Limitations of gravity models in predicting fine-scale spatial-temporal urban mobility networks

TL;DR

This paper examines the limitations of gravity models in predicting fine-scale urban mobility networks, revealing their reduced accuracy at hourly and daily scales and emphasizing the need for alternative modeling approaches.

Contribution

It demonstrates that gravity models are less effective for fine-grained, high-resolution urban mobility data and highlights the necessity for new models or machine learning methods.

Findings

Finer-scale mobility networks are not scale-free.

Gravity model performance declines at hourly/daily scales.

Population density variations do not significantly affect model performance.

Abstract

This study identifies the limitations and underlying characteristics of urban mobility networks that influence the performance of the gravity model. The gravity model is a widely-used approach for estimating and predicting population flows in urban mobility networks, assuming the scale-free property. Prior studies have reported good performance results for the gravity model at certain levels of aggregation. However, the characteristics of urban mobility networks might vary depending on the spatial and temporal resolutions of data. Hence, the sensitivity of gravity model performance to variation in the level of aggregation of data and the temporal and spatial scale of urban mobility networks needs to be examined. The basic gravity model is tested on urban mobility networks on an hourly and daily scale using fine-grained location-based human mobility data for multiple US metropolitan counties to address this gap. The findings suggest that: (1) finer-scale urban mobility networks do not demonstrate a scale-free property; (2) the performance of the basic gravity model decays for predicting population flow in the finer-scale urban mobility networks; (3) the variations in population density distribution and mobility network structure and properties across counties do not significantly influence the performance of gravity models. Hence, gravity models may not be suitable for modeling urban mobility networks with daily or hourly aggregation of census tract to census tract movements. The findings highlight the need for new urban mobility network models or machine learning approaches to predict fine-scale and high temporal-resolution urban mobility networks better.