Proposal: Survey of Climate Data Mining

Global climate change has been a discussion topic for years, finally culminating as a significant problem and a national defense issue according to the 2010 Quadrennial Defense Review. From weather to ecological data, the planet is continuously being monitored by researchers to collect spatio-temporal climate data. With the immense amount of data collected, the challenge is making sense of the data by building models and studying the climatic events that deviate and correlate to the models. Traditionally, statistics have been used to study and build climate models, but now data mining is being used for this purpose. This presents a unique challenge for researchers because there are four general issues involved in climate data mining: preprocessing data and applying data mining techniques, outlier and anomaly detection, evaluation techniques and methods, and data visualization. For each of these issues, there are special challenges and techniques. The purpose of this proposed survey paper is to broadly explore these four issues and the techniques used to resolve the challenges. The proposed survey paper will be organized as follows. Section 2 will give a general overview of the four issues. Section 3 and 4 will discuss preprocessing and data mining techniques. Section 5 discusses outlier and anomaly detection. Section 6 discusses evaluation techniques. Finally, section 7 will discuss visualization.