How to Visualize the Impact of Urbanization on Public Health Using EDA

Urbanization is transforming cities at an unprecedented pace, influencing many aspects of daily life—including public health. As populations increasingly shift from rural to urban areas, understanding the complex interplay between urbanization and public health becomes critical. Exploratory Data Analysis (EDA) serves as a powerful tool to visualize, interpret, and draw insights from these multidimensional relationships. Here’s a comprehensive guide on how to visualize the impact of urbanization on public health using EDA techniques.

Understanding Key Variables

To begin visualizing the impact of urbanization on public health, it’s essential to identify the variables that define each aspect.

Urbanization Indicators:

• Population density

• Rate of urban population growth

• Built-up area expansion

• Access to urban infrastructure (water, electricity, sanitation)

• Transportation density

• Urban green space availability

Public Health Metrics:

• Incidence of communicable and non-communicable diseases

• Air quality index (AQI)

• Mortality and morbidity rates

• Hospital and healthcare facility accessibility

• Mental health statistics

• Waterborne disease outbreaks

• Noise pollution levels

Data Sources

Gathering relevant datasets is the first step in conducting EDA. Reputable sources include:

• World Health Organization (WHO)

• World Bank Open Data

• United Nations Department of Economic and Social Affairs

• National statistical bureaus

• OpenStreetMap and satellite imagery for spatial data

• Environmental Protection Agencies for pollution data

Data Preprocessing and Cleaning

Before visualization, data must be preprocessed:

• Handling missing values through imputation or deletion.

• Standardizing units (e.g., converting all temperature values to Celsius).

• Normalizing data for variables with different scales.

• Parsing and formatting dates, ensuring consistency across datasets.

• Geocoding for mapping spatial information.

Univariate Analysis

Start with simple visualizations of single variables to understand distributions and anomalies.

Visualizations:

• Histograms for understanding distributions of health metrics like disease incidence rates.

• Box plots to highlight outliers in variables like AQI or mortality rates.

• Bar charts to compare public health infrastructure availability across cities.

Example: A histogram showing the distribution of PM2.5 levels across urban areas reveals which cities have dangerously high pollution levels.

Bivariate and Multivariate Analysis

This phase reveals relationships between urbanization indicators and health outcomes.

Scatter Plots:

• Population density vs. air quality index

• Urban sprawl vs. incidence of asthma or respiratory diseases

• Green space per capita vs. mental health disorder prevalence

Heatmaps:

• Correlation heatmaps help visualize multivariate relationships, such as how strongly different urbanization indicators correlate with various public health outcomes.

Pair Plots:

• Pair plots (scatterplot matrix) can be useful to explore multiple bivariate relationships simultaneously.

Regression Plots:

• Use regression lines in scatter plots to assess linear relationships. For instance, plotting built-up area percentage against respiratory illness incidence can show trends.

Time-Series Visualization

Urbanization and its impact on health evolve over time. Time-series analysis helps identify long-term trends and seasonality.

Line Graphs:

• Urban population growth vs. respiratory disease rates over a decade

• AQI vs. hospitalization rates year over year

Area Charts:

• Useful to demonstrate cumulative impacts, such as the increasing burden of lifestyle-related diseases in urban populations.

Rolling Averages:

• Smooth out short-term fluctuations in time-series health data for clearer trends.

Spatial Analysis

Geospatial visualization is crucial to understand the geographic spread of health impacts due to urbanization.

Tools:

• GIS platforms (ArcGIS, QGIS)

• Python libraries (Folium, Geopandas)

• Heatmaps over maps using tools like Leaflet or Plotly

Visualizations:

• Choropleth maps to show disease incidence or pollution levels by city or neighborhood.

• Dot density maps to indicate hospital or clinic locations relative to population clusters.

• Urban heat island visualizations using satellite imagery and temperature data to connect to heat-related illnesses.

Example: A choropleth map of dengue fever cases in relation to urban water stagnation zones highlights vulnerable urban neighborhoods.

Categorical Analysis

EDA on categorical variables helps understand how demographic or socio-economic groups are affected differently.

Bar Charts and Count Plots:

• Compare healthcare access across income or ethnic groups in urban areas.

• Examine the prevalence of chronic diseases among different age brackets in densely populated areas.

Mosaic Plots:

• Depict the relationship between multiple categorical variables such as gender, income level, and disease type.

Advanced Visualization Techniques

Incorporating advanced EDA techniques can enhance interpretability and insight.

Cluster Analysis:

• Use clustering (e.g., K-means) to group cities or districts by similar urban and health characteristics.

• Visualize clusters using colored scatter plots or maps.

Dimensionality Reduction:

• Principal Component Analysis (PCA) helps reduce data complexity and reveal key factors affecting urban health.

• Use 2D PCA plots to visualize major trends.

Interactive Dashboards:

• Tools like Tableau, Power BI, or Dash can create interactive visualizations for stakeholders to explore EDA results.

• Dashboards may include filters for year, location, and metric type to make the data exploration dynamic.

Case Study Example

Consider a case study comparing 10 rapidly urbanizing cities over a 20-year period. EDA steps could include:

• Visualizing urban expansion using satellite-derived built-up area layers.

• Overlaying AQI data to reveal trends in pollution hotspots.

• Mapping public health clinics and overlaying population density.

• Comparing disease incidence rates before and after significant urban growth events.

Findings may include:

• Strong positive correlation between population density and respiratory disease.

• Inverse correlation between green space and mental health disorders.

• Clustering of disease outbreaks near unplanned urban slums.

Conclusion

EDA offers a multifaceted way to visualize and understand the impact of urbanization on public health. By integrating temporal, spatial, and multivariate techniques, stakeholders can derive actionable insights for urban planning, policy-making, and public health interventions. The power of visualization lies in its ability to make complex data intuitive and compelling, ultimately helping to design healthier cities in an increasingly urbanized world.