AI-Based Weather Forecasting and Natural Language Summary Generation Using Time Series Modeling

Weather forecasting is a critical task with wide-ranging applications in agriculture, transportation, public safety, and climate-sensitive industries. While traditional forecasting methods rely on numerical simulations of physical atmospheric processes, modern data-driven techniques such as machine learning and time series modeling offer accessible alternatives. This research presents an AI-powered weather prediction system that utilizes Facebook Prophet for time series forecasting and integrates Natural Language Processing (NLP) to automatically generate human-readable summaries of predictions. The model forecasts daily minimum temperatures based on historical weather data and converts forecasted data into descriptive monthly insights. Additionally, the system supports basic natural language queries such as “Tell me about July” by extracting intent using rule-based parsing. Our experiments demonstrate that the model delivers accurate forecasts and enhances interpretability, offering a robust tool for non-expert users.