Kriging and Semivariogram Deconvolution in the Presence of Irregular Geographical Units

摘要: This paper presents a methodology to conduct geostatistical variography and interpolation on areal data measured over geographical units (or blocks) with different sizes shapes, while accounting for heterogeneous weight or kernel functions within those units. The deconvolution method is iterative seeks the point-support model that minimizes difference between theoretically regularized semivariogram fitted data. then used in area-to-point (ATP) kriging map spatial distribution of attribute interest each unit. coherence constraint ensures weighted average kriged estimates equals datum. This approach illustrated using health (cancer rates aggregated at county level) population density surface as function. Simulations are conducted two regions contrasting geographies: state Indiana four states Western United States. In both regions, yields point support reasonably close simulated values. use this ATP more accurate prediction than naive simply collapses into its geographic centroid. reduces smoothing effect robust respect small differences model. Important features semivariogram, such nugget effect, can never be fully validated from user may want narrow down set solutions based his knowledge phenomenon (e.g., zero). presented avoids visual bias associated interpretation choropleth maps should facilitate analysis relationships variables supports.