Flow colocation quotient: Measuring bivariate spatial association for flow data

Bivariate association analyses of spatial flows can reveal the spatial dependence of two types of spatial flows. However, existing studies involving the detection of the bivariate associations of spatial flows focused on two distinct populations without considering the effect of joint distribution patterns. In this paper, the authors propose the flow colocation quotient (FCLQ), which is extended from the colocation quotient, to measure the bivariate association between categories of spatial flows considering the joint distribution. The authors developed two versions of the FCLQ: the global FCLQ, which is applied to measure the overall spatial association pattern, and the local FCLQ, which is used to identify the spatial heterogeneity of spatial association. To further test the statistical significance of the FCLQ values, the authors perform a Monte Carlo simulation under the null hypothesis with random labeling. Six synthetic datasets with different preset patterns are applied to verify the effectiveness of the FCLQ approach. A case study of bike-sharing trip data from Xiamen Island demonstrates the usefulness of the FCLQ in comparative analyses of three bike sharing platforms.

Language

  • English

Media Info

Subject/Index Terms

Filing Info

  • Accession Number: 01877721
  • Record Type: Publication
  • Files: TRIS
  • Created Date: Mar 28 2023 9:56AM