🔬 physics

Mobility shapes heat exposure inequalities in cities

By analyzing daily mobility patterns across 23 Spanish cities, this study reveals that low-income groups face significantly higher heat exposure than high-income groups due to the interplay between spatial heat gradients and the unequal organization of daily activities, particularly during commuting, a disparity that is better captured by radiation models than gravity models.

Original authors: Marc Duran-Sala, Mattia Mazzoli, Martin Hendrick, Gabriele Manoli

Published 2026-03-27

📖 5 min read🧠 Deep dive

CC BY 4.0

Original authors: Marc Duran-Sala, Mattia Mazzoli, Martin Hendrick, Gabriele Manoli

Original paper licensed under CC BY 4.0 (http://creativecommons.org/licenses/by/4.0/). ✨ This is an AI-generated explanation of the paper below. It is not written or endorsed by the authors. For technical accuracy, refer to the original paper. Read full disclaimer

The Big Idea: It's Not Just Where You Live, It's Where You Go

Imagine a city as a giant, unevenly heated pizza. Some slices are scorching hot (the "heat islands" with lots of concrete and no trees), and some are cool and refreshing (the green parks and shaded neighborhoods).

For a long time, scientists thought the only thing that mattered for heat exposure was where you live. If you lived on a hot slice, you were suffering. If you lived on a cool slice, you were safe.

This paper says that's only half the story.

It turns out that where you go during the day matters just as much. Your daily commute, your shopping trips, and your work location act like a "heat elevator." Depending on your income and age, your daily movements might either keep you on the cool slices or drag you onto the burning hot ones.

The Main Characters: The "Heat Commuters"

The researchers looked at 23 Spanish cities using mobile phone data (anonymized, so no one's privacy was invaded) to track where people went in the summer of 2022 and 2023. They found two main patterns:

1. The Income Gap: The "Heat Trap" vs. The "Cool Escape"

The Low-Income Commuter: Imagine a worker who lives in a hot, dense neighborhood. They take a bus or train to work, which is often in another hot, dense part of the city (like a downtown business district). Their entire day is spent moving between "hot slices" of the pizza. They are trapped in a Heat Trap.
The High-Income Commuter: Imagine a wealthier person who lives in a cooler, greener suburb. They might drive to a job in a modern, air-conditioned office park or a cooler part of the city. Their daily path connects "cool slices." They are enjoying a Cool Escape.

The Result: Low-income people are consistently exposed to more heat than rich people, not just at home, but all day long. The study found that commuting (going to work or school) makes this gap even wider. Routine trips reinforce the inequality.

2. The Age Gap: The "Young & Hot" vs. The "Old & Cool"

The Younger Group: Younger people (students, young workers) tend to move around more and visit slightly hotter areas than the elderly.
The Elderly Group: Older people tend to stay in slightly cooler areas.
The Catch: While the age gap exists, it is much smaller than the money gap. However, the most vulnerable group of all is older people who are also poor. They live in hot areas and don't have the resources to escape them, making them the "double trouble" group for heat risks.

The "Magic Crystal Ball" Test: Can We Predict This?

The researchers asked a fascinating question: If we just look at a map of where people live and how far apart neighborhoods are, can we predict who gets hot and who stays cool?

They tested two famous "crystal balls" (mathematical models) used to predict human movement:

The Gravity Model: This model assumes people move like planets: they are pulled to big cities (high population) and pushed away by distance.
- The Verdict: This model failed. It was too simple. It thought everyone mixes together evenly, so it underestimated the heat inequality. It didn't realize that poor people and rich people actually stick to their own "thermal neighborhoods."
The Radiation Model: This model is smarter. It assumes people look at all the opportunities between where they are and where they want to go. It accounts for the fact that you might skip a nearby hot spot to go to a slightly farther cool spot if there's a better opportunity there.
- The Verdict: This model worked! It successfully predicted the heat gaps.

What this tells us: The heat inequality isn't just because poor people choose to go to hot places. It's because the structure of the city forces them into a pattern where their daily activities (work, school, shopping) are located in hotter zones, while rich people's activities are in cooler zones. The city layout itself is the culprit.

The Takeaway: Fixing the City, Not Just the People

Think of the city as a game board. If the board is set up so that the "poor player" always lands on the "burning square" and the "rich player" lands on the "ice square," changing the players' behavior won't fix the game. You have to change the board.

What should city planners do?

Don't just look at homes: When planning for climate change, don't just check which neighborhoods are hot. Check where people work and shop.
Cool the commute: If the route to work is a "heat tunnel," we need to add trees, shade, and cooling stations along the way.
Fairness is key: We need to ensure that essential services (jobs, schools, hospitals) aren't all clustered in the hottest parts of town, leaving the most vulnerable people to suffer the most.

In a Nutshell

This paper proves that heat inequality is a moving target. It follows you from your home to your job. Poor people are stuck in a daily loop of heat, while wealthier people can often find cooler paths. To protect everyone from a warming world, cities need to design their streets and transit systems so that everyone has a cool escape, not just the rich.

1. Problem Statement

Urban heat islands pose significant public health and economic risks, but existing assessments of heat exposure inequality are largely static, relying on residence-based metrics. These approaches fail to capture the dynamic nature of human exposure, which depends on where individuals spend their time throughout the day (commuting, working, leisure).

The Gap: It remains unclear how daily mobility patterns, embedded within heterogeneous urban heat fields, shape heat exposure inequalities across different sociodemographic groups (income, age, gender).
The Question: Do daily movements amplify or attenuate existing residential heat disparities? Can standard population-based mobility models (which ignore socioeconomic attributes) reproduce these observed inequalities?

2. Methodology

The authors developed a network-based framework to quantify mobility-driven heat exposure across 23 Spanish cities during the summer months (July–August) of 2022 and 2023.

Data Sources

Mobility Data: Anonymized mobile phone records from the Spanish Ministry of Transport (MITMA), providing daily Origin-Destination (OD) flows between administrative districts. Data is stratified by:
- Income: Low (<€10k), Mid (€10k–€15k), High (>€15k).
- Age: Young (0–25), Mid (25–45), Older (45–65), Elderly (65–100).
- Gender: Male/Female.
Thermal Data: High-resolution Land Surface Temperature (LST) from NASA MODIS (1km resolution). Districts were ranked into temperature deciles (1–10) based on mean summer LST.
Demographics: Population and income data from the Spanish National Statistical Office (INE).

Analytical Framework

Network Construction: Districts are nodes; weighted directed edges represent trips between temperature deciles.
Key Metrics:
- Heat Exposure ( $E$ ): The trip-weighted mean temperature decile of visited destinations.
- Heat Assortativity ( $\rho$ ): The tendency for trips to connect districts of similar temperature (positive $\rho$ implies "heat traps" or segregation).
- Heat Directionality ( $R$ ): The net bias of movement toward hotter ( $R>0$ ) or cooler ( $R<0$ ) districts.
Inequality Quantification: Calculated as the difference in exposure between groups (e.g., $\Delta E_{inc} = E_{low} - E_{high}$ ).
Model Validation: Tested two state-of-the-art, parameter-free population mobility models to see if they could reproduce observed inequalities without explicit socioeconomic inputs:
- Gravity Model: Flows depend on origin/destination population masses and distance.
- Radiation Model: Flows depend on intervening opportunities (population between origin and destination).

3. Key Results

A. Systematic Income and Age Inequalities

Income: Low-income groups consistently experience higher heat exposure than high-income groups across most cities ( $\Delta E_{inc} > 0$ $Δ E_{in c} > 0$ ).
- Example: In Madrid, low-income residents visit significantly hotter districts than high-income residents.
- Exception: In a few cities (e.g., Murcia), the pattern reverses, correlating with local reversals in the income-heat spatial structure.
Age: Younger individuals (0–25) experience slightly higher heat exposure than the elderly (65–100), though the magnitude of this disparity is smaller than income-based inequality.
Vulnerability Intersection: Elderly low-income residents exhibit the highest combined risk: they face high exposure levels (similar to younger low-income groups) while possessing higher physiological vulnerability.

B. The Role of Trip Purpose

Commuting vs. Non-Routine: Routine trips (work/study) exhibit higher heat exposure and stronger heat assortativity than non-routine trips.
Amplification: Commuting reinforces spatial heat gradients more than leisure travel. However, non-routine mobility does not eliminate inequalities, suggesting the disparity is structurally anchored in the city's spatial organization of employment and activity hubs.

C. Model Performance (Gravity vs. Radiation)

Gravity Model: Underestimates income and age-related exposure differences. It tends to homogenize flows toward central, large districts, compressing thermal contrasts between groups.
Radiation Model: Successfully reproduces most observed disparities. By accounting for intervening opportunities, it better preserves the spatial structure of flows between peripheral districts with distinct thermal profiles.
Implication: Heat exposure inequalities emerge primarily from the interplay between unequal origin patterns (where groups live) and the spatial structure of the city, rather than from group-specific behavioral differences in destination choice.

D. Robustness Checks

Placebo Tests: Shuffling income/age labels eliminated inequalities, confirming the results are not statistical artifacts.
Temporal Stability: Inequalities persist daily and across weekends, though weekend dynamics show slight attenuation.
Gender: No systematic heat exposure inequality was found between genders, likely due to smaller mobility differences compared to income/age.

4. Key Contributions

Dynamic Exposure Framework: Moves beyond static residence-based assessments to a network-based approach that captures how daily mobility redistributes populations across urban thermal gradients.
Structural Mechanism Identification: Demonstrates that heat inequality is not just a result of where people live, but how cities spatially organize daily activities. The inequality is "structurally embedded" in the urban form.
Modeling Insight: Establishes that the Radiation Model is superior to the Gravity Model for capturing environmental inequalities because it better accounts for the spatial competition of opportunities, which is crucial for modeling flows in heterogeneous thermal landscapes.
Policy-Relevant Metrics: Provides a generalizable method to quantify mobility-driven exposure, applicable to cities lacking detailed behavioral data.

5. Significance and Implications

Urban Planning: Strategies to mitigate heat inequality must go beyond residential cooling. Planners must consider the spatial distribution of employment centers, services, and cooling infrastructure relative to where low-income populations commute.
Public Health: Interventions should target the "most exposed and vulnerable" (e.g., elderly low-income residents) and consider that routine commuting is a major driver of cumulative heat stress.
Climate Resilience: As climate change intensifies heat extremes, understanding how mobility patterns redistribute risk is critical for designing equitable, climate-resilient cities. The study suggests that parsimonious models (like Radiation) can effectively guide these strategies without requiring complex, group-specific behavioral data.

In conclusion, the paper proves that mobility acts as a multiplier of environmental inequality, reinforcing the thermal disadvantages of low-income and younger populations through daily movement patterns that are structurally determined by the city's layout.