Prediction of Cumulonimbus Clouds in Airport Vicinity using NOAA Satellite Imagery and Random Forest Models

This paper demonstrates a machine learning approach using Random Forest models and NOAA satellite imagery to effectively predict cumulonimbus (Cb) cloud formation, achieving over 75% accuracy.Unlike previous studies that rely on groundbased data, our method leverages satellite images to provide enhanced coverage and perspective.Of the algorithms tested, Random Forest significantly outperforms other models like Support Vector Machines and Neural Networks.Our results indicate that employing NOAA satellite images and Random Forest models holds promise for improving Cb prediction to support aviation safety