Scaling transition for long-range dependent Gaussian random fields

Abstract: In Puplinskaite and Surgailis (2014) we introduced the notion of scaling transition for stationary random fields $X$ on $\mathbb{Z}^2$ in terms of partial sums limits, or scaling limits, of $X$ over rectangles whose sides grow at possibly different rate. The present paper establishes the existence of scaling transition for a natural class of stationary Gaussian random fields on $\mathbb{Z}^2$ with long-range dependence. The scaling limits of such random fields are identified and characterized by dependence properties of rectangular increments.