GANG

home

people

projects

< previous | next >

All the clouds shown here have the same C1, H, alpha values (0.1, 0.333, 1.8). In a study on nearly 1000 satellite radiances at visible and infra red wavelengths (Lovejoy, Schertzer, Stanway 2001), it was found that the isotropic statistics for alpha, C1 were indeed very close to these values. However, the value of H was quite variable: multifractal simulations with cloud liquid water density with H=0.333 (the value obtained from aircraft is 0.3) indicated that various H values for the radiation could be obtained depending on the optical properties and illumination of the clouds. In short, the radiative effects can apparently account for the varying H of the result. In the following we show purely visually the result of varying the cloud anisotropy. All the pictures are believed compatible with the observations of real anisotropic clouds. This gallery is a visual attempt to demonstrate that even if the basic C1, alpha, H value of the liquid water of all clouds were identical, purely via the scale depndent anisotropy that we could nevertheless still account for the observed diversity of cloud shapes and morphologies.

Here are some clouds rendered with perspective:

Top: rendered in Mathematica with a custom raytracer, isotropic phase function, single scattering only.

Middle three: rendered in POV ray with single scattering only, anisotropic phase function (courtesey of Eric De Giuli).

Bottom: A subset of a 256x256x128 cloud, rendered in POV ray with single scattering only, isotropic phase function (courtesey of Eric De Giuli).

As you can see, the possibilities are quite varied for rendering 3D clouds. We have barely begun to scratch the surface!

< previous | next >