PLACE: 214 Atanasoff

SPEAKER:  Dan Nordman, Department of Statistics, Iowa State University

TITLE: Optimal spatial subsample size for variance estimation
 

ABSTRACT:

The talk addresses a problem in determining the optimal block (or subsample) size with a subsampling method for spatial processes observed on regular grids.   We describe two subsample-based variance estimators based on overlapping and non-overlapping subsampling schemes. Expansions for the mean square error of each estimator are derived, which yield expressions for the theoretical optimal block size with each subsample type.

 The optimal block size is shown to depend in an intricate way on the geometry of the spatial sampling region as well as on characteristics of the underlying random field. Final expressions for the optimal block size make use of some nontrivial estimates of lattice point counts in translated convex sets. The expressions for the optimal block size are computed for sampling regions of a number of commonly-encountered shapes.
 
 

COFFEE:  8:40 a.m., 104 Snedecor Hall