Network-Level Measures of Mobility from Aggregated Origin-Destination Data

TL;DR

This work develops a network-level framework for mobility using aggregated origin-destination data by casting flows as time-dependent Markov operators. It defines time-elapsed measures—net OD flows, distance-based metrics via time-elapsed paths, effective distance, and return-to-origin distances—within a privacy-preserving, collective description, validated on a synthetic network and real NetMob 2024 data. The key contributions include a principled pseudo Markov-chain model, a network-driven synthetic data generator for interpretation, and empirical demonstrations that meaningful large-scale structure emerges despite aggregation and coarser temporal resolution. These measures illuminate structural routing constraints, directional patterns, and mobility scales, offering tools for urban analysis and planning using aggregated mobility products when individual trajectories are unavailable.

Abstract

We introduce a framework for defining and interpreting collective mobility measures from spatially and temporally aggregated origin--destination (OD) data. Rather than characterizing individual behavior, these measures describe properties of the mobility system itself: how network organization, spatial structure, and routing constraints shape and channel population movement. In this view, aggregate mobility flows reveal aspects of connectivity, functional organization, and large-scale daily activity patterns encoded in the underlying transport and spatial network. To support interpretation and provide a controlled reference for the proposed time-elapsed calculations, we first employ an independent, network-driven synthetic data generator in which trajectories arise from prescribed system structure rather than observed data. This controlled setting provides a concrete reference for understanding how the proposed measures reflect network organization and flow constraints. We then apply the measures to fully anonymized data from the NetMob 2024 Data Challenge, examining their behavior under realistic limitations of spatial and temporal aggregation. While such data constraints restrict dynamical resolution, the resulting metrics still exhibit interpretable large-scale structure and temporal variation at the city scale.

Figures (20)

• Figure 1: Synthetic mobility network used in all experiments. Nodes correspond to resolution 6 cells over the Atlanta metropolitan area. Thicker nodes denote hub locations, with node color indicating distance bands from the designated center. Thicker edges correspond to high-capacity metro connections, while thinner edges represent local connections. This network serves as the fixed spatial structure on which time-dependent Markov dynamics generate synthetic mobility flows.
• Figure 2: Periodic fixed-point population distribution of the synthetic mobility model, corresponding to the midnight distribution of population-equivalent persons across H3 tiles. Darker colors indicate higher population density. Amber-circled nodes denote the designated center of the network, while all remaining nodes form the periphery.
• Figure 3: Synthetic morning net flows on an H3 resolution 6 spatial grid. Top $75^{\text{th}}$ percentile of net flows shown. Extending the time-elapsed window transforms sparse, local flows into a coherent radial pattern directed toward central hub nodes, revealing network-scale inbound structure.
• Figure 4: Synthetic afternoon and evening net flows on an H3 resolution 6 grid. Top $75^{\text{th}}$ percentile of net flows shown. Extending the time-elapsed window reveals a clear reversal of dominant flow directions, with outward redistribution from central hubs toward peripheral locations.
• Figure 5: Mexico City, 06--05--2019. Morning time-elapsed net flows showing the top $75^{\text{th}}$ percentile of net flows. Extending the time-elapsed window incorporates longer-distance inflows toward central locations, increasing both the density and spatial reach of inward flows.