حل مسئله معكوس در چهارچوب بيزين بر روي ميدان جابجايي حاصل از تداخل سنجي راداري با دريچه مصنوعي در زلزله قشم 2008

Bayesian inversion of InSAR observations for 2008 Qeshm earthquake

Synthetic Aperture Radar has proven to be a powerful geodetic technique for measuring surface displacement with cm-scale accuracy. In this paper, interferometric observations associated with the September 10, 2008 Qeshm earthquake in Iran obtained by C-band Envisat and L-band ALOS are used to extract the displacement field and model the source parameters of earthquakes. By solving the inversion problem we resolve the source parameters of displacement field. The inversion is done in a two-step process. First we derive, using the Simulated Annealing (SA) method, the posterior marginal distribution of parameters to determine their sensitivity to the observations. This step is important as it determines which parameters can be resolved reliably by the inversion of geodetic data and which ones need to be constrained by other sources. Following the above procedure, we use a Genetic Algorithm (GA) to find the maximum Bayesian probability for each parameter. Bayeʹs theorem provides the possibility of the simultaneous use of observations and prior information in that a probability level is assigned to a solution according to its compatibility with observations and prior information. Thus, using an optimization algorithm such as a genetic algorithm we obtain the most probable solution in the Bayesian framework.