Skip to Main Content
Geoscientists have a constant need to query into large-scale multidimensional array-based datasets. The most efficient way to accelerate queries is indexing. We focus on the climate datasets and propose a novel and efficient indexing method called the chunk-locality index. The main idea of this method is to take advantage of the spatial-temporal data similarity in climate datasets. We evaluate the performance of chunk-locality index in various chunk sizes with two practical climate datasets, and compare the performance results with the bitmap index. The comparison results show that the chunk-locality index presents better performance than the bitmap index not only in improving the efficiency of data queries but also in the index building time and the index size.