Visualization for climate modeling

Techniques for visualizing several climate variables, particularly clouds and wind velocities, are discussed. The techniques are volume rendering, textured contour surfaces, and vector field rendering. The application of these techniques is discussed in the context of a high-definition television (HDTV) animation produced to show global cloud motion during ten simulated days in January. The computational grid is defined by nearly a million grid points-320 evenly spaced longitudes, 160 unevenly spaced latitudes, and 19 unevenly spaced surfaces of constant geopotential. The climate variables at every grid point were output at every hour of simulated time, resulting in 380 MB of data per simulated day.<>