1.1.1 Reasons for Modeling
1.1.2 Dimensions of Modeling
1.1.3 Techniques for Digital Modeling
1.2.1 Coordinate Systems
1.2.2 Map Coordinates
1.2.3 Pixels and Color Space
1.2.4 Points, Lines, Polygons, and Curves
1.3.1 Surfaces
1.3.2 Solid Models
1.4.1 Organization: Repetition, Combination, et al.
1.4.2 Boolean Operations: Intersection, Difference, Union, et al.
1.4.3 Procedural Operations and Scripting
1.5.1 Ray-tracing
1.5.2 Multichannel Textures
1.5.3 Procedural Textures
1.6.1 Lighting and Shadows
1.6.2 Camera Frame and Viewpoint
1.7.1 Database Models
1.8.1 Generation
1.8.2 Movement Through
1.8.3 Movement Of
1.8.4 Interaction With
Figure 1.29 Animated Sunlight and Ambient Light - University Commons Model
Figure 1.30 Animated Water Surface
This book is about modeling the landscape, and so it has both an action-oriented purpose - modeling - and an object-oriented one - landscape. Modeling simply means making representations, such as drawings, paintings, and cardboard mock-ups; or, more specifically, using digital computers and computer software to organize information in the form of numbers or bits, then creating images on a computer screen or printed on paper; or creating a series of images to form an animation; or even producing a three-dimensional artifact, such as a physical model created by a numerically controlled machine. Modeling by computer is similar in some ways to drawing or painting with pencil or brush, but is quite radically different in other ways - the differences are mostly what this book is about.
Landscape means the natural world, in which we live, garden, work, and build, including both natural systems such as plants and weather, and also built systems, such as roads and cities. Though we may sometimes speak of "the landscape," that is misleading, as there are many different landscapes in this world, and many different perceptions of them. In this text, "the landscape" is used in the same spirit as when we speak of "the human race," meaning to focus on the commonalities and shared attributes, but without ignoring or demeaning the variety and individuality to be found within it.
Four essential elements of the landscape - landform, plants, water, and the atmosphere - are the focus of this book. The first three are the traditional palette of landscape architects, and are the essential components of the natural world, without people or buildings. Of course, in the real world that we live in, the landscape includes structures of all kinds, including buildings and bridges and cars, and a wide variety of animals whose activities are vital to the function and look of the landscape. There is a vast literature on using computers to model buildings and structures, using Computer Aided Design (CAD) software, and while this book assumes some familiarity with those ideas, it does not focus on making models of buildings.
... (Note: This Website contains abbreviated text. For
the complete text,
click here to order the Landscape Modeling
book.)
Figure 1.1 The landscape of the Fens, in Boston, by landscape architect Frederick Law Olmsted, bridge by H.H. Richardson. Most landscapes are made up of varying proportions of landform, including rocks, vegetation, structures, water, atmosphere, and animals including people.
There are many reasons why people need or want to make models of landscapes and landscape elements, and there are many ways of doing so. Modelers include landscape architects, garden designers, architects, planners, engineers, illustrators, scene designers, and others who are engaged in synthetic design processes. Some modelers make models so as to portray landscapes as "scenery" like classical landscape paintings do, or to be used as backgrounds, much like stage sets, or environments for computer games. In quite another vein, scientists and planners may seek to model landscapes and landscape processes so as to be able to simulate or understand them, exploring scientific hypotheses, or measuring aspects of quantitative simulations, such as soil erosion, hydrologic process, or vegetative succession.
... (Note: This Website contains abbreviated text. For
the complete text,
click here to order the Landscape Modeling
book.)
Figure 1.2 "Texas Redbuds" Landscape Painting by A.R. McTee c.1935. Here as in all landscapes, landform, vegetation, and water combine to create a sense of place. The landscape painter's eye is focused on composition, texture, color, and light.
Figure 1.3 Multiple representations: a. Photograph
of a view down an allee of trees in a garden outside of Rome. Italy. b. A black-and-white
"sketch" made by image processing filter operations. c. A computer generated
view of a 3D model.
Figure 1.4 Wireframe representation of the University Commons Model in 3d Studio Max. The abstract colored lines on a computer screen are indicators of landscape elements.
Figure 1.5 A view across raked gravel into the pine forest, and stream with stone bridge, at the Ginka-kuji temple, Kyoto, Japan. Photograph courtesy of Christian Tschumi.
Figure 1.6 The 'Blue Steps' at Naumkeag, a composition of treees, shrubs, landforms, structures, and water by landscape architect Fletcher Steele.
The "real world" in which we operate, the alternative worlds designed by landscape architects and urban planners, and even the imaginary worlds imagined by computer games designers or stage-set creators, is usually thought of as being three-dimensional (3D) - occupying the spatial dimensions X (width), Y (breadth), and Z (height, or depth). This conventional coordinate system, attributed to Euclid and DesCartes, enables detailed descriptions of solid geometry, and is the underpinning of all Computer Aided design (CAD) and Geographic Information Systems (GIS) representations. At the same time, the most common representations used, certainly for centuries and still today, are only two-dimensional (2D), such as maps, plans, images on flat paper, or computer screens. The art and science of perspective projection, representing three dimensions in only two, has been developed by painters, first three or four hundred years ago, and more recently by computer scientists in the last thirty or forty years.
... (Note: This Website contains abbreviated text. For
the complete text,
click here to order the Landscape Modeling
book.)
Figure 1.7 Cartographic projection forces a spherical
surface to a rectangle, necessarily distorting it in the process.
Figure 1.8 Three frames from an animated walkthough
of the historic Park WŸrlitz, near Dessau, Germany. Courtesy of the Institut
fŸr Neue Medien, Frankfurt, Germany.
Figure 1.9 Complex forms created by procedural operations (spiral twisting, e.g.) and combinations of simple elements (in this case, stacking, and decreasing size).
The rest of this chapter is a summary review of basic computer modeling and computer graphics terms, tools, and concepts. Readers who are intimately familiar with 2D, 3D, and 4D modeling may skip ahead to the next chapter which starts the landscape-specific treatment, but since terms that are used throughout the rest of the book are introduced and defined, this summary is recommended for all. It is not intended to be a comprehensive review, for which a good reference text on computer graphics or digital media should be consulted. It assumes a general level of knowledge about computers, and such terms as memory, disk space, processor speed, and a beginning-to-intermediate level of knowledge with 2D and 3D modeling and rendering software in general. Although step-by-step tutorials are provided for many of the topics in the next four chapters, they are not software- or platform-specific.
... (Note: This Website contains abbreviated text. For
the complete text,
click here to order the Landscape Modeling
book.)
Figure 1.10 Computer model of the Department of Housing and Urban Development in Washington D.C. HUD Plaza Improvements, Martha Schwartz, Inc. Courtesy Martha Schwartz, Inc.
The most common models we know of in the design world are 2D: drawings, photographs, maps, plans, sections, projections, et al. In analog representations, much of the power and variety comes from the diverse media, ranging from rough paper to mylar to photographic film, and the tools, including cameras, bristle brushes, technical pens, felt tip markers, and others. The science of computer graphics and the related technologies of digital printing have not supplanted these - especially in such ergonomic matters as portability, ease of use and reliability - but have offered surrogates for almost all of them.
... (Note: This Website contains abbreviated text. For
the complete text,
click here to order the Landscape Modeling
book.)
Positions in 2D space, whether on paper, the surface of the earth, or a computer display screen, can be described by a pair of numbers, representing distance along two orthogonal axes, whose crossing is labeled the origin, and has a value of zero on each axis. In the Cartesian system, these coordinates are typically called x for the horizontal, or width dimensions, and y for the vertical, or height dimensions. On a spherical surface, like the earth, they are called longitude and latitude; in a two dimensional image used as a texture they are sometimes called u and v, or s and t. Two dimensional coordinates can also use the polar system, in which any position has an angle and a distance from the origin; these are sometimes encountered in landscape work, as in the descriptions of land surveys, for example.
... (Note: This Website contains abbreviated text. For
the complete text,
click here to order the Landscape Modeling
book.)
Figure 1.11 a. Cartesian coordinate axes; each point
is represented as an (x,y) pair. b. Polar coordinates; each point is represented
as (distance, angle)or (r, ¯) from the origin.
In the special case of map coordinates, indicating positions on the earth's surface (or any sphere, such as a planet or moon), there are two additional considerations: units and projections.
... (Note: This Website contains abbreviated text. For
the complete text,
click here to order the Landscape Modeling
book.)
Figure 1.12 Three coordinate systems shown on USGS
topographic quad sheet: Geographic (longitude and latitude), UTM (in meters)
and State Plane System (in feet).
Figure 1.13 Three different projections of the land mass of Greenland: Mercator, Geographic and Peters projection.
Whereas the input data for landscape modelers may come in a variety of projections, units and systems, the output is much simpler - almost always a raster array of "picture elements," or pixels, which make up an image on a computer screen (or printed page). Pixels are usually measured by rows and columns, and the spacing between them, in dots per inch, or dpi (also measured as dot pitch, which is the actual distance between pixels on a display device). Pixels, and display screens, may also have an aspect ratio, which is the ratio of width to height. A perfectly square screen with pixels equidistant in both directions would have an aspect ratio of 1.0; most actual display devices and pixels are slightly elongated in the y-axis, so the aspect ratio is slightly less than 1.0. All computer graphics, whether line drawings or photographic images, are represented in pixels. The total number of pixels, for a display screen, or the dpi of a printed image, are sometimes referred to as the resolution of the display or image. Most computer screens are around 1000 by 1000 pixels (for one megapixel resolution), and between 72 dpi and 100 dpi. Printed images are often around 600 dpi or more for ordinary laser printers, and 2000 dpi or more for high quality printing.
... (Note: This Website contains abbreviated text. For
the complete text,
click here to order the Landscape Modeling
book.)
Figure 1.14 a. 24-bit (full RGB) color. b. 8-bit (indexed)
color. c. 8-bit gray scale. d. 1-bit (monochrome) image.
A pixel may correspond to a point in a 2D or 3D model, but because of scale variations, a pixel may also represent far more than that. In a satellite image, for example, a single pixel may represent an area of 30m x 30m on the ground; or in a perspective view, a distant tree in the background may be reduced to a single pixel. In general, pixels are the final output result; 2D and 3D modeling works with more mathematical coordinates and constructs, which only become pixels when rendered to a screen.
All 2D and 3D geometric models are composed of some combination of:
... (Note: This Website contains abbreviated text. For
the complete text,
click here to order the Landscape Modeling
book.)
Figure 1.15 A typical site plan, constructed in a CAD or GIS system. Street lines, parcels and building footprints are polygons composed of line segments; building centroids and other spot features are represented as points.
Elaborating slightly on this foundation, curved lines and curved polygons may be added in; although any curve can be approximated by a series of short straight lines, (or on screen, by a series of pixels), curves such as circular and parabolic arcs, or polynomial spline curves, or curved areas such circles or ellipses, may be represented in a purely mathematical way (with a center point and radius, for example).
... (Note: This Website contains abbreviated text. For
the complete text,
click here to order the Landscape Modeling
book.)
Figure 1.16 Data tables from a GIS system showing the features - points, lines and polygons - illustrated in the image at left. Each row represents one item; each column is some attribute about the item, includng its geometry (the X,Y,Z coordinates, which are hidden in this case), and other non-graphic attributes.
Whereas 2D models are described in 2D planar space, 3D models occupy a volume of space. A third axis, usually called Z, is orthogonal to the first two, and each point has three coordinates, expressed as a triple (x,y,z). This mathematical description encompasses what we think of as the geometry of the real world, in which things have width, depth, and height, and corollary properties such as volume, mass, and centroid. Among three dimensional objects, there is an important distinction in computer modeling: between pure surfaces, which may exist in three dimensional space, and are more than a single flat plane, but are mathematically infinitely thin, having only a front side and a back; and solids, which have thickness, and an interior and exterior. As an example, 4 points in space may be connected to form a surface (a curved hyperbolic paraboloid) or a solid (a tetrahedron).
Surfaces are created by many modeling systems by default, since they are easier to represent and manipulate than solids. A simple planar surface, such as a rectangle, or circular disk, is usually only a surface. For rendering purposes, it is necessary to determine the surface normal, a vector that is perpendicular to the face of the surface; this vector determines which direction a surface is facing. In some systems, it may be necessary to explicitly compute these, or to force the object to have two sides visible; otherwise, one face of the surface may be invisible to the rendering system.
Figure 1.17 The same four points may define a surface (a quadratic patch) or a solid (a tetrahedron).
Solids are more complex than surfaces, since they have thickness and other attributes, but this makes them more robust for some modeling purposes. (They can be sliced through, for example, and their mass or center of gravity can be computed.) The simplest solids are created by just extruding a 2D polygon by some amount (sometimes called 2.5D); others are created by sweeping or revolving a profile; and more complex ones can be created by using various modeling tools. Most modeling software offer a selection of geometric primitives including cubes, spheres, cylinders, and cones which can be parametrically varied to produce a wide variety of shapes alone and in combination.
... (Note: This Website contains abbreviated text. For
the complete text,
click here to order the Landscape Modeling
book.)
Figure 1.18 Image of a 3D solid model of the Villa Rotunda, by Palladio. As with too many architectural models, it is floating in a void, with no landscape elements to ground it.
Building solid models, or surface representations, can be a process requiring primarily eye-hand coordination, using various primitives in the modeling software to build up a complete model, mostly by mouse-clicks and menu selections. But many real world environments have systems and regularities about them, including repetitive elements, grids, or other rules of alignment, which can make computer modeling especially appropriate and facilitate accurate constructions. Grid snaps, duplications, controlled offsets, and regular arrays are operations offered by all modeling software which can be used to create landscape elements that actually form space and give design form to the composition.
Linear repetition, specifying a spacing between and a total number of elements, can be used to create an allee of trees, or rows of lamp posts along a street; regular rectangular arrays and circular arrays may organize shrubs or stone bollards. The overuse of strict geometry and symmetry leads to static, overly formal compositions, but some organizing systems and regularity can give a controlled, navigable sense to landscape designs.
... (Note: This Website contains abbreviated text. For
the complete text,
click here to order the Landscape Modeling
book.)
Figure 1.19 Simple lollipop trees combined in various
ways to create space-forming enclosures: a rectangular array, or bosque; b.
a double row, or allee; c. a circular bosque d. a clearing in a thicket.
Simple grouping causes subelements to be treated as one, but does not actually fuse the elements together - they can always be ungrouped. An additional set of modifiers are available in some modeling systems, which enable the so-called boolean operations of intersection, union, and difference. (These are named after George Boole, the French mathematician and logician, who described the logical operations and and or, first as operations on sets, and then as the primitive operators in binary arithmetic, which is the theoretical basis on which all digital computers operate.)
... (Note: This Website contains abbreviated text. For
the complete text,
click here to order the Landscape Modeling
book.)
Figure 1.20 Boolean operations. The brown cyliner (A) and grey rectangular solids (B) are combined to create: A minus B = ("Lincoln log" form); B minusA = (blocks with cylindrical notches cut out); and A intersects B = (the small brown cylindrical "chips" from the log).
The combining operations, grouping, and boolean operations give a great deal of power to modeling, but still require hand-control of every individual step. In many modeling systems, you can gain even greater power for automation of tasks and creation of complex models, by using a built-in scripting, macro, or programming language. Sometimes a general purpose programming language such as C, Basic, or Lisp can be used to control the modeling environment, sometimes a unique proprietary programming language is required. In any case, the basic elements of programming: assigning variable names and values, using arithmetic, trigonometric, and other operations and functions, conditional control with some kind of "if-then" statements, and some form of repetition, are all available in some form or other. Writing programs, or scripts, or macros (which are all equivalent terms), can be very helpful in generating variations, where several variables are given values to create, for example, stairs with varying numbers of steps, and varying riser and tread dimensions, or more complex forms, such as branching trees or eroded landscapes. Simple programs are usually easy to get started with; more elaborate programs may require study, effort, and reference to a good text on programming.
Figure 1.21 A fragment of a program in the JAVA language, which calculates and displayscolors for a sunrise simulation.
The process of creating a 2D image from a 3D model is called rendering. Some modeling software has rendering integrated into it; some software is for rendering only, taking input in various standard formats. In all cases, a virtual camera is located somewhere in 3D space, looking at the model, and an image is created onto a picture plane located in virtual space between the camera and the model elements. The projection of the image may be perspective, or it may be one of the orthographic projections, such as a plan, elevation, or axonometric. Some rendering systems enable you to specify camera lens characteristics, such as focal length; this is useful when you are trying to match a 3D computer model with real world photographs. The size of the image is specified in pixels, width and height, and the basic process of rendering is to determine what color each pixel in the image should be. A wide variety of rendering algorithms exist, and this has been an area of research and invention in computer graphics for many years.
Many rendering systems, especially those that seek to create "realistic" images, use some form of ray-tracing. In ray-tracing, a three-dimensional vector, or ray, is simulated as if it were passing from the camera, through each pixel of the picture plane, and into the volume of the 3D model. When the ray intersects an object, the pixel's color value is set to whatever color the object is, modified by the presence of lights, and shadows, and surface texture characteristics such as bumpiness, or shininess, etc. Ray-tracing is particularly effective at modeling reflectivity, transparency, and refraction, since the ray need not stop at the first object it hits, but can be continued on, bouncing through the 3D model, so that an object defined as a mirror, rather than creating pixels with its own color, will get its surface color from the next object hit, reflecting that object's color.
... (Note: This Website contains abbreviated text. For
the complete text,
click here to order the Landscape Modeling
book.)
Images are the end result of modeling and rendering, but they may also be used as inputs, as well. A digital image - created by a digital camera, or a scanner, for example - can be manipulated using image processing software to change pixel colors, scale and change resolution, apply filters and special effects, and so on. These images can also be used as textures, applied to objects in a 3D model, to give color and other surface attributes to geometric objects.
... (Note: This Website contains abbreviated text. For
the complete text,
click here to order the Landscape Modeling
book.)
Figure 1.22 A typical "color picker" from a CAD, modeling,
or redendring program, shows colors andshades along with corresponding RGB and
CMYK values. From AutoCAD.
Figure 1.23 a. 3D models of primitive solid objects:
cube, sphere, sylinder, cone, pyramid, torus and plane. b.Simple solid colors
applied, with no lighting (only ambient light). c. Various textures and effects
applied and rendered with ray-tracing: brick texture on cube, reflective surface
on sphere, wood etxture on cylinder, transparent blue on cone, bumpy surface
on pyramid, chrmoe reflectivity on torus, and bitmap of a leaf on the plane.
Some simple repeating textures, like bricks in a running bond, or complex textures like the veins of marble, or puffs of smoke, can be procedurally generated by computer, without any need for a photographic image. These procedural textures are usually generated based on a small number of input variables, or parameters. Some systems may provide a mechanism for writing your own procedural textures. Such procedural textures are sometimes called shaders, and many have been developed for the advanced rendering system known as "RenderMan," which is available both in commercial packaged form, and as an open-source specification, so that some free or public-domain software can use its formats.
... (Note: This Website contains abbreviated text. For
the complete text,
click here to order the Landscape Modeling
book.)
The end goal of most modeling is to create a presentation, either of an image or several images, an animation with many images, or sometimes a 3D artifact such as a computer generated cardboard or resin model. (The technology of computer aided manufacturing or CAD/CAM, is beyond the scope of this text, but many of the 3D modeling techniques are relevant to the preparation of files for computer controlled fabrication machines.) For images, the content is most important, but presentation is important too.
... (Note: This Website contains abbreviated text. For
the complete text,
click here to order the Landscape Modeling
book.)
All renderers are sensitive to the lighting conditions on the model, and all recognize at least two distinct kinds of light: ambient light, which is directionless, overall illumination; and directional light, which has location, intensity, and beam properties, such as the parallel rays of sunlight, or the cone shaped beam of an electric spotlight. Often, multiple light sources must be placed in a scene to provide the illumination and emphasis desired .
Shadows are the inverse of lighting, and also serve to add depth and atmosphere to renderings. Strong shadows are created by direct light, whether originating from sunlight or from artificial fixtures. Indirect light, as on a hazy day, produces only very weak shadows, or none at all.
Shadow casting is a computationally expensive operation, and so it is often foregone, especially in quick studies or animations, where rendering time is an important consideration. There are several short-cut methods of casting approximate shadows, such as shadow maps which can also be used. (See Chapter 5 for more on lighting and shadows in the landscape.)
Figure 1.24 Photograph of a golf course landscape; gray clouds and indirect light contribute to luminous colors, few shadows.
Framing a picture is an essential part of the composition process. Choosing between landscape mode (a rectangle with a longer horizontal axis), portrait mode (a longer vertical axis), or even a perfect square, is a first step. Many landscape images naturally look best in a landscape format. Historically, one aspect ratio has dominated computer-generated visualizations: the 4:3 ratio of most computer display screens (640 x 480 pixels in early models, 1024 x 768 or 1600 x 1200 more recently). This rectangle still is not really very elongated, and so some landscape representations have tended for even more extreme aspect ratios, such as 8:3, so that two frames fit vertically in a standard computer window. The advent of high definition digital television (HDTV) has brought its new standard aspect ratio of 16:9 (nearly 2:1) into regular use, a format more suited to landscape scenes (and large movie screens). See Chapter 6 for more on frames and viewpoints in landscape modeling.
Figure 1.25 The effect of various frames on an image.a)
landscape (5:3) b) Portrait (3:5) c) Square (1:1) d) Panorama (4:1)
One benefit of making digital landscape models, beyond the range of visualization and presentation options, is that the database created may be used for additional analytical and calculation purposes. In addition, other existing digital data base information can be incorporated and used to inform and shape the model.
Although 2D and 3D models are composed primarily of the graphic and geometric primitives surveyed above, they are usually organized by the software into a database including higher level aggregations, including groups of elements as named objects. These objects may have attributes, such as color, which are stored in the application's file, but in some systems you may also be able to link to attribute information stored in external files. A geometric file of a forest, for example, may include each tree's position as a 3D spot location, matched to a terrain model's elevation, but information about species, and size, and other characteristics such as age, or potential harvest dates, may be stored in a separate database file. When the two are linked, the rendering system might be able to make use of this additional database information to create appropriate tree textures, or densities, etc.
... (Note: This Website contains abbreviated text. For
the complete text,
click here to order the Landscape Modeling
book.)
Figure 1.26 Visualization of two alternative scenarios
for land-use growth and change. Images at top are derived from GIS data in maps
shown below. Courtesy of David Hulse, The Institute for a Sustainable Envonment,
University of Oregon. Images by David Diethelm.
Although 3D models of landscape are mostly what this book is about, they fail to capture an essential characteristic of landscapes: that they are dynamic. They change over time, daily and seasonally. Our perception of them is often in motion, as we walk or travel through. Using digital models gives us the opportunity to begin to represent these dynamics in ways which are impossible with traditional analog media, by adding the fourth dimension of time and creating animations and other dynamic 4D models.
One way that computer models can capture landscape dynamics is by using the power of procedural, or algorithmic, generation. Plant growth, for example, can be modeled by a computer program with rules about how buds, leaves, stems, and branches form. Then, to create a model of a plant at a particular stage of growth, the generative procedure can be given the desired age, and generate an appropriate model. Using generative procedures can be far more efficient and powerful than modeling "by hand," and makes possible the generation of variations based on some changing parameters. Often, this involves programming. Many modeling and rendering systems have some embedded programming language, or scripting or macro capability; sometimes it may be necessary to use a computer language such as C or Java to create the procedures. Although programming is not for everybody, sometimes it is the best way to achieve fine control, or to create dynamic effects, in your 3D models.
Figure 1.27 Two variants of a branching form (tree?)
generated by a simple computer program, the
JavaTree program included on the attached Website -ROM.
The simplest kinds of dynamic effects to produce with typical 3D modeling software are animations, made by generating a series of still images, each with a slightly changed camera position and view. This results in an animation, or digital movie, which can be played back with appropriate software; at frame rates above about 15 frames/second or more, the human visual system blends these images together to create the illusion of motion.
... (Note: This Website contains abbreviated text. For
the complete text,
click here to order the Landscape Modeling
book.)
Figure 1.28 Diagram comparing a fly-over animation
with a fly through animation. Both include a Canmera in two time positions and
a path element.
The same techniques used to animate camera paths can sometimes be used to animate the motion of other landscape elements. An artificial sunlight, for example, can be animated over a circular path through the sky, and so be used to generate an animation showing changing shadow patterns over the course of the day. In some animation systems, not only position, but all of an object's modeling parameters can be animated over time. In this way, geometric properties such as the diameter of a cylinder, or its length, or thickness can be animated, or non-geometric properties such as a color or texture can be changed over time.
... (Note: This Website contains abbreviated text. For
the complete text,
click here to order the Landscape Modeling
book.).
Figure 1.29 Animated Sunlight and Ambient Light -
University Commons Model
Click here to see the Animation.
Figure 1.30 Animated Water Surface
Click here to see the Animation.
An important form of dynamics in landscapes is their dynamic interaction with observers, or participants, or computer-users. In the real landscape, if you sweep a branch out of your way, it moves, then springs back when you release it; if you press the lever on a gate handle, the gate opens. Using modern technology including hyperlinks, Java, and VRML, "active worlds" can be modeled digitally, which react in various ways when you "touch" them. Usually, this touching involves clicking a mouse somewhere on an image, or using a virtual pointer of some sort in a virtual 3D world. The result can be anything from jumping to another page on a website, or another node in a virtual model, to haptic force-feedback from the model transmitted back to the user through some mechanical interface. The details of VRML and haptic interfaces are well beyond the scope of this book, but some of the modeling techniques herein can be used to make the virtual worlds and landscapes within which these interactions can happen.
Figure 1.31 An interactive World Wide Web page, at
http://www.watershedatlas.org,
using hyper-links, or "hot links" on a map to provide educational information
about the Alleghany River Watershed. Courtesy of Suzy Meyer, Image-Earth,
Inc
Digital landscape modeling, like all other digital techniques, is really less than 40 years old in the year 2000. Many of the techniques presented throughout this book have been developed by researchers, programmers, and practicioners in academia, computer science, the military, the entertainment industries, and the disciplines of architecture and landscape architecture over that time.
... (Note: This Website contains abbreviated text. For
the complete text,
click here to order the Landscape Modeling
book.)
Figure 1.32 a. GIS map of the DELMARVA study. area,
printed on impact matrix line printer, c. 1966. Courtesy of Carl Steinitz. b.
Monochrome green-on-black 3D perspective, created on an Apple ][+ computer,
c. 1980. Courtesy of Stephen Ervin. c. Detail of a tree planting, in an early
version of AutoCAD, c. 1984. Courtesy of Art Kulak. d. Pen plotter drawing of
landform, c. 1985. Courtesy of Steve Estrada. e. Color 3D plot of forest, from
the "Perspective Plot" program, c. 1979. Courtesy of the US Forest Service.
f. Deschutes National Forest. Mt. Bachelor Visual Assessment. US Forest Service:
Office of the Landscape Architect. Printed on pen plotter c. 1979 Courtesy of
the U.S. Forest Service.
Landscape modeling depends upon basic computer graphics, 3D modeling, and GIS software techniques and conventions, but as the next four chapters show, requires attention to the specific challenges of modeling landscape elements: landform, vegetation, water, and atmosphere. As with all representations, landscape models can span a broad range of styles, presentation media, and ranges of realism from "photographic" to impressionistic and highly abstract. Choosing and using the right level of abstraction, and the appropriate medium for presentation and communication, requires both technical knowledge of the media, and professional design judgement. This book attempts to provide the former, only time and experience can inculcate the latter.
... (Note: This Website contains abbreviated text. For
the complete text,
click here to order the Landscape Modeling
book.)
Ashford, Janet, and John Odam. Getting Started with 3D: A Designer's Guide to 3D Graphics Illustration. Berkeley, CA: Peachpit Press, 1998.
Foley, James D., and Andries Van Dam. Computer Graphics, Principles and Practice, 2nd edition. Reading, MA: Addison-Wesley, 1992.
Laurini, Robert, and Derek Thompson. Fundamentals of Spatial Information Systems. London, England: Academic Press, Harcourt Brace & Co, 1992.
Mitchell, William J., and Malcolm McCullough. Digital Design Media, 2nd edition. New York, NY: Van Nostrand Reinhold, 1995.
Mitchell, William J, Robin Liggett, and Thomas Kvan. The Art of Computer Graphics Programming: A Structured Introduction for Architects and Designers. New York, NY: Van Nostrand Reinhold, 1987.
Motloch, John L. Introduction to Landscape Design. New York, NY: Van Nostrand Reinhold, 1991.
Sullivan, Chip. Drawing the Landscape. New York, NY: Van Nostrand Reinhold, 1995.
VonWodtke, Mark. Design with Digital Tools: Using New Media Creatively. New York, NY: McGraw-Hill, 1999.