Crustal heterogeneities and complexities of fault processes

In this work we study the relation between crustal heterogeneities and complexities in fault processes. The first kind of heterogeneity considered involves the concept of asperity. The presence of an asperity in the hypocentral region of the M = 6.5 earthquake of June 17-th, 2000 in the South Iceland Seismic Zone was invoked to explain the change of seismicity pattern before and after the mainshock: in particular, the spatial distribution of foreshock epicentres trends NW while the strike of the main fault is N 7◦ E and aftershocks trend accordingly; the foreshock depths were typically deeper than average aftershock depths. A model is devised which simulates the presence of an asperity in terms of a spherical inclusion, within a softer elastic medium in a transform domain with a deviatoric stress field imposed at remote distances (compressive NE − SW, tensile NW − SE). An isotropic compressive stress component is induced outside the asperity, in the direction of the compressive stress axis, and a tensile component in the direction of the tensile axis; as a consequence, fluid flow is inhibited in the compressive quadrants while it is favoured in tensile quadrants. Within the asperity the isotropic stress vanishes but the deviatoric stress increases substantially, without any significant change in the principal stress directions. Hydrofracture processes in the tensile quadrants and viscoelastic relaxation at depth may contribute to lower the effective rigidity of the medium surrounding the asperity. According to the present model, foreshocks may be interpreted as induced, close to the brittle-ductile transition, by high pressure fluids migrating upwards within the tensile quadrants; this process increases the deviatoric stress within the asperity which eventually fails, becoming the hypocenter of the mainshock, on the optimally oriented fault plane. In the second part of our work we study the complexities induced in fault processes by the layered structure of the crust. In the first model proposed we study the case in which fault bending takes place in a shallow layer. The problem can be addressed in terms of a deep vertical planar crack, interacting with a shallower inclined planar crack. An asymptotic study of the singular behaviour of the dislocation density at the interface reveals that the density distribution has an algebraic singularity at the interface of degree ω between -1 and 0, depending on the dip angle of the upper crack section and on the rigidity contrast between the two media. From the welded boundary condition at the interface between medium 1 and 2, a stress drop discontinuity condition is obtained which can be fulfilled if the stress drop in the upper medium is lower than required for a planar trough-going surface: as a corollary, a vertically dipping strike-slip fault at depth may cross the interface with a sedimentary layer, provided that the shallower section is suitably inclined (fault "refraction"); this results has important implications for our understanding of the complexity of the fault system in the SISZ; in particular, we may understand the observed offset of secondary surface fractures with respect to the strike direction of the seismic fault. The results of this model also suggest that further fractures can develop in the opposite quadrant and so a second model describing fault branching in the upper layer is proposed. As the previous model, this model can be applied only when the stress drop in the shallow layer is lower than the value prescribed for a vertical planar crack surface. Alternative solutions must be considered if the stress drop in the upper layer is higher than in the other layer, which may be the case when anelastic processes relax deviatoric stress in layer 2. In such a case one through-going crack cannot fulfil the welded boundary conditions and unwelding of the interface may take place. We have solved this problem within the theory of fracture mechanics, employing the boundary element method. The fault terminates against the interface in a T-shaped configuration, whose segments interact among each other: the lateral extent of the unwelded surface can be computed in terms of the main fault parameters and the stress field resulting in the shallower layer can be modelled. A wide stripe of high and nearly uniform shear stress develops above the unwelded surface, whose width is controlled by the lateral extension of unwelding. Secondary shear fractures may then open within this stripe, according to the Coulomb failure criterion, and the depth of open fractures opening in mixed mode may be computed and compared with the well studied fault complexities observed in the field. In absence of the T-shaped decollement structure, stress concentration above the seismic fault would be difficult to reconcile with observations, being much higher and narrower.

Maremoti in bacini limitati

The work of my thesis is focused on the impact of tsunami waves in limited basins. By limited basins I mean here those basins capable of modifying significantly the tsunami signal with respect to the surrounding open sea. Based on this definition, we consider limited basins not only harbours but also straits, channels, seamounts and oceanic shelves. I have considered two different examples, one dealing with the Seychelles Island platform in the Indian Ocean, the second focussing on the Messina Strait and the harbour of the Messina city itself (Italy). The Seychelles platform is differentiated at bathymetric level from the surrounding ocean, with rapid changes from 2 km to 70 meters over short horizontal distances. The study of the platform response to the tsunami propagation is based on the simulation of the mega-event occurred on 26 December 2004. Based on a hypothesis for the earthquake causative fault, the ensuing tsunami has been numerically simulated. I analysed synthetic tide gauge records at several virtual tide gauges aligned along the direction going from the source to the platform. A substantial uniformity of tsunami signals in all calculated open ocean tide-gauge records is observed, while the signals calculated in two points of the Seychelles platform show different features both in terms of amplitude and period of the perturbation. To better understand the content in frequency of different calculated marigrams, a spectral analysis was carried out. In particular the ratio between the calculated tide-gauge records spectrum on the platform and the average tide-gauge records in the open ocean was considered. The main result is that, while in the average spectrum in the open ocean the fundamental peak is related to the source, the platform introduces further peaks linked both to the bathymetric configuration and to coastal geometry. The Messina Strait represents an interesting case because it consists in a sort of a channel open both in the north and in the south and furthermore contains the limited basin of the Messina harbour. In this case the study has been carried out in a different way with respect to the Seychelles case. The basin was forced along a boundary of the computational domain with sinusoidal functions having different periods within the typical tsunami frequencies. The tsunami has been simulated numerically and in particular the tide-gauge records were calculated for every forcing function in different points both externally and internally of the channel and of the Messina harbour. Apart from the tide-gauge records in the source region that almost immediately reach stationarity, all the computed signals in the channel and in the Messina harbour present a transient variable amplitude followed by a stationary part. Based exclusively on this last part, I calculated the amplification curves for each site. I found that the maximum amplification is obtained for forcing periods of approximately 10 minutes.

Comparing Different Approaches for Clustering Categorical Data

There are different ways to do cluster analysis of categorical data in the literature and the choice among them is strongly related to the aim of the researcher, if we do not take into account time and economical constraints. Main approaches for clustering are usually distinguished into model-based and distance-based methods: the former assume that objects belonging to the same class are similar in the sense that their observed values come from the same probability distribution, whose parameters are unknown and need to be estimated; the latter evaluate distances among objects by a defined dissimilarity measure and, basing on it, allocate units to the closest group. In clustering, one may be interested in the classification of similar objects into groups, and one may be interested in finding observations that come from the same true homogeneous distribution. But do both of these aims lead to the same clustering? And how good are clustering methods designed to fulfil one of these aims in terms of the other? In order to answer, two approaches, namely a latent class model (mixture of multinomial distributions) and a partition around medoids one, are evaluated and compared by Adjusted Rand Index, Average Silhouette Width and Pearson-Gamma indexes in a fairly wide simulation study. Simulation outcomes are plotted in bi-dimensional graphs via Multidimensional Scaling; size of points is proportional to the number of points that overlap and different colours are used according to the cluster membership.

Employing Distances in Design-based Spatial Estimation

In the last couple of decades we assisted to a reappraisal of spatial design-based techniques. Usually the spatial information regarding the spatial location of the individuals of a population has been used to develop efficient sampling designs. This thesis aims at offering a new technique for both inference on individual values and global population values able to employ the spatial information available before sampling at estimation level by rewriting a deterministic interpolator under a design-based framework. The achieved point estimator of the individual values is treated both in the case of finite spatial populations and continuous spatial domains, while the theory on the estimator of the population global value covers the finite population case only. A fairly broad simulation study compares the results of the point estimator with the simple random sampling without replacement estimator in predictive form and the kriging, which is the benchmark technique for inference on spatial data. The Monte Carlo experiment is carried out on populations generated according to different superpopulation methods in order to manage different aspects of the spatial structure. The simulation outcomes point out that the proposed point estimator has almost the same behaviour as the kriging predictor regardless of the parameters adopted for generating the populations, especially for low sampling fractions. Moreover, the use of the spatial information improves substantially design-based spatial inference on individual values.

On the Luminous and Dark Matter Distribution in Early-Type Galaxies

The way mass is distributed in galaxies plays a major role in shaping their evolution across cosmic time. The galaxy's total mass is usually determined by tracing the motion of stars in its potential, which can be probed observationally by measuring stellar spectra at different distances from the galactic centre, whose kinematics is used to constrain dynamical models. A class of such models, commonly used to accurately determine the distribution of luminous and dark matter in galaxies, is that of equilibrium models. In this Thesis, a novel approach to the design of equilibrium dynamical models, in which the distribution function is an analytic function of the action integrals, is presented. Axisymmetric and rotating models are used to explain observations of a sample of nearby early-type galaxies in the Calar Alto Legacy Integral Field Area survey. Photometric and spectroscopic data for round and flattened galaxies are well fitted by the models, which are then used to get the galaxies' total mass distribution and orbital anisotropy. The time evolution of massive early-type galaxies is also investigated with numerical models. Their structural properties (mass, size, velocity dispersion) are observed to evolve, on average, with redshift. In particular, they appear to be significantly more compact at higher redshift, at fixed stellar mass, so it is interesting to investigate what drives such evolution. This Thesis focuses on the role played by dark-matter haloes: their mass-size and mass-velocity dispersion correlations evolve similarly to the analogous correlations of ellipticals; at fixed halo mass, the haloes are more compact at higher redshift, similarly to massive galaxies; a simple model, in which all the galaxy's size and velocity-dispersion evolution is due to the cosmological evolution of the underlying halo population, reproduces the observed size and velocity-dispersion of massive compact early-type galaxies up to redshift of about 2.