Intensity Estimation of Tropical Cyclone using Different Satellite Imagery and Random Forest Classifier

Intensity estimation of Tropical cyclones is an important task due to its impact on lives and property. Intensity estimation using satellite images is a potential domain where re-searchers are working to automate the estimation process with-out much human involvement. Most of the TC intensity estimations techniques are involved Infrared (IR) images. However, this work uses IR, Visible (VIS), Water vapor (WV), and Special Sensor Microwave/Imager (SSM/I) images for TC intensity estimation through feature extraction and Random Forest (RF) classi-fication technique. A comparison between the abovementioned major types of satellite images is shown by the classification accuracy of TC images. The study shows that IR images work best in different resolutions compared to others.