Deciphering Climate-Mobility Links: How patterns vary across altitude areas in the Metropolitan City of Turin

Technical Summary: Deciphering Climate-Mobility Links in the Metropolitan City of Turin

Problem Statement and Context
While migration is recognized as a complex, multicausal process, existing literature on climate-mobility links remains fragmented, particularly regarding the specific influence of climate indicators on mobility patterns. Research has predominantly focused on the Global South, leaving a significant gap in understanding whether similar dynamics apply in the Global North, specifically within European mountainous regions. Furthermore, the extent to which climate change shapes migration within relatively small geographical areas, and how these patterns vary across altitude (lowlands vs. highlands), remains insufficiently explored. This study addresses these gaps by investigating the Metropolitan City of Turin (CMTO), Italy—a region encompassing over 52% mountainous terrain and situated in a recognized climate hotspot. The central research question asks: To what extent have people migrated to areas less affected by heat and drought, and how do these patterns differ between highland and lowland destinations?

Methodology
The study employs a mixed-methods approach utilizing residential data from 2006 to 2022 across 312 municipalities in the CMTO. The analysis is grounded in a tri-dimensional conceptual framework covering social, urban-built, and natural environments.

1. Data Sources:

   • Demographic/Mobility: Annual data on new and cancelled residencies, population structure, and socioeconomic indicators provided by the Metropolitan City of Turin and ISTAT.
   • Climate Data: Monthly averaged data on skin temperature, total precipitation, and relative humidity derived from ERA5 (Copernicus Climate Change Service).
   • Variables: The study calculates annual anomalies for summer temperature, precipitation, and humidity by comparing the 2006–2022 period against the 1971–2000 reference period.

2. Analytical Techniques:

   • Bivariate Mapping: Used to visualize spatial variations between lowlands and highlands. This method maps net mobility (inbound minus outbound migration as a percentage of total population) against climate anomalies (temperature, precipitation, humidity) to identify territorial patterns.
   • Fixed-Effects Regression: A panel data model was employed to assess the correlation between climate indicators and mobility flows while controlling for time-invariant characteristics (e.g., culture, location) and macroeconomic fluctuations. The model includes covariates for the social environment (population size, age, unemployment, income) and the urban-built environment (housing prices).
   • Stratification: Municipalities were classified into three altitude categories (Mountains, Hills, Lowlands) to examine altitude-specific differences.
   • Robustness Checks: The study utilized pooled OLS regression with lagged variables to test the temporal influence of climate anomalies on migration decisions.

Key Results
The analysis reveals that the relationship between climate change and mobility is non-linear and highly dependent on altitude and socioeconomic context.

• Spatial Patterns: Bivariate mapping indicates that while approximately 55% of the sample consists of municipalities heavily impacted by heat and drought, these areas remain primary destinations for inbound mobility. No single, clear altitude-related pattern emerged from the visual mapping alone, suggesting that climate indicators individually do not strongly drive outward migration without considering other factors.
• Regression Findings (Temperature): Rising summer temperatures are significantly associated with outmigration trends. In the full sample, a one-degree Celsius increase in temperature anomaly correlates with a decrease in net mobility. This negative correlation is most pronounced and statistically significant in lowland and hilly municipalities.
• Regression Findings (Humidity and Precipitation):
  • In lowlands, higher temperatures and lower humidity (drought conditions) are associated with reduced inbound mobility.
  • In mountainous areas, temperature anomalies do not show a significant impact on mobility. Conversely, higher humidity is positively associated with net mobility in mountains, while reduced precipitation correlates with higher immigration.
  • In hilly areas, the combination of higher temperatures and lower humidity reduces attractiveness for new residents.
• Socioeconomic Drivers: Socioeconomic factors remain the dominant drivers of mobility. Unemployment rates and the percentage of non-Italian residents show strong, consistent correlations with net mobility across all altitude levels. Population size and the percentage of the population under 18 also show significant negative correlations with mobility, suggesting families and larger populations are less likely to relocate.
• Thermal Discomfort: The "Discomfort Index" (a measure of thermal stress) showed a positive correlation with net mobility in lowlands, suggesting that individuals may migrate to areas with moderate thermal discomfort if other socioeconomic factors are favorable, challenging the notion that heat alone drives immediate relocation.

Key Contributions

• Geographical Scope: The paper provides novel evidence on climate-mobility links in a less-studied geographical area (the European Alps in the Global North), challenging the predominance of Global South-focused literature.
• Altitude Differentiation: It demonstrates that climate impacts on mobility are not uniform; they vary significantly between lowlands, hills, and mountains. Specifically, it highlights that while heat and drought affect lowlands, mountainous areas exhibit different sensitivities, particularly regarding humidity and precipitation.
• Methodological Integration: By combining bivariate mapping with fixed-effects regression and a tri-environmental framework, the study offers a nuanced view of how slow-onset climate events interact with socioeconomic drivers to shape residential choices.
• Indicator Expansion: The study broadens the scope of climate indicators by including relative humidity and the discomfort index, which are often overlooked in climate mobility research.

Significance and Claims
The authors claim that their findings emphasize the necessity of considering the inherent characteristics of specific areas when analyzing climate-mobility links. The study asserts that while heat and drought-affected lowlands appear less attractive for inbound mobility, these patterns cannot be generalized across all territories. The research provides essential insights for developing effective adaptation strategies and long-term resilience plans, particularly for urban planning in regions susceptible to climate change. The paper concludes that climate change is a relevant factor in migration decisions, but its influence is mediated by local socioeconomic conditions and altitude, requiring tailored policy responses rather than broad generalizations.