Intrinsic Autoregressions and Related Models on the Two-Dimensional Lattice
将二维格点上的平稳自回归推广到内在模型（仅假设增量平稳），推导预测公式和半变异函数的渐近行为，并提出近似极大似然估计方法。
Stationary autoregressions on a two-dimensional lattice are generalized to intrinsic models where only increments are assumed to be stationary.Prediction formulae and the asymptotic behaviour of the semivariogram are derived.For parameter estimation we propose an approximate maximum likelihood estimator, a generalization of Whittle's estimator; it is derived also for general intrinsic models.