Influence of noise-reduction schemes on statistics of the diffusion-limited aggregation model at meso- and macroscales

This paper is devoted to the problem of noise reduction in the diffusion-limited aggregation (DLA) model. We investigate the influence of multiple-hit averaging on statistical properties of DLA clusters at meso- and macroscopic length scales. For this purpose, we analyze the long-range correlation of the symmetry broken at the microscopic level via the introduction of a chirality to the original Witten-Sander DLA algorithm. Instead of the usual circular growth units, we consider them to consist of chiral (i.e., without inversion symmetry) combinations of microblocks. Extensive Monte Carlo simulations performed for off- and on-square-lattice conditions reveal the asymptotically achiral behavior of the model, which also occurs when one proceeds to noise-reducing DLA modifications such as multiple-hit averaging with and without erasing. This allows us to conclude that neither of the noise-reducing schemes studied breaks the original balance between deterministic and stochastic forces governing the long-range statistics of the DLA model.