several configurations of dislocation densities may correspond to the same deformations as measured at ground surface. Among the various models proposed in the literature the most popular one is certainly the Mogi-Yamakawa approach (MOGI, 1958). This model provides means to evaluate ground movements at the free surface due to a buried strain nucleus. The medium is an elastic, isotropic half- space and the pressure within the source is assumed uniform, while the source radius is much smaller than its distance to the surface. Field observations on sills, laccoliths and dikes lead WALSH and DECKER (1971) and, later, DIETERICH and DECKER (1975) to consider line sources or more complicated axi-symmetrical bodies to explain inflation or deflation episodes by assuming uniform displacements of the walls of magmatic chambers in response to pressure changes. One of the major shortcomings of these elastic models is that they predict quite similar vertical deformations in the near-field, and they do not explain the observed differences between vertical and horizontal deforma-