Critical phases in the raise and peel model

TL;DR

This paper investigates the critical behavior of the raise and peel model (RPM), revealing a conformally invariant point at u=1, a roughness transition, and two distinct critical phases for u>1 with different surface profile properties.

Contribution

The study provides a detailed analysis of the RPM's phase diagram, identifying a conformal invariance point, a roughness transition, and the existence of two separate critical phases for u>1, with insights into their universality classes.

Findings

At u=1, RPM is conformally invariant with critical exponents z=1 and α=0.

For u>1, RPM exhibits two critical phases with different roughness exponents.

At short scales, RPM shows KPZ universality behavior, but differs at large scales.

Abstract

The raise and peel model (RPM) is a nonlocal stochastic model describing the space and time fluctuations of an evolving one dimensional interface. Its relevant parameter $u$ is the ratio between the rates of local adsorption and nonlocal desorption processes (avalanches) processes. The model at $u=1$ give us the first example of a conformally invariant stochastic model. For small values $u<u_0$ the model is known to be noncritical, while for $u>u_0$ it is critical. By calculating the structure function of the height profiles in the reciprocal space we confirm with good precision that indeed $u_0=1$. We establish that at the conformal invariant point $u=1$ the RPM has a roughness transition with dynamical and roughness critical exponents $z=1$ and $\alpha=0$, respectively. For $u>1$ the model is critical with an $u$-dependent dynamical critical exponent $z(u)$ that tends towards zero as $u\to \infty$. However at $1/u=0$ the RPM is exactly mapped into the totally asymmetric exclusion problem (TASEP). This last model is known to be noncritical (critical) for open (periodic) boundary conditions. Our studies indicate that the RPM as $u \to \infty$, due to its nonlocal dynamics processes, has the same large-distance physics no matter what boundary condition we chose. For $u>1$, our analysis show that differently from previous predictions, the region is composed by two distinct critical phases. For $u\leq u < u_c\approx 40$ the height profiles are rough ($\alpha = \alpha(u) >0$), and for $u>u_c$ the height profiles are flat at large distances ($\alpha = \alpha(u) <0$). We also observed that in both critical phases ($u>1$) the RPM at short length scales, has an effective behavior in the Kardar-Parisi-Zhang (KPZ) critical universality class, that is not the true behavior of the system at large length scales.