Spatio-Temporal Pattern Detection in Climate Data

In this paper, a unique approach to the problem of spatiotemporal pattern detection is discussed in relation to climate data; this can be understood as discovering dependent climate events that occur over space. An accurate solution in this domain will provide climate scientists with highly valuable data which can be used to improve climate models and add to the knowledge base of climate science. This will in turn be beneficial to policy makers to make better informed decisions that can be based on improved data. This problem is one in which their are many valid solutions, which all must take into account the general problems that arise when dealing with big data, such as scaling and data complexity among other issues. The approach taken here calls upon the many concepts from the field of data-mining, and applies those concepts to the field of climate science to yield new and interesting knowledge. The concepts that will be discussed in this paper can also be used to locate and identify events in climate data-sets, which can be applied to improve the value of the climate data-sets that we have today by cross-referencing with other knowledge bases and datasets.