Stochastic Weather Modelling to Generate Rain, Snowfall and Wind

Ceegan Kohere and Reza Ryan Part of CW18

Within the past decade development and quality of weather phenomena in virtual environments has rapidly increased. However, there is a lack of documented framework to create a dynamic and optimized weather system suitable for real-time environment.

This research is a critical inquiry of current research and the implementation required to create such a weather system in real-time. In this research a dynamic weather model was created using different weather component generation techniques such as Particle emission, Markov chains, Cellular Automata, Tri-Planar projection and Depth mapping. The weather model was designed and tested through the design science research methodology to ensure functionality. This framework can be easily integrated into existing real-time engines.