Clearing The Skies

Linking Aviation Emissions to Climate Impacts on At-Risk Communities

By: Shanyu Gowdu

Background

The aviation industry accounts for around 2-3% of global CO2 emissions.
While this might seem small compared to other sectors like energy or manufacturing, aviation is a growing industry, and emissions from aviation are not evenly distributed across the globe.

*I did not create this visualization* - provided by aviationbenefits.org

Aviation emissions are released high in the atmosphere, where they have a disproportionately high impact on climate change due to the altitude at which they occur
Aviation is also one of the hardest sectors to decarbonize due to technological limitations
Expecting huge rapid growth

*I did not create this visualization* - provided by Oliver Wyman Analysis

ND-Gain Climate Vulnerability score: measures a country's exposure, sensitivity and ability to adapt to the negative impact of climate change.
The irony is that the countries with high vulnerability scores—countries like Haiti, Madagascar, and Vanuatu—produce negligible aviation emissions but face the greatest climate risks.

Data Collection: OECD - Aviation Emissions, University of Notre Dame - Climate Vulnerability Score, World Bank - CO2 emissions and GDP, United Nations - Population
Data Cleaning: Used Microsoft Excel and Generative AI to filter/clean data for 20 specific countries
Visualizations: Plotly for interactive, real-time visualizations and Matplotlib for static
Predictive Model: I used linear regression from the Scikit-learn package to predict aviation emissions for countries based on their GDP, population, and CO2 emissions.

Feature Selection: The features I chose for the predictive model were GDP (current US$), Population(Thousands), and CO2 Emissions (Metric Tons), as they are key factors influencing aviation activity and emissions.
Target Variable: The target variable was Aviation Emissions (Tons), which I aimed to predict based on the selected features.
Train-Test Split: I split the data into training and test sets using train_test_split, with 80% of the data for training and 20% for testing. This helped ensure that the model was trained on most of the data but validated on unseen data to test accuracy.
Model Training: I used Linear Regression (LinearRegression class from Scikit-learn) to train the model on the training data. The model learned the relationship between GDP, population, CO2 emissions, and aviation emissions.

Incorporate More Comprehensive Variables: Expand the range of variables
Time-Series Analysis Improvements:

Longer Time Period, Seasonal Adjustments
Improve Predictive Modeling: Non-linear models, Cross-validation, Hyperparameter tuning
Scenario-Based Modeling: predict how different policy implementations (e.g., carbon taxes, SAF adoption) might impact future aviation emissions
More Advanced Data Visualizations: Interactive Dashboards - Allow users to view future projections under scenario-based modeling

Policy Changes: Global Carbon Tax for Aviation, Frequent Flyer Levy, Sustainable Aviation Fuel (SAF) Incentives, International Collaboration and Regulatory Alignment
Investment in Rail: In the U.S., 2/3 of aviation emissions come from domestic flights, largely because of the lack of affordable reliable interstate public transportation like high-speed rail.
Enhanced Carbon Offsetting: Improve carbon offset programs to ensure they genuinely benefit high-vulnerability countries.

Aviation Dataset: https://data-explorer.oecd.org/vis?tm=aviation&pg=0&snb=24&df[ds]=dsDisseminateFinalDMZ&df[id]=DSD_AIR_TRANSPORT%40DF_AIR_TRANSPORT&df[ag]=OECD.SDD.NAD.SEEA&df[vs]=1.0&dq=.Q.......&pd=%2C&to[TIME_PERIOD]=false
GDP Dataset: https://datacatalog.worldbank.org/search/dataset/0037712/World-Development-Indicators
Population Dataset: https://population.un.org/wpp/Download/Standard/MostUsed/
CO2 emissions Dataset: https://data.worldbank.org/indicator/EN.ATM.CO2E.PC?end=2020&start=1990&view=chart
Generative AI was used to create code snippets