Premium
Interpolating Spatial Interaction Data 1
Author(s) -
Jang Woo,
Yao Xiaobai
Publication year - 2011
Publication title -
transactions in gis
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 0.721
H-Index - 63
eISSN - 1467-9671
pISSN - 1361-1682
DOI - 10.1111/j.1467-9671.2011.01273.x
Subject(s) - interpolation (computer graphics) , zoning , multivariate interpolation , computer science , line (geometry) , spatial analysis , nearest neighbor interpolation , bilinear interpolation , flow (mathematics) , algorithm , trilinear interpolation , data mining , geography , mathematics , artificial intelligence , remote sensing , computer vision , geometry , engineering , motion (physics) , civil engineering
Spatial interpolation has been widely used to improve the spatial granularity of data, or to mediate between inconsistent zoning schemes of spatial data. Traditional areal interpolation methods translate values of source zones to those of target zones. These methods have difficulty in dealing with flow data, as each instance is associated with a pair of zones. This study develops a new concept, flow line interpolation, to fill the abovementioned gap. We also develop a first flow line interpolation method to estimate commuting flow data between spatial units in a target zoning scheme based on such data in a source zoning scheme. Three models (i.e., areal‐weighted, intelligent, and gravity‐type flow line interpolation) are presented. To test the estimation accuracy and the application potential of these models, a case study of Fulton County in Georgia is conducted. The results reveal that both the areal‐weighted and intelligent models are very promising flow line interpolation methods.