Sky models are quantitative representations of natural luminance of the sky under various atmospheric conditions. They have been used extensively in studies of architectural design for nearly a century, and more recently for rendering objects in the field of computer graphics. The objectives of this paper are to (1) describe sky models, (2) demonstrate how map designers can render terrain under various sky models in a typical geographic information system (GIS), (3) illustrate potential enhancements to terrain renderings using sky models, and (4) discuss how sky models, with their well-established standards from a different discipline, might contribute to a virtual geographic environment (VGE).
Current GIS hill-shading tools use the Lambertian assumption which can be related to a simple point light source at an infinite distance to render terrain. General sky models allow the map designer to choose from a gamut of sky models standardized by the International Commission on Illumination (CIE). We present a computer application that allows the map designer to select a general sky model and to use existing GIS tools to illuminate any terrain under that model. The application determines the orientations and weights of many discrete point light sources that, in the aggregate, approximate the illumination provided by the chosen sky model. We discuss specific enhancements to terrains that are shaded and shadowed with these general sky models, including additional detail of secondary landforms with soft shadows and more realistic shading contrasts. We also illustrate how non-directional illumination models result in renderings that lack the perceptual relief effect. Additionally, we argue that this process of creating hill-shaded visualizations of terrain with sky models shows parallels to other geo-simulations, and that basing such work on standards from the computer graphics industry shows potential for its use in VGE. 相似文献
The staurolite–biotite–garnet–cordierite–andalusite–plagioclase–muscovite–quartz metapelitic mineral assemblage has been frequently interpreted in the literature as a result of superimposition of various metamorphic events, for example, in polymetamorphic sequences. The assemblage was identified in schists from the Ancasti metamorphic complex (Sierras Pampeanas of Argentina) where previous authors have favoured the polymetamorphic genetic interpretation. A pseudosection in the MnNCKFMASH system for the analysed XRF bulk composition predicts the stability of the sub‐assemblage staurolite–biotite–garnet–plagioclase–muscovite–quartz, and the compositional isopleths also agree with measured mineral compositions. Nevertheless, the XRF pseudosection does not predict any field with staurolite, andalusite and cordierite being stable together. As a result of more detailed modelling making use of the effective bulk composition concept, our interpretation is that the staurolite–biotite–garnet–plagioclase–muscovite–quartz sub‐assemblage was present at peak metamorphic conditions, 590 °C and 5.2 kbar, but that andalusite and cordierite grew later along a continuous P–T path. These minerals are not in mutual contact and are observed in separate microstructural domains with different proportions of staurolite. These domains are explained as a result of local reaction equilibrium subsystems developed during decompression and influenced by the previous peak crystal size and local modal distribution of staurolite porphyroblasts that have remained metastable. Thus, andalusite and cordierite grew synchronously, although in separate microdomains, and represent the decompression stage at 565 °C and 3.5 kbar. 相似文献