Comment on "Restricted curvature model with suppression of extremal height"

Recently Jeong and Kim [Phys. Rev. E 66, 051605 (2002)] investigated the scaling properties of equilibrium self-flattening surfaces subject to a restricted curvature constraint. In one dimension (1D), they found numerically that the stationary roughness exponent alpha approximately 0.561 and the window exponent delta approximately 0.423. We present an analytic argument for general self-flattening surfaces in D dimensions, leading to alpha=Dalpha(0)/(D+alpha(0)) and delta=D/(D+alpha(0)), where alpha(0) is the roughness exponent for equilibrium surfaces without the self-flattening mechanism. In case of surfaces subject to a restricted curvature constraint, it is known exactly that alpha(0)=3/2 in 1D, which leads to alpha=3/5 and delta=2/5. Small discrepancies between our analytic values and their numerical values may be attributed to finite size effects.