Voxel-based Analysis and Visualization of Rainfall Data

The remote sensing of hydrologic variables such as rainfall, soil moisture and streamflow is becoming increasingly more important as we strive to improve our understanding of the earth’s climate. Because these variables change more rapidly than other land cover characteristics, a high frequency of image acquisition is required. This in turn results in a large volume of data that must be compared to detect changes and trends in moisture flux. Traditional remote sensing tools are not well suited to this analysis because of the large number of scenes involved. In this paper, we explore the use of 3D geovisualization techniques to analyze a large time series of satellite imagery. We take advantage of the 3 dimensional voxel data structure to transform the image collection in a space-time cube. We employ 3D voxel-based visualization methods also known as volume visualization to perform feature analysis, change detection and segmentation of this large spatio-temporal dataset. We demonstrate that the use of the spacetime cube offers even those hydrometeorologists that prefer to look at 2D maps and graphs with improved access to the time series data through software tools for dynamic graphing and slicing. In a practical application, we use the space-time cube to present the evolution of the annual rainfall cycle over Africa as a single visual image.