Stochastic sandpile model on small-world networks: Scaling and crossover

Abstract A dissipative stochastic sandpile model is constructed on one and two dimensional small-world networks with shortcut density ϕ , ϕ = 0 represents a regular lattice whereas ϕ = 1 represents a random network. The effect of the transformation of the regular lattice to a small-world network on the critical behaviour of the model as well as the role of dimensionality of the underlying regular lattice are explored studying different geometrical properties of the avalanches as a function of avalanche size s in the small-world regime ( 2 − 12 ≤ ϕ ≤ 0 . 1 ). For both the dimensions, three regions of s , separated by two crossover sizes s 1 and s 2 ( s 1 s 2 ), are identified analysing the scaling behaviour of average height and area of the toppling surface associated with an avalanche. It is found that avalanches of size s s 1 are compact and follow Manna scaling on the regular lattice whereas the avalanches with size s > s 1 are sparse as they are on network and follow mean-field scaling. Coexistence of different scaling forms in the small-world regime leads to violation of usual finite-size scaling which were valid on the regular lattice as well as on the random network independently. Simultaneous appearance of multiple scaling forms are characterized by developing a coexistence scaling theory. As SWN evolves from regular lattice to random network, a crossover from diffusive to super-diffusive nature of sand transport is observed and scaling forms of such crossover is developed and verified.