A Moving Neighborhood Kriging Algorithm Using Delaunay-Fixed Distance Neighborhood Selection

Yu-jian DU,De-yun XIAO

DOI: https://doi.org/10.3969/j.issn.1003-0158.2005.02.013

2005-01-01

Abstract:In geostatistics, Kriging is a kind of algorithm that interpolates the given discrete data with spatially correlated errors. As a regional algorithm, neighborhood selection is crucial in the Kriging applications. The principles of neighborhood selection in Kriging are outlined in particular through a case study of temperature field’s calculation. Delaunay triangulation and fixed distance metrics are integrated in neighborhood selection. In the calculation of temperature field, the new algorithm gives higher precision result compared with moving neighborhood algorithm using fixed-distance alone.