Climatic Change Analytics in the Philippines: A Time-Series Predictive Modeling Approach

The Philippines ranks among the most climate-vulnerable nations, facing intensified extreme weather and rising heat due to climate change. This study presents a time-series predictive modeling approach using the Global Weather Repository dataset, focusing on Manila (2024– 2025). We engineer features—including a wet-bulb temperature index—to enhance forecasting of daily temperatures, a critical climate variable amid increasing heat extremes. Through exploratory analysis (Python, Tableau) and rigorous evaluation, we compare classical and machine learning models. A tuned SARIMA model with engineered inputs achieves high predictive accuracy (~90% R<sup xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">2</sup>) on validation data without overfitting. Visualizations illustrate historical trends, forecast performance, and modeling workflow. Our approach provides a scalable, data-driven framework for localized climate analytics in the Philippines. It supports evidence-based adaptation strategies by identifying temperature patterns, extreme heat risks, and forecasting capabilities. This work contributes an original methodology combining engineered climate indicators with time-series modeling, offering practical insights for policy and research in the context of global warming.