Mobility shapes heat exposure inequalities in cities

Abstract

Segregation has long been recognized as a driver of environmental inequalities, with disadvantaged groups often living in neighborhoods where heat-related risks are highest. Yet, it remains unclear how daily mobility patterns, embedded within heterogeneous urban heat fields, shape heat exposure inequalities across sociodemographic groups. Using a mobile phone dataset of daily mobility flows and urban temperature fields across 23 Spanish cities, we develop a network-based framework to quantify how different sociodemographic groups experience heat through their daily movements. We find systematic income-related inequalities, with low-income groups consistently experiencing higher exposure than high-income groups, while age-related disparities are smaller in magnitude, with younger individuals slightly more exposed than elderly ones. These inequalities intensify during commuting trips, indicating that routine mobility amplifies spatial heat gradients more than non-routine movements. We further assess whether state-of-the-art population-based mobility models can capture these observed inequalities. The gravity model underestimates income- and age-related exposure differences, whereas the parameter-free radiation model captures most of the observed disparities. This indicates that heat exposure inequalities largely emerge from the interplay between the unequal organization of daily activities across sociodemographic groups and urban heat gradients, rather than from group-specific behavioral differences. Our findings provide a generalizable framework to characterize mobility-driven heat exposure inequalities and inform climate-resilient urban planning and public health strategies as cities face intensifying climate-related risks.

Figures (22)

• Figure 1: Mobility networks at the district level stratified by age and income groups across Spanish cities. (a) City-level mean temperature averaged over July–August 2022 and 2023 for the 23 cities included in the analysis. (b) Mobility network at district-level in Madrid aggregated over July–August 2022 and 2023. Only the top $5\%$ of links by share of trips are displayed, with edge width proportional to the absolute number of trips and edge color indicating the relative share of trips. Nodes are colored by deciles of average district temperature over the same period.
• Figure 2: Mobility-driven heat exposure inequalities across income and age groups. (a-b) Origin-destination matrices (ODs) between temperature deciles for income and age groups in Madrid. (c-d) Mobility-driven heat exposure of high versus low income groups, and elderly versus young populations across cities. Colored points highlight selected cities for visual reference; grey points correspond to the remaining cities. The $y=x$ line indicates no exposure inequality between groups. (e) Histogram of exposure difference $\Delta E$ across cities. Positive values indicate that low-income groups and young populations are more exposed than high-income groups and elderly populations, respectively. (f) Population-weighted average exposure across cities for all income-age groups. Error bars denote the standard error of the population-weighted average.
• Figure 3: Daily dynamics of mobility-driven heat exposure, heat assortativity, and heat directionality differ between commuting and non-routine trips across income and age groups. Daily population-weighted average heat exposure, assortativity, and directionality across all cities in August 2023, shown separately for work/education trips and non-routine trips, for income and age groups. Error bands represent the standard error of the population-weighted mean. Grey shading indicates weekends. The dashed red line marks 15th August. A 2-day centered window smooths short-term noise.
• Figure 4: Parsimonious mobility models largely reproduce income- and age-related heat exposure inequalities. (a) Illustration of the mobility networks derived from empirical data and from gravity and radiation models for Barcelona city during July-August 2022 and 2023 for high income group. Only the top $200$ links ranked by share of trips are displayed. Edge width scales with the absolute number of trips, while edge color encodes the relative share of trips. Nodes are colored by deciles of average district temperature over the same period. Observed versus predicted heat exposure differences for (b) income ($\Delta E^{inc}$) and (c) age ($\Delta E^{age}$) for each city. The dashed gray $y=x$ line indicates perfect agreement between observations and model predictions. (d) Summary of regression metrics from the linear fit of predicted vs observed exposure differences across cities. Bars show slope (top) and coefficient of determination $R^2$ (bottom).
• Figure S.1: Comparison between aggregated and daily temperature deciles. On the left, for each city, district-level temperature deciles computed from mean-summer LST are compared with deciles computed for a single day. Each point represents a district’s aggregated decile versus its daily decile; the solid diagonal indicates perfect agreement, and dashed lines mark $\pm 1$ and $\pm 2$ deciles. On the right pooled comparison across all cities, showing the count of district–day pairs in each aggregated–daily decile combination.