Cloudburst Prediction System

Cloudbursts present significant risks to urban infrastructure and public safety due to their abrupt and localized characteristics, frequently leading to flash floods and landslides. This study introduces the Advanced Cloudburst Prediction System, a hybrid AI-driven framework aimed at providing real-time assessments of cloudburst risks specific to cities. The system combines a Random Forest classifier with an LSTM neural network, utilizing both historical simulations and current weather data sourced from the OpenWeatherMap API. Its outputs feature dynamic risk probabilities, visual analytics, regional risk maps, and emergency notifications through a Gradio web interface. By delivering timely warnings and practical insights, this system enables both authorities and citizens to improve their disaster preparedness and response strategies.