Roughness distribution of multiple hit and long surface diffusion length noise reduced discrete growth models

Conventionally, the universality class of a discrete growth model is identified via the scaling of interface width. This method requires large-scale simulations to minimize finite-size effects on the results. The multiple hit noise reduction techniques (m>1 NRT) and the long surface diffusion length noise reduction techniques (ℓ>1 NRT) have been used to promote the asymptotic behaviors of the growth models. Lately, an alternative method involving comparison of roughness distribution in the steady state has been proposed. In this work, the roughness distribution of the (2+1)-dimensional Das Sarma–Tamborenea (DT), Wolf–Villain (WV), and Larger Curvature (LC) models, with and without NRTs, are calculated in order to investigate effects of the NRTs on the roughness distribution. Additionally, effective growth exponents of the noise reduced (2+1)-dimensional DT, WV and LC models are also calculated. Our results indicate that the NRTs affect the interface width both in the growth and the saturation regimes. In the steady state, the NRTs do not seem to have any impact on the roughness distribution of the DT model, but it significantly changes the roughness distribution of the WV and LC models to the normal distribution curves.