A kriging interpolation model for geographical flows

Abstract

The kriging model can accommodate various spatial supports and has been extensively applied in hydrology, meteorology, soil science, and other domains. With the expansion of applications, it is essential to extend the kriging model for new spatial support of high-dimensional data. Geographical flows can depict the movements of geographical objects and imply the underlying mobility patterns in geographical phenomena. However, due to the bias, sparsity, and uneven quality of flow data in the real world, research about flows remains hindered by the lack of complete flow data and effective flow interpolation methods. In this study, we design a kriging interpolation model for flows based on several flow-related concepts and the autocorrelation of flows. We also analyze the second-order stationarity and anisotropy in the flow spatial random field. To illustrate the effectiveness and applicability of our method, we conduct two case studies. The former case study compares several experiments of flow density interpolation using Beijing mobile signaling data and illustrates the conditions of applicable areas. The latter case study extends our model to other flow attributes, such as travel time uncertainty, using Beijing taxi origin-destination flow data. The results of these cases demonstrate the effectiveness and high accuracy of our model.

Keywords:

• Flow space
• flow kriging interpolation method
• origin-destination (OD) flow
• second-order stationarity
• travel time uncertainty

Acknowledgments

The authors thank the editor and the anonymous reviewers for their helpful comments on an earlier draft of this paper.

Disclosure statement

No potential conflict of interest was reported by the author(s).

Data and codes availability statement

The data and codes that support the findings of this study are available in 'figshare.com’ with the identifier(s): https://doi.org/10.6084/m9.figshare.21322401.v10.

Additional information

Funding

This work was supported by the National Natural Science Foundation of China (Grant No. 42071436, 42071435) and the Innovation Project of LREIS (Grant No. KPI002, YPI006).

Notes on contributors

Ya Fang

Ya Fang is a Master’s candidate at the Institute of Geographical Sciences and Natural Resources Research, CAS. She designed the study, performed the Kriging interpolation experiment, and drafted the manuscript.

Tao Pei

Tao Pei is a Professor at the Institute of Geographical Sciences and Natural Resources Research, CAS. He conceived of the study, participated in its design, and conduct to draft the manuscript.

Ci Song

Ci Song is an Associate Professor at the Institute of Geographical Sciences and Natural Resources Research, CAS. He conceived of the study, participated in its design, and conduct to draft the manuscript.

Jie Chen

Jie Chen is an Associate Professor at the Institute of Geographical Sciences and Natural Resources Research, CAS. She participated in the investigation of the study region and result analyses.

Xi Wang

Xi Wang is a Doctoral candidate at the Institute of Geographical Sciences and Natural Resources Research, CAS. He preprocessed the mobile phone data and taxi OD flow data.

Xiao Chen

Xiao Chen is a Doctoral candidate at the Institute of Geographical Sciences and Natural Resources Research, CAS. She participated in the generation of experimental data.

Yaxi Liu

Yaxi Liu is a Doctoral candidate at the Institute of Geographical Sciences and Natural Resources Research, CAS. He participated in the extraction and validation of OD flow data.