This document is the hypertext version of an article of the same title, which appeared in the OCtober 1998 issue of Landscape Architecture Magazine.
"The real value of computing for designers and planners is not the ability to produce impressive graphics; but rather to systematically explore alternatives ..."
Landscape architects and planners are now using computers in a wide variety of ways in their practice, research, and teaching. This magazine and others have chronicled developments in Computer Aided Design, including drafting, modelling, rendering and visualization for preparation of presentations and construction documents, Geographic Information Systems for spatial analysis and cartographic production, digital Image Processing, Animation and a host of other design-oriented software. In the past few years, 'Moore's Law' (which says, in effect, that computer capacity per dollar approximately doubles every year) has held, as photorealistic images in milliions of colors, maps and drawings with millions of vectors, models and animations with millions of polygons and disks holding hundreds of millions of bytes, have all become readily accesible on the desktop for a few thousands of dollars.
This access to huge disks and fast processors -- along with increasingly sophisticated software -- continues to make computing ever more attractive to landscape architects. But 'raw power and speed' alone, as useful as they are, are not the real source of the leverage provided to designers by computing. Rather, it's the ability to explore alternatives intelligently, and to get answers to 'What-If' questions in a systematic way. (Faster answers, with more colors, are very nice, to be sure!)
Getting nice results -- graphically rich drawings, maps, renderings, digital movies -- is possible largely because of the complex and rapid mathematical capabilities of computers: computing perspective projections, removing hidden lines, calculating angles of light rays' incidence and reflection, generating fractal surfaces, etc. But more deeply seated in every digital computer than its mathematical calculation abilities is its logical foundation. Every chip from Silicon Valley, from the MOS 6502 in the earliest Apple to the Intel Pentium II-MMX in the lastest PC, is primarily a 'logical' processor, employing 'Boolean algebra' (the mathematics of logical decisions and choices). And every programming language, while certainly enabling complex algebraic calculations, primarily depends on logical operations such as 'IF-THEN' branches and other forms of conditional controls.
This deeply rooted 'IF-THEN' capability in modern computers resonates with designers' quest for alternatives and consequences. Design is not so much about making beautiful renderings (art) or calculating precise results (engineering) as it is a process of exploring alternatives, and asking 'What-If?'
In the early 1980's, it was the computer program Visi-Calc (the first 'spreadsheet') which is credited with bringing on the personal computer revolution precisely because of its ability to support asking What-If? In that case, it was limited mathematical questions of the form "What if we double the unit price ?", or "What if the interest rate goes up 1%?", which are are so critical to business decisions. The ability to enter a changed value -- unit price, or interest rate, etc. -- and to 'instantly' see the results was revolutionary, compared to paper and pencil, ledger sheets and calculators (in the old days you could actually see the figures ripple on screen as the calculations propagated and values were updated).
Today, the kinds, forms and complexities of "What-If?' that are supported by computing are many more varied and subtle. For professional landscape architects, calculating profit percentages, cut and fill ratios, and unit cost discounts are important, certainly, but a range of equally interesting 'What-If' questions can now be computed, such as: "What if we put downlights in the branches of the trees?" "What if we paint the transmission tower blue?" "What if the habitat for that species is destroyed?" "What if the forest fire comes from the west?"...
Many of these kinds of "What-If's" have to do with visual consequences, of course, but not exclusively so. The many kinds of "What-If's" of interest to landscape architects can be broadly divided into several classes:
These may be mixed and integrated (as in a visual simulation, or
animation, of a land use growth scenario), and may be increasingly
multi-media (as in a VR
experience, which implicitly asks the question, "What if we were
really there?")
Image of Park Woerlitz, Germany, from Simulation by Institute for New Media
Purely numeric 'What-If' questions, which can be asked and answered in a spreadsheet format, are still important. Cost-estimates, earthworks cut-and-fill quantities, and special-purpose formulae for everything from hotel room occupancy calculations to skateboard jump trajectories may be called for in the course of a landscape architect's work. The ability to enter a changed value, or several, and get back instantly -- or even within a few hours or days -- the results, feeds in to the fluid nature of preliminary design, and can add essential information for decision making.
Hydrological calculations, ranging from fairly simple summations of area times runoff coefficient, to much more complex formulations with mulitple parameters and non-linear calculations, have become increasingly important for professionals involved in watershed planning. In the EPA- and DOD- funded study of 'Alternative Futures for the Region of Camp Pendleton, California', Dr. Michael Binford, then at Harvard's Graduate School of Design (HGSD), used GIS and other hydrological modelling tools to ask the question "What will happen to peak discharge in the San Juan River basin, if the paved area in the upstream watershed increases in the next twenty years due to planned new housing developments?" The answer -- that peak storm flood events will increase dramatically in frequency and intensity -- should be no surprise to land planners, but is one of many lessons that needs to be given over and over again in the process of informing public stakeholders and decision makers.
( Fig. 1 Illustration of peak hydrograph curves from Camp Pendleton report. Blue line shows Existing conditions, Red line shows conditions in possible futrure scenario with increased urbanization in upstream watershed. Horizontal axis is time, vertical axis is discharge in cubic feet per second of water.)
More visual 'What-If' implications are also now part and parcel of professional practice. Image processing -- the art and science of blending actual photographs with proposed changes -- has now taken 'artist's renderings' of proposed projects to new levels of detail and accuracy. Computing precise viewpoints and perspective parameters, accounting for atmosperic haze and other environmental conditions, and matching existing to proposed conditiions down to the pixel, are everyday undertakings for professional responsible for visual simulations.
In one study for a microwave tower transmission company, Maryland-based landscape architect and computer artist Curt Westergard was able to illustrate the visibility and visual impact of proposed new transmission towers, first by photographing proposed sites from several vantage points, and using photo-montage to digitally insert 3-D CAD models of the structures, and then by demonstrating the impact of different paint colors on the towers ("What if we paint the transmission towers blue, or green, etc?")
(Fig 2 llustration of transmission tower studies from Curt Westergard. Created with 3DStudio Max and PSP software.)
Sacramento, California landscape architect Brent Thrams of the firm Acanthus also uses photography and artful hand-editing, with Adobe Photoshop, CAD modeling and other software, to present photo-realistic images of new construction projects around Lake Tahoe and throughout northern California. Working from architects' plans and panoramic photographs, his renderings of proposed new structures are used in scenic impact analysis ("What if we double the size of this residence, and clad it in darker wood siding, can we mitigate the impacts with landscape screening and other design modifications?") Such detailed visualizations, and the exploration of visual alternatives, are becoming de-riguer for visual impact assessments in the Lake Tahoe region as well as in other parts of the country.
( Fig 3 illustration of Lake Tahoe images from Brent Thrams, Acanthus Inc., using Photoshop software.)
Visualizations based on 3D models can also be used to ask 'What-If?' for special considerations such as microclimate, sun-angle and shadow studies ("What if the ten story building is on the south side of the park?")
Many regions of the country (and the world!) are facing rapid
gowth and land-use change. In these conditions, landscape architects
and planners are often asked to present 'alternative scenarios' for
future development; increasingly, these scenarios, both in their
development and evaluation, are part of community-involvement
projects. Spelling out consequences -- often long range, sometimes
invisible or subtle -- for the general public, is a difficult task.
Professor
Carl Steinitz, of the HGSD, and a number of his students and
colleagues, have accumulated a wealth of experience in these
'alternative scenario' studies, bringing together synthetic (design)
approaches and analytic (scientific) methods to ask a whole series of
'What-if' questions. ("What if population doubles in the next twenty
years? What if the development is scattered over the landscape by
uncontrolled 'sprawl' processes? What if it's concentrated in certain
zones and carefully designed? ")
A series of studies by Steinitz and colleagues, including 'Alternative Futures for Monroe County Pensylvania', and the above cited Camp Pendleton study, have used a range of techniques -- numeric, visual, animations -- to convey their messages.
Fig. 4 Map of 'Landscape Ecological Pattern', from CampPendleton study by Steinitz, et al Created using ESRI Arc/Info software..
Professor David Hulse, of the Institute for a Sustainable Environment, at the University of Oregon, Eugene, has followed a similar approach in his group's study of "Possible Futures for the Muddy Creek Watershed". In addition, they've made their results available over the web (see http://ise.uoregon.edu) as a further step towards involving local (and remote) citzenry in envisioning and evaluating alternatives. Through these web pages, a variety of alternative future scenarios, their rationale and predicted impacts, can be explored and compared.
(FIg 5 Image of Muddy Creek web page from David Hulse / ISE, enabling viwers to select and compare existing and possible alternative conditions. Created with MacGIS and other software.)
Landscape architects, educators and computer scientists at the Center for Landscape Research (CLR) at the Universioty of Toronto have developed landscape simulation and visualization software (POLYTRIMS, CLRVIEW) which they have used extensively in research and teaching, and a variety of international collaborative projects. Among many projects, they have used the software to address questions such as "What if various height restrictions are / are not implemented in downtown Ottawa?" to measure impact on such factors as visibility of the monumental capitol buildings and incident sunlight on public open spaces.
Landscape architects
Eckart
Lange and
Sigrid
Hehl-Lange, at the
Instutute for National,
Regional and Local Planning (ORL) at ETH-Zurich (The Swiss
Federal Institute of Technology) , have used high-powered SGI
computers and the software developed at the CLR to for landscape
visualizations (See Figure 8 below, e.g.) , including
a
series of projections of landscape change in Switzerland.
Professor
Dick Klostermann, a planner at the University of Akron, Ohio, has
developed a software system designed to augment traditional GIS
facilities with more interactive numeric and logical modelling. The
system, called appropriately enough 'What-If?', has been used in
public presentations and public participation planning meetings to
develop and evaluate planning and growth policies for the region of
Cincinnatti, Ohio. Users can enter a number of assumptions about
growth rates, employment and economic trends, demographics and other
future-influences; the software uses existing polygon-based GIS land
use maps to conduct a classic suitability analysis, and produces
either projections of land use change, based on allocation of future
needs to available resources, or evaluations of proposed scenarios,
both based on the preliminary suitability analysis. This way,
planners and citizens can ask questions such as "What if we build new
sewers south of town?" or "What if we re-zone the agricultural land
to permit cluster housing?", and see the results, both as maps and as
charts and other numeric analyses.
(FIg 6 Image of screen-shot from 'What-If', software from Richard Klostermann)
The visual images described so far as answers to 'What-if questions' are static -- single frames, like photographs, or at best series of several of these. Computers, however, can also be used to generate animations and other, more interactive representations. Computer capabilities now include creating animations and digital movies, produced as eye-level paths through 3D models, or as 2D frame-based animations using software like Adobe Premiere and others. "What-if we approach the new site from the south?", or "What if the access road is curved with the contours" are visual choreographic/experiential questions, which are routinely being asked and answered using video animations.
In exploring alternatives with dynamic systems, such as the weather, or wild fires, or beach dynamcs, animations and other simulations can be more vivid and informative than single frames. Some software, such as 'Stella', has been designed specifically to create dynamic simulations, where several input parameters describe such characteritics as growth, or death, or infiltration or evaporation processes; the simulations so created are allowed to run through multiple time steps, and the results then examined.
Mike Flaxman, a doctoral student at HGSD, is using computer simulations to study the impacts of forest fire management strategies. Coupling GIS base maps, photo image processing, QuickTime movie production and dynamic programming using 'cellular automata' software ('StarLogo'), Flaxman is producing a decision support tool for land use management and risk analysis (enabling users to ask "What if we clear cut for 100 feet around all structures?, and to see both 'What will it look like?', and 'What effect might it have on fire spread?")
In the images below, a wild fire starts as a small red spot in the left hand side of the image, then grows to the east . In each frame, green represents unburnt vegetation, red represents active fire; black represents burned-over vegetation.
(Fig 7 Image of Fire Simulation movie from Flaxman . Frames from a dynamic simulation of Fire spreading over terrain, fire spread a function of vegetation type and other factors. The top strip of 3 images represents fire-spread without mitigation; the bottom stri pof three 'with mitigation' Simulation created using 'StarLogo' software )
The examples cited above are just a representative handful from the many projects by landscape architects and planners around the world using computers to ask "What-If"? in a widening variety of ways.These and other allied professionals have benefited greatly from the advances in digital technology including, CAD and GIS, image processing, and sophisticated software techniques for simulation and visualization. Although many designers still find they they are more comfortable quickly generating and illustrating ideas using old-fashioned graphics and 'back of the envelope' sketches, when it comes to systematically exploring alternatives and elucidating consequences, the leverage provided by computing is rapidly coming to be invaluable.
Fig 8. 'Integrated Image' , making visible plume of pollution, in synthesized image of landscape (scanned topographic model draped over Terrain Model, with image processed trees in foreground) Eckart Lange & Siggrid Hehl-Lange, ORL-ETH. See the MPEG movie at http://www.orl.arch.ethz.ch/%7ELange/rheintal/rhein.html
About the author: Stephen M. Ervin, Assistant Dean for Information Technology, and Lecturer in the Department of Landscape Architecture, Graduate School of Design, Harvard University teaches and conducts research in the areas of design, computing, media and technology, including Geographic Information Systems, Computer Aided Design, Image Processing and their intersection. He has taught and lectured worldwide and published articles in a number of journals including Process Architecture, GIS World, Landscape Architecture Magazine, Computers, Environment and Urban Systems, Computer Graphics and Applications, and others The founding chairman of the American Society of Landscape Architects' Open Committee on Computers in Landscape Architecture, he holds a Master's degree in Landscape Architecture from the University of Massachusetts at Amherst and a PhD in Urban Studies from the Massachusetts Institute of Technology.