Cloud movement prediction using GPS information

Abstract

Cloud movement prediction is the process of forecasting the direction and speed of movement of clouds in the atmosphere. By reflecting sunlight back into space and trapping heat within the atmosphere, clouds are an essential part of the Earth's climate system and play a significant role in the energy balance of the planet. To predict cloud movements in the sky, ground-based cameras are used progressively nowadays to capture the image of cloud movements. The image of cloud is categorised into 4 categories: • Category 1 – Glare • Category 2 – Dark Cloud • Category 3 – Clear Sky • Category 4 – Cloud Overcast While cloud classification is typically done manually by looking through images one at a time, the accuracy will be affected due to different people's perceptions of the various cloud types, and it is extremely time consuming. By automating this process, this research aims to improve cloud classification. Different methods in improving the accuracy of cloud image classification are discussed and explored in this report.