The present work aims at building a non-stationary random field model for the slip distribution on the rupture plane. The estimates are arrived based on 230 slip fields available in the SRCMOD database. The evaluation is performed by segregating and quantifying the trend and fluctuation components of the field. Here, the trend portion of the slip is extracted by fitting a 2D elliptical Gaussian surface. The remaining fluctuation part is observed to be stationary following a normal distribution. Further, we propose scaling relations as a function of magnitude for all the unknowns in trend and fluctuation part of the field. Furthermore, an additive model to generate a slip field for a given magnitude combining the deterministic trend part and randomly generated fluctuation part is also developed in the study. Ground motion simulations performed with the components of the slip field showed that the trend portion controls the low frequency, and the fluctuation portion influences the high-frequency characteristics of ground motion. The model developed from the study can be used to generate an ensemble of slip fields for ground motion simulations.