A Comprehensive Study on Earth System Science using Machine Learning and Process Mining

Machine Learning has emerged as a powerful tool in Earth System Science offering new avenues for data analysis and predictive modeling. This paper deep dive into the application of machine learning techniques to understand and predict air quality a critical aspect of Earth System Science. Process mining techniques are incorporated to explore data patterns and relationships. Our work explores the use of machine learning and process mining techniques in Earth System Science, with a specific focus on air quality assessment. The study involves data collection, preprocessing and the development of machine learning models including Linear Regression, Decision Trees, Random Forest, XGBoost and CatBoost to predict Air Quality Index (AQI). Performance evaluation using Root Mean Squared Error (RMSE) and R-squared (R
²
) metrics reveals that XGBoost and CatBoost demonstrate superior predictive accuracy. Process mining techniques are applied to datasets containing city-specific NOx (nitrogen oxides) measurements, resulting in the creation of heuristic process models. This research serves as a foundational exploration of the integration of machine learning models and process mining techniques in Earth System Science with a focus on air quality.