Digital scene modeling provides a powerful tool for the simulation of visual and infrared imagery because it allows control of all aspects of the scene. Specific targets can be defined and placed in arbitrary back-grounds, and specific atmospheric conditions and diurnal effects can be included and changed with relative ease, allowing a comprehensive study of factors critical to a real-world scenario. A major stumbling block to digital simulation has been the computation load required to process scenes containing a realistic representation of natural detail. The traditional approach to computer modeling of natural scenes represents natural detail explicitly, requiring a complex geometric data base which is costly to render. An alternative approach to modeling natural scenes is to use a small number of surfaces to define the major geometry of scene features and to use texturing to imply surface detail. Techniques developed at the Grumman Corporate Research Center use a few simple surfaces to define major scene features, such as hills, trees, and clouds, and a mathematical texturing function to define minor topographical detail by modulating surface shading and translucence. This technique pro-duces realistic visual images with much less computation than the traditional approach. Continuing research at Grumman in cooperation with the Keweenaw Research Institute and the U.S. Army Tank Automotive Command (TACOM) has extended the use of this technology to the simulation of infrared imagery by using the statistical characteristics of measured IR data to control textural shading on scene surfaces. This allows the simulation of visual and IR imagery using the same geometric data base and provides a cost-effective tool that can be integrated in a comprehensive target acquisition simulation system which will include missile dynamics, multi-sensor simulation, image processing, and pat-tern recognition.