The International Conference for High Performance Computing, Networking, Storage and Analysis
Visualization of Three-Dimensional Temperature Distribution in Google Earth.
Authors: Fumiaki Araki (Japanese Agency for Marine-Earth Science and Technology), Tooru Sugiyama (Japanese Agency for Marine-Earth Science and Technology), Shintaro Kawahara (Japanese Agency for Marine-Earth Science and Technology), Keiko Takahashi (Japanese Agency for Marine-Earth Science and Technology)
Abstract: The graphical contents described on Google Earth are mainly classified according to a number of spatial dimensions from zero to two (2D): the point source of an earthquake-epicenter, the curves of cruising-track, and the surface of satellite image are examples on the geo-scientific topics. This classification below 2D comes from that Keyhole Markup Language (KML) used in Google Earth does not support three-dimensional objects except the statistical charts or the model buildings. Therefore, a new idea is expected to describe the three-dimensional contents, especially obtained from numerical simulations in which the data are defined on spatially three-dimensional grids.
Here we propose a novel technique to represent the three-dimensional scalar data on Google Earth. The key is opacity. Appropriate value of the opacity is assigned to data. As a result, we can easily see multiple data aligned the view direction. We select PNG image-format because the alpha channel for opacity is supported.
The simulation is performed on Earth Simulator using the urban-scale atmospheric simulation model called the Multi-Scale Simulator for the Geoenvironment (MSSG). The shown data set is the three-dimensional, time-developing temperature distribution in the area of about 4km-square in Minato Mirai 21 district, Yokohama, Japan, The spatial resolution is about 5 meter (~ 0.000045 degree) with the grid number of 968 (longitude) x 792 (latitude) x 82 (layers). The calculated time is 30 minutes with interval of 10 seconds. We successfully obtain quite nice visualized movie which is similar to the volume rendering method.