The International Archives of the Photogrammetry, Remote Sensing and Spatial Information Sciences
Download
Publications Copernicus
Download
Citation
Articles | Volume XLVIII-4/W1-2022
Int. Arch. Photogramm. Remote Sens. Spatial Inf. Sci., XLVIII-4/W1-2022, 279–286, 2022
https://doi.org/10.5194/isprs-archives-XLVIII-4-W1-2022-279-2022
Int. Arch. Photogramm. Remote Sens. Spatial Inf. Sci., XLVIII-4/W1-2022, 279–286, 2022
https://doi.org/10.5194/isprs-archives-XLVIII-4-W1-2022-279-2022
 
05 Aug 2022
05 Aug 2022

EXPLORING JITTERING AND ROUTING OPTIONS FOR CONVERTING ORIGIN-DESTINATION DATA INTO ROUTE NETWORKS: TOWARDS ACCURATE ESTIMATES OF MOVEMENT AT THE STREET LEVEL

R. Lovelace1, R. Félix2, and D. Carlino3 R. Lovelace et al.
  • 1Institute for Transport Studies, University of Leeds, UK
  • 2CERIS, Instituto Superior Técnico, University of Lisbon, Portugal
  • 3Alan Turing Institute, UK

Keywords: Origin-Destination data, Methods, Jittering, Active transport, Road network, Validation

Abstract. Origin-Destination (OD) datasets provide vital information on how people travel between areas in many cities, regions and countries worldwide. OD datasets are usually represented geographically with straight lines or routes between zone centroids. For active travel, this geographic representation has substantial limitations, especially when zone origins and centroids are large: only using a single centroid origin/destination for each large zone results in sparse route networks covering only a small fraction of likely walking and cycling routes. This paper implements and explores the use of jittering and different routing options to overcome this limitation, thereby adding value to aggregate OD data to support investment in sustainable transport infrastructure. The route network results — generated from on an open dataset representing cycling trips in Lisbon, Portugal — were compared with a ground-truth dataset from 67 count locations distributed throughout the city. This approach enabled exploration of which jittering parameters and routing options lead to the most accurate route network results approximating the real geographic distribution of cycling trips in the study area. We found that jittering and disaggregating OD data, combined with routing using low level of traffic stress (quieter) preferences resulted in the most accurate route networks. We conclude that a combined approach involving 1) jittering with intermediate levels of disaggregation and 2) careful selection of routing options can lead to much more realistic route networks than using established OD processing techniques. The methods can be deployed to support evidence-based investment in strategic cycling and other sustainable transport networks in cities worldwide.