Title:Correlation functions in the two-dimensional random-field Ising model

Abstract: Transfer-matrix methods are used to study the probability distributions of spin-spin correlation functions $G$ in the two-dimensional random-field Ising model, on long strips of width $L = 3 - 15$ sites, for binary field distributions at generic distance $R$, temperature $T$ and field intensity $h_0$. For moderately high $T$, and $h_0$ of the order of magnitude used in most experiments, the distributions are singly-peaked, though rather asymmetric. For low temperatures the single-peaked shape deteriorates, crossing over towards a double-$\delta$ ground-state structure. A connection is obtained between the probability distribution for correlation functions and the underlying distribution of accumulated field fluctuations. Analytical expressions are in good agreement with numerical results for $R/L \gtrsim 1$, low $T$, $h_0$ not too small, and near G=1. From a finite-size {\it ansatz} at $T=T_c (h_0=0)$, $h_0 \to 0$, averaged correlation functions are predicted to scale with $L^y h_0$, $y =7/8$. From numerical data we estimate y=0.875 \pm 0.025$, in excellent agreement with theory. In the same region, the RMS relative width $W$ of the probability distributions varies for fixed $R/L=1$ as $W \sim h_0^{\kappa} f(L h_0^u)$ with $\kappa \simeq 0.45$, $u \simeq 0.8$ ; $f(x)$ appears to saturate when $x \to \infty$, thus implying $W \sim h_0^{\kappa}$ in $d=2$.