Extensive studies on linear growth processes with spatiotemporally correlated noise in arbitrary substrate dimensions

An extensive analytical and numerical study on a class of growth processes with spatiotemporally correlated noise in arbitrary dimension is undertaken. In addition to the conventional investigation on the interface morphology and interfacial widths, we pay special attention to exploring the characteristics of the slope-slope correlation function S(r,t) and the [Q]-th degree residual local interfacial width w[Q](l,t), whose importance has been somewhat overlooked in the literature. Based on the above analysis, we give a plausible theoretical explanation about the various experimental observations of kinetically and thermodynamically unstable surface growth. Furthermore, through explicit examples, we show that the statistical methods of calculating the exponents (including the dynamic exponent z, the global roughness exponent α, and the local roughness exponent α(loc)), based on the scaling of S(r,t) and w[Q](l,t), are very reliable and rarely influenced by the finite time and/or finite-size effects. Another important issue we focus on in this paper is related to numerical calculation. For the specific class of growth processes discussed in this paper, we develop a very efficient and accurate algorithm for numerical calculation of the dynamics of interface configuration, the structure factor, the various correlation functions, the interfacial width and its variants in arbitrary dimensions, even with very large system size and very late time. The proposed systematical algorithm can be easily generalized to other linear processes and some special nonlinear processes.