SEISMIC NOISE MODELING

In general, different classes of noise can be observed in a seismogram . Typical sources of noise are ground roll, multiples, out-of-plane di ffractions and ambient disturbances (i.e. produced by rain, wind, tide s, vibrations, etc.). In many cases, when generating synthetic data fo r test purposes, it is important that realistic noise models can be su perimposed on the data. However, in most studies in the literature, no ise components are synonymous with Gaussian white noise or similar ass umptions. In this paper we therefore propose an extended computational method, which reflects the complex and varied nature of noise in a be tter way.