Weather prediction using fuzzy sets and inference methodology

This paper presents the use of fuzzy inference methodologies in the prediction of the weather, as a viable alternative to the classical complex, costly and time-consuming weather prediction techniques. Basic weather elements which affect weather characteristics and which underlie the factors used in advanced weather prediction methods are fuzzified. They are then used in fuzzy weather prediction models based on fuzzy inferences. These models are simulated and the crisp results obtained using different existing and recently developed defuzzification strategies are compared with the actual weather data. Further, conclusive comments are provided regarding the effectiveness of the introduced models and the avenues which need to be explored in order to make them increasingly efficient.