The NE flank of Stromboli volcano referred to as Sciara del Fuoco (SdF), due to its slope instability and the strombolian
type activity is subject to landslides. Analysis performed on ground displacements measured at the SdF by an automatic
monitoring system referred to as THEODOROS (THEOdolite and Distancemeter Robot Observatory of Stromboli),
have shown that the recorded cumulative probability of displacements are power-law distributed. The volcano flank
seems affected by movements that could be explained using a mechanical model. The proposed model consists of a
slider mass driven-noise placed on an inclined plane with the same incline angle of the SdF (θ=36.8°). The aim of our
study is to investigate the effects of noise on the cumulative size distribution of the proposed model. In particular we
show that appropriately choosing the range of variation for the intensity of a Gaussian distributed noise source with zero
mean it is possible to model the observed displacements. This is an example of noise-induced critical phenomena.
The concept of stochastic resonance introduced the idea that the presence of noise in nonlinear systems may have benefic effects. In this paper different regular topologies of populations of FitzHugh-Nagumo neurons have been investigated with respect to the presence of noise in the network. Each neuron is subjected to an independent source of noise. In these conditions the behavior of the population depend on the connection among the elements. In population of uncoupled neurons the so-called stochastic resonance without tuning was observed. Moreover, we show that globally coupled neurons have increasing response-to-stimulus coherence for increasing values of the coupling strength. In locally coupled neurons the performance depend on the neighborhood radius and in general are higher than in the case of uncoupled neurons.