Survival Probability in a Random Velocity Field

Abstract

The time dependence of the survival probability, S(t), is determined for diffusing particles in two dimensions which are also driven by a random unidirectional zero-mean velocity field, v_x(y). For a semi-infinite system with unbounded y and x>0, and with particle absorption at x=0, a qualitative argument is presented which indicates that S(t)~t^{-1/4}. This prediction is supported by numerical simulations. A heuristic argument is also given which suggests that the longitudinal probability distribution of the surviving particles has the scaling form P(x,t)~ t^{-1}u^{1/3}g(u). Here the scaling variable u is proportional to x/t^{3/4}, so that the overall time dependence of P(x,t) is proportional to t^{-5/4}, and the scaling function g(u) has the limiting dependences g(u) approaching a constant as u--->0 and g(u)~exp(-u^{4/3}) as u--->infinity. This argument also suggests an effective continuum equation of motion for the infinite system which reproduces the correct asymptotic longitudinal probability distribution.

6 pages, RevTeX, 5 figures includes, to be submitted to Phys. Rev. E

References in corpus (1)

• Life and Death at the Edge of a Windy Cliff

Cited by in corpus (12)

• Persistence and First-Passage Properties in Non-equilibrium Systems
• Persistence of a particle in the Matheron-de Marsily velocity field
• First-Passage Duality
• Persistence of Randomly Coupled Fluctuating Interfaces
• Tracer Diffusion on a Crowded Random Manhattan Lattice
• Persistence of a Rouse polymer chain under transverse shear flow
• Survival of a Diffusing Particle in a Transverse Shear Flow: A First-Passage Problem with Continuously Varying Persistence Exponent
• Persistence of fractional Brownian motion with moving boundaries and applications
• Persistence probabilities of two-sided (integrated) sums of correlated stationary Gaussian sequences
• Motion of a random walker in a quenched power law correlated velocity field
• Persistence probabilities for stationary increment processes
• On the survival probability in the Matheron - De Marsily model