f’ght to save

® seemingly PACE 475

•doomed town A double look
PAGE 438 at some personal
Looking toward a gardens
new town site
page 468
PACE 444
page 456 Planning a place
Suroeying the where urbanites
Overlay mapping: aftermath of a
views both histone can play
park-destroying
and current tornado
PAGE 463
An ill wind that
allowed for some
needed change
of records the paths
tom^j^^' midwestem
courtesy of
Chi° Pujita, University of
and^° Dan Hobbs
Steve Hall.

PAGE

Richard ^Johnston 432

Ihejown That Said No: How the Corps of Engineers_Was------------- Robert Royston 438

Steinitz, PauXPai!<£lJ=2W!l?J5^^ 444

the Feel for a New Town Site andJts_De^
Carl
tjggd-Drawn Overlays — Their History and_Prospective__ses Allen J. Share 456

Campbell E. Miller 463

]ornado I: Restoration of a Tornado Ravaged Park
Walter H. Kehm 468
]22?clo II: Recovery and Re-Discovery^^ Resurrection
421
Z^'^onto Creates Bronte Park, Of and For the People
426
P®opie _
475
and Fills
484
La Reviews
'f>dex to Advertisers

PuOll.l-0 O-monl«, in VSoTdTw Am.«=.n Sw.»

jgf, Editor
Gatewood. Advertising and Business Manager
Pickering. Managing Editor „90 E..L ‘'XX
Witr^' Editorial Associate
'^rn Morgan, Book Review Editor
Ela? Steve Hall, Graphic Consultants
Schweitzer, Circulation Manager
Gibson. Assistant Circulation Manager
 Hand-Drawn Overlays:
Their History and Prospective Uses
by Carl Steinitz, Paul Parker, and Lawrie Jordan

The design professions the characteristic elements

have long advocated the and methodology of "over-
necessity of basing design lay processes." The authors
decisions on accurate data, have no doubt there are
analyzed by competent other, possibly older, ex-
professionals. The types amples of each analytic
and scales of data used as step illustrated in this
well as the methods and paper. We are not invent-
techniques of analysis have ing these methods; we have
varied over time. However, a professional literature,
the common practice of albeit a miserably under-
drawing different data developed one.
maps such as slope, soils,
vegetation, etc., of over- Manning, 1912, overlay
laying them, and of making interpretation of land data.
analytic inferences from The earliest evidence of
such composites is the most the use of data overlays as
widely used approach. It an analysis technique in
seems to represent a logical this country is a study done
and obvious basis for ana- for the town of Billerica,
lyzing relationships among Massachusetts in 1912 by
different elements of the Warren Manning land-
landscape. scape architect and one-
When, where, and by time associate of Frederick
whom was this style of Law Olmsted. Billerica, 22
analysis first applied? What miles northwest of Boston,
has been its development was, at the time of the
from its origins to the pre- study, a farming and
sent? What is its future? manufacturing town. Man-
First, we describe the ning, working directly and
historical development of closely with town residents,
hand-drawn map overlay prepared at least four dif-
techniques; next, we pro- ferent data maps of the
pose a variation on the pre- town and one plan showing
vious methods so as to gain recommendations and
the analytic efficiency of a changes in the town's cir-
computer-like approach to culation routes and land
data files, and an increased use pattern. Maps and plan
graphic flexibility. were all drawn at the same
scale and published in the
I. Hand-Drawn Map Over- town magazine, Billerica.
Lays: An Historical An accompanying article
Perspective describing the Billerica
The following is a brief study was published in
and incomplete description LANDSCAPE ARCHITEC- Fig. 1.2
of some past projects that TURE Quarterly, April,
either exemplify or discuss 1913.

444
 Although no direct men- region. Rather, the em-
tion is made of actually phasis appears to be on cir-
employing an overlay tech- culation. The natural land-
nique, the LANDSCAPE scape is, however, de-
ARCHITECTURE article scribed topographically
does discuss soil and vege- and only sites of higher
tation information (Figs. elevation are recommended
1.1 and 1.2) together with for future residential de-
topography, and their com- velopment, while the low-
bined relationship to land lands are set aside for agri-
use. There is little doubt culture and industry.
that Manning combined A series of regional data
and analyzed the various maps at the same scale
Billerica data elements were prepared, followed by
through an overlay process a general land use and cir-
and presented the project culation recommendation
to the town in exactly the map. The maps include a
same manner that many spatial accessibility plan
practicing professionals showing proximity to exist-
would today. ing roads (Fig. 1.3), while
future accessibility is ex-
Dusseldorf, 1912, time- amined as the proximity to
series change overlays. two alternative additional
In the same year, results ring roads (Fig. 1.4). The
of a design competition use of overlays as a tech-
for the city plan for Dussel- nique is again not specifi-
dorf were published in cally mentioned. It does,
Bonder-Katalog Fur die however, seem reasonable
Gruppe Stadtebau der to assume that the several
Stadteausstellung zu Dus- data maps, prepared at the
seldorf 1912. Included were same scale and including
five plans depicting Dussel- accessibility interpreta-
dorf's development from tions, were combined and
1874 to 1912. All were analyzed through overlay
drawn to the same scale. procedures.
By overlapping them one
could easily trace the city's New York, 1923, economic
growth and see which areas and demographic overlays,
were responding more varied base scales.
readily to development In 1929, The Survey of
pressures. Again, no expli- New Tork and Its Environs
cit mention is made regard- was published. The data
ing the use of overlays and gathered and presented for
we can only infer their use the regional survey of New
in the analysis of historic York consisted of many
growth, volumes and represents a
wealth of facts and infor-
Abercrombie and Johnson, mation. Most of its maps
1922, accessibility isolines. and tables were prepared
Ten years later, a re- between 1923 and 1925,
gional planning scheme for and refer to that period.
Fig. 14
Doncaster, England, was In Vol. II, Population,
Pioneering example of data overlays was 1912 analysis prepared and published by Land Values and Govern-
of Billerica, Mass., by Warren Manning [Figs. 1.1 and Patrick Abercrombie and ment, we have an indica-
1.2]. In early 1930s Abercrombie and Johnson produced Thomas Johnson in The tion that overlays were
spatial accessibility maps for Doncaster region [figs. 1.3 Doncaster Regional Plan- part of the analysis pro-
and 1.4], implying the overlay process. ning Scheme. While this cedure.
was an early effort in re- New York was analyzed
gional planning, it typified over time in relation to
many regional studies of population densities in a
the 1920s, Relatively little manner similar to that of
attention was given to the the Dusseldorf study. Two
natural elements of the series of maps display pop-

LANDSCAPE ARCHITECTURE— SEPTEMBER 1976 445

 ulation density and distri- lay technique is made, but
bution for the years 1850, component data maps at
1900, and 1920 at a region- the same scale and format
al as well as a city scale. as the proposed plan are
In the same volume, 1920 shown. Figure 1.9 shows
population densities are the results of a survey
compared to 1923 land which compared the quan-
values. A map of existing tities and location of exist-
distribution of population ing open spaces with pop-
for 1920 is shown (Fig. ulation and location stan-
1.5), as is a ten-color map dards. It locates areas of
which defined the range of excess supply and defi-
land values (Fig. 1.6). A ciency. The open space
third colored map com- plan is a direct result of Fig. 1.5 NYC Population Fig. 1.6 NYC Land Values
bined both land values and combining the two ana-
population density as a lyses of supply and de-
separate analysis (Fig. 1.7). ficiency.
It is clear that this map is
the result of a simple over- Tyrwhitt, 1950, an explicit
lay which compares two description.
related factors on a geo- The 1950 English publi-
graphic basis. ,cation, Town and Country
Planning Textbook, pro-
New York, 1929, base map vides the first explicit dis-
underlays for reference. cussion of "the overlay
Similar examples are technique." The book con-
found in other volumes of sists of a series of articles
the study, but perhaps the dealing with planning and
best example is in Vol. design methods and issues.
Fig. 1.7 NYC Population/Land Value
Composite
I, Graphic Regional Plan Written and compiled by
of New York and Its En- a number of different auth-
virons. At the end of the ors and published as a text-
book are five general--- maps
- book, this publication
of the region drawn to the might well be considered
same scale. Each map is an accurate description of
printed on a base map of the state of the art of
relative topographic eleva- town and country plan-
tion and shows in simpli- ning at the time. In Chap-
fied form the existing and ter VII, "Surveys for Plan-
proposed features of one * . .. ■
ning" by Jacqueline Tyr-
important element of the whitt, we find a clear ex-
graphic plan, such as the #
ample and description of
plan of the park system the typical overlay process. Fig. 1.8 NYC Plan of Existing and
(Fig. 1.8). Each can be re- The author explains how Proposed Park System
moved from the pocket all maps are drawn to the Fig. 1.9 NYC Open Space Deficiency Survey
part of the book and over- same scale and repeat one
laid on a line drawing base general feature as a com- COUNTY Of LONOON Pt AN

map for quick comparison mon reference guide. She

of the different elements of also states that "for actual
the plan. planning purposes all sur-
vey maps should be on
London County Council, transparent paper." In an
1943, "standards," excess example, four data maps
and deficiency. (relief, rock types, hydro-
The County of London logy and soil drainage, and
Plan, published in 1943, by farmland) are combined
the London County Coun- and analyzed into one map
cil in anticipation of post- entitled, "Land Character-
war redevelopment, uses istics," (Fig. 1.10). The land
several typical character- characteristics map is de-
istics of an overlay pro- scribed as a "synthesis and
cess. Once again, no ex- interpretation of the fore-
plicit mention of an over- going maps" and as a "judi-

446
 cious blending of the first Philip Lewis as principal
HO CMARACttflISTia
four maps." How better consultant, uses overlays
could a contemporary pro- to evaluate natural re-
fessional explain this pro- sources for the entire state
cess? of Wisconsin. Individual
landscape elements of
Whittle, 1950, limitations water, wetlands, vegeta-
and data handling stra- tion, and significant topo-
tegies. graphy, are combined
In Appendix I of the through overlays into a
same book Jack Whittle composite map (Fig. 1.11)
wrote an article, "The Pre- depicting Wisconsin's "en-
paration of Planning Maps: vironmental corridors" —
Sources, Notations and areas of prime environ-
Methods of Reproduction." mental importance. In-
He begins by describing cluded in the study is a
the limitations of maps, graphic interpretation of
and discusses the difficul- the more detailed analysis
ties of illustrating the dyna- (Fig. 1.12), obviously gen-
Fig 1.10 TyrwnHI-s Land Characteristics eralized for purposes of
mic qualities of an area;
"A pictorial technique of public communication.
two dimensions cannot
deal adequately with pre- Alexander and Mannheim,
sentation problems of such 1962, many factors, hier-
complexity, but there are archical combinations,
two methods in use which, weightings.
although they do not solve Another study completed
the problem, avoid it. The in 1962 applies overlay
first, and more common techniques in a more com-
method, is to use several plex manner. The study
base maps of the area, and was done by Christopher
show a deliberate restricted Alexander and Marvin
amount of information on Mannheim at the Civil En-
each map, e.g., land use gineering Systems Labor-
on one map, traffic flow atory, Massachusetts In-
on another, etc. The second stitute of Technology, and
method uses one base map is entitled. The Use of Dia-
PiJ SUWACI «KTI«
on which the basic matter grams in Highway Route
of the survey (often land Location: An Experiment
jXl lOUiT. WCXJOtA*® * MiTUtf use) is drawn, and places Twenty-six different dia-
K] UUAMIATIOK over this a series of trans- grams, each depicting the
WtTLAND
g n ieic an t t o e >o g r aph t s u r f ac e WATER, parencies, each 'overlay' relatively "utility" or desir-
ig. 1.11 Philip Lewis' composite Map showing one subject." It is ability of the highway
clear from the literature location for one specific
that by 1950 transparent requirement of the high-
overlays were being used way, are considered for
by planners and designers analysis. The study recog-
both as a standard land nizes that not all factors
analysis technique and as a can be geographically over-
presentation method. laid at one time, and pro-
poses a formal procedure
Lewis, 1962, very large "tree" for the combination
area, generalized. and weighting of overlays.
In more recent times (Fig. 1.13).
there are many examples This study surpasses tra-
displaying overlay analysis ditional overlay applica-
techniques including appli- tions in several ways; in its
cations to very large geo- use of an explicit weighing
graphical areas. Recreation system for the combination
in Wisconsin, published in of different components, in
1963 by the Department of the idea of superimposing
Resource Development, the different diagrams or
... ■■environmental Corridors" State of Wisconsin, with maps photographically as
1.12 Lewis tnvinv

,'c d t c mRER 1976 447

separately timed exposures both in the general litera-
on one print, and in the ture as well as in specific
procedure used to structure projects and studies. De-
the synthesis of the 26 dif- signers rarely describe their
ferent diagrams. As a methods of analysis and
whole this study has had design. This was true in the
much to offer and it has past and it is unfortunately
resulted in one of the rare true today as well. It is not
methodological debates to uncommon in a publication
be found in the design liter- to find maps of existing
ature. conditions followed by a
proposed design or master
McHarg, 1969, "layered" plan, with little discussion
overlay sets, photographic of how the designer ad-
combination vanced from one to the
Ian McHarg directed and other. Is this accidental?
published a number of What does it imply about Fig. 1 13 Alexander/Mannheim Procedural "Tree"
planning studies during the a profession?
1960s that were based on The emergence of over-
an overlay process. In- lays as an analysis tool
cluded in these are the may have been connected
1-95 Richmond Parkway, to a change in the formats
the Plan for the Valleys and types of data avail-
and the Staten Island able to the designer. Early
Study. designers probably visited
The Staten Island Study, the site personally and re-
published in Design with corded basic site data (i.e.
Nature, is one of the most soil conditions, vegetation,
graphically elaborate ex- water, rock outcrops, etc.)
amples of overlay analysis. on one map. They com-
In Design with Nature, piled their own data base
McHarg explains a process as they perceived the site,
of mapping the individual and perhaps analyzed its
elements and then photo- problems and potentials as
graphing these as trans- they proceeded. Because of Fig. 1.14 Utility Diagrams and Their Resultant Graphic Composite
parent prints. The trans- the lack of formalized, re-
parent prints of light and corded data, the designer
dark values are superim- had no choice other than
posed over each other to to gather information di-
construct the necessary rectly. As resource data
composite analysis maps. became available in stan-
These composites are called dardized forms from gov-
"suitability maps" and refer ernment agencies and
to broad land use classifi- others, the designer had
cations such as conserva- more of the data avail-
tion, urbanization, and able as separate mapped
recreation. These were elements. The different
then combined with each specialized data sources
other as overlays to pro- such as U.S.G.S. and the
duce an overall composite S.C.S. each produced their
suitability map (Fig. 1.14). own individual maps which
Because of the range and had to be combined in
number of composite val- some way by the designer.
ues, the final map is dis- Thus, the designer found
played graphically in four in overlays a simple method
different colors and eight of logical association by
different tones within each. which to analyze the dif-
As we reviewed the evi- ferent relationships among
dence which is partly pre- different patterns of land-
sented here, it is obvious scape characteristics, and
that there has existed a in this manner gain a more
serious lack of method- comprehensive knowledge
ological documentation. of the site. Fig. 1.15 McHarg's "Overall Composite Suitability" Map of Staten Island

448
Hand-drawn data map Another possible expla- withstood the test of time color code does not com-
process requires multiple nation for the increased and diverse application. municate this well.
re coding and re drawing use of overlay techniques Possibly, however, it is in-
of maps such as these, a is related to the type and adequate, inflexible, and The hand-drawn data
serious liability when more scale of projects for which inefficient — particularly map process
is required for design they were used and to their given the revolution in In the typical "hand-
decisions. presentation. As designers contemporary graphic tech- drawn data map process"
began working on larger nologies and the ever-in- shown in Fig. 2.1., we usu-
and more complex prob- creasing complexity of our ally make a data map as
lems, they could not as- project needs. follows: first, we get the
semble, analyze or explain Is it reasonable to expect data source (usually an-
the necessary data on a that we will continue to other map), then we define
Hand-Drawn single base map. For pur- rely on manual graphic and make outlines of
Data Map Process poses of clearer graphic methods in the next half- homogeneous zones and
presentation as well as ana- century? And don't changes select a color code. Then,
lysis, the data had to be in graphic technology and we "color it in" just as
mapped as separately com- project needs imply changes children do. Finally, on
binable components. Over- in the organization of over- the light table, we pray that
lays may have been a na- lay processes? We believe we can make sense of it in
tural result of this graphic that the answer is "yes!" combination with a limited
dilemma. number of other data
These overlay methods II. The Hand-Drawn Data maps. I have continually
of analysis have aided the File Process observed this approach at
design professions for well An Alternative several professional offices
over half a century and While working with Bill and schools, including my
continue as a fundamental Johnson of Johnson, John- own. The data map pro-
concept. Their greatest role son, and Roy about three duct, while fine for display
has been to help us realize years ago, the question was purposes, is usually not
that a better understanding raised, "What are the tech- fine for analysis purposes
of the whole is derived nical limits of hand-drawn without recoding and re-
from a knowledge of the overlay analysis tech- drawing. This graphic in-
parts and how they relate niques?" The project was efficiency and inflexibility
to each other. The applica- the evaluation of the na- is a serious liability, par-
tions have become more tural environment of the ticularly as projects get
complex, diverse and tech- Upper Michigan Peninsula. larger and more complex,
nically sophisticated, but A prototype study area had or have more basic options
as we examine the use of to be done on short time to consider.
the overlay approaches and budget. All requisite
from their early develop- base data had been col- The computer data
ment to present applica- lected. The organization of file process
tions, it is clear that the a computer data base was A computer data file
basic methodology and the not feasible. The issue was process approaches the
underlying logic have how to set up a hand- problem differently. One
changed little. We combine drawn graphic analysis sys- still has to get the data
data maps of soil, slope, tem which could flexibly source map (or other data
and other elements in the and appropriately use the source which is already in
same manner that Warren available data to perform a computer-compatible form)
Manning probably did in large number of analyses and one still has to define
1912. We commonly hand- without the expense of re- or locate the various homo-
draw each data element as drawing 200 potential eva- geneous zones. At that
a complete separate map luations. Clearly, the data point the traditional hand-
and then combine maps maps, once drawn and drawn and computer ap-
through overlays — visual- colored, could not be used proaches diverge. Regard-
ly, by redrawing or by as overlays because the less of whether the com-
photographic reproduction graphic codes would have puter encoding process is
_ into some sort of com- to be changed (and the based on a grid or on ir-
posite representing the maps redrawn) each time regular polygons, the com-
analysis. the subvariables were to be puter data file process en-
This apparent stability of used differently in an ana- codes and provides separ-
approach may be a good lysis. For some things, steep able access to each sub-
thing, attesting to an ef- slopes are good, and for variable. Thus, in effect,
ficient, accurate, profes- some things steep slopes one makes not one data
sional process which has are bad; and one slope map map, but a separable, map-

.,.^_SEPTEMBER 1976 449

LAk.
 pable file component for thus, the hand-drawn data FIGURE 2.2
each subvariable. The com- file process. Instead of fol- The Computer Data File Process
puter allows the selective lowing traditional practice
recall and recoding of these shown in Fig. 2.1 and pre-
subvariables, as required maturely making data
by the criteria in any par- maps, we followed a pro-
ticular analysis. A com- cess similar to the com-
puter graphics program can puter approach in Fig. 2.2.
also create a new complete This is illustrated in Fig.
data map (Fig. 2.2). 2.3.
The advantage of the As in the other processes,
computer data file ap- one still had to identify
proach, and its difference data needs, get the data
from the hand-drawn data source maps and define
map process, is its flexi- [;| U S' G S Data Source Map [2] Slope Zones
various homogeneous
bility in analysis. The same zones. At this point, each In computer data file pro-
data variable can be re- subvariable was hand- cess, the computer sum-
coded many different ways; drawn on a separate black mons data in comparable
and the more and varied ink-on-paper overlay of the displays, as shown in Fig.
uses to which the data var- outlined base map. Each of 2.2. However [opposite
iable is put, the more effi- these "plates" was then page] the hand-drawn data
cient the computer-like transformed into a colored file process permits you to
processes become. Thus, in plastic overlay, and these weigh some criteria as
Fig. 2.2, if the slope vari- overlay sheets became the more important than
able is in the data file, we data file.2 others, using colored trans-
can quickly transform the One color was selected parent plastic maps.
initial coding into analyses for all data file overlays. |3| Encoding Key
such as flat = good, or When the individual trans-
steep = good, with simple parent overlays are com-
or complex recoding com- bined the areas that over-
mands. This also offers lap become darker shades
flexibility in displaying of the same color.
data. In order to create a data
To provide the equiv- display map, duplicate
alent two recodings as copies of the same sub-
shown in Fig. 2.2, an ori- variable were used to
ginal hand-drawn data map achieve different shades of
would be redrawn twice, color or black and white.
or combined in a letter or Thus, no added drawing is (4] Water [51 Flat [0-6% slope]
number code, or trans- needed to create the two
formed in the designer's data display maps in Fig.
head. Both lack graphic 2.3.; one shows steeper
clarity and have serious slopes as darker and one
complexity limitations. shows flatter slopes as
darker.
The hand-drawn data Similarly, an analysis
file process composite containing four
The JJR problem was individual overlays would
flexible and efficient ana- display a range of four dif-
lysis. Twenty land use acti- ferent shades of one color.
vities had to be evaluated This in turn can be dup- [6] Medium [7-t5% slope] 17) Sleep ()6+ % stope]
for ten possible impacts. licated inexpensively in
Each of the resulting 200 gray tones.
analysis maps potentially If one thinks of the
needed a different map amount of coloring as the
based on different criteria difference in work between
redrawn from the original the two hand-drawn ap-
data maps. The resolution proaches, one can see that
proposed was an attempt each area of the source
to combine the efficiency map in Fig. 2.1, Step 3,
of a computer-like data file and in Fig. 2.3, Step 3, has
process, with the necessity been "colored in" once for
to use hand-drawn maps; either method. Thus, the [8] Data Display Map |9| Data Display Map
’Flatter is Darker" Steeper is Darker "

450
FIGURE 2.3
The Hand-Drawn
Data Map Process

I?] Stope Zones [3] Encoding Key

(4a] Water [46] Flat [0-6%]

15a] Slooe/Waler [56] Stope

hand work is the same. In matrix, and overlaying locate on a site; and eval-
the traditional approach, them on a line drawing-on- uations of the vulnerability
the product is one colored plastic base map. One can of the site to the impacts
data map, and all sub- then photograph or inex- caused by construction
sequent analysis maps must pensively duplicate the dis- maintenance and any acti-
be redrawn. In the com- play of each combination vitiesS associated with
puter-like approach, the with no added coloring physical change. Both in-
product is as many maps time. When done, the data volve the definition of cri-
as there are subvariables, file element maps are re- teria and result in mapped
but the data work is the turned to the file for the analyses. Both must be
same. Subsequent analyses next use, as in a computer combined in a proposal/
■6a] Composite: I-latter = Darker
do not require redrawing. data file. synthesis.
The data file elements Working backwards
were then used for both the A design analysis approach from what appears to be a
diverse analyses and data using the hand-drawn data traditional practice, we first
display. In the JJR study,3 file process define the problem: the
a matrix was first prepared In our own use of the aims, the site and the pro-
to show which criteria hand-drawn data file pro- gram of physical changes
(data file maps) were to be cedure^ in projects of to be considered for the
used for each analysis. Any greatly varying scales, we site area. We then identify
overlay analysis consisted find it useful to distinguish the analyses by which we
simply of getting the ap- between separate analyses will prepare and evaluate
propriate subvariables from of the attractiveness for our proposed designs. Both
[60] Composite Steeper = Darker the file according to the activities which seek to attractiveness and vulnera-

LANDSCAPE ARCHITECTURE—SEPTEMBER 1976 451

 bility analyses are described for this example (as in FIGURE 2.4 The Hand-Drawn Data File Process
in terms of criteria which Fig. 2.4) consists of 16
in turn define the data colored transparent plastic
needs. We then, and only maps, each separately de-
then, prepare the data file fined, coded and num-
as previously described. bered. Multiple copies have
been made in red and green
An Example as previously described —
Suppose we have a site quite inexpensively. Using
on which we want to eval- these, or any other colors
uate locations for low den- we want to produce, we
sity housing and a shop- then reconstruct the data
ping center; and for vul- maps that all design pre- Pl Slope 16 Soils
nerability to erosion caused sentations seemingly call
by extension grading and for, using the color Xerox
to visual change.6 Step 1 is process or slides.
to identify the criteria for As a graphic and ana-
these analyses which must lytic convention, we have
be supported by data, and found it useful to prepare
their relative importance. the attractiveness criteria
This leads directly to Step composites in green and
2, in which criteria are col- the vulnerability criteria
lected into data needs and composites in red; using
sub-classifications. Step 3 black for the conditions
is to find data sources, to which are absolutely con- (2) Water Water
encode each of the data strained. In effect, we are
elements on a consistent using graphic analogs to
base map, and to prepare criteria and weightings as
the hand-drawn data file follows: black = absolute-
as shown in Fig. 2.4. ly constrained (never!);
A key advantage of the 2 red = very vulnerable
hand-drawn data file ap- (very bad); red = vulner-
proach over the data map able (bad); clear = neu-
method occurs when you tral, not applicable; green
want to weight some cri- = attractive (good);
teria as being more im- 2 green = very attractive
portant than others. If you (very good). It is easy to
[6] Coarse
limit the weightings to communicate that green =
whole numbers, you can "go" and do it; and that
graphically weight by dup- red = "bad" and stop.
licating the necessary data People who can read traf-
file element and inserting fic lights understand this.
the appropriate number of In Step 4, we then pro-
identical maps into the duce the analyses — the
analysis "sandwich." (Re- attractiveness and vulnera-
member, it involves no bility analyses listed in
more drawing and it is Step 1 — using the data
cheap). In the commercial file to select criteria map
analysis, a criterion data components and to create
{yi Pear anrt Organic
item is weighted twice. In the appropriate overlay
the erosion analysis, some composites. These are then
slopes are weighted once reproduced via color Xerox
and some twice. The data on single, color-on-clear-
file approach easily pro- plastic sheets. These, in
vides graphic analogs for turn, can be used as over-
this analytic flexibility. lays as required by the de-
You can follow this ex- sign analyses or presenta-
ample through Fig. 2.5. tion (Fig. 2.5).
In preparing the data Step 5 is returning the
file, we usually make five data elements to the data
copies of each element: file.
1 black, 2 reds and 2 |‘)J S/eep[Z6+%J I /O] Fine-Medium
greens. Thus, the data file Some advantages in design

452
 Among the simpler syn- infrastructure alternative,
thesis steps derived from for example, the changes
Fig. 2.5 and shown in Fig. in attractiveness for com-
2.6 are the following: one merce and housing caused
can identify parts of the by the addition of a major
site that are generally at- new road. Our design work
tractive by overlaying all using computers has clear-
or some of the attractive- ly shown the efficacy of
ness analyses. These can iterative synthesis methods,
also be weighted as be- with rapid evaluation being
fore, by duplicating the the key to trial-and-error
summary analyses. design. This style of probe-
One can identify areas of and-test is difficult to do
multiple-vulnerability by with the traditional data
overlaying all or some of map process, since all ana-
the vulnerability analyses. lyses typically have to be
Again, component analyses redrawn for each alterna-
of these danger zones can tive. In the data file pro-
be weighted graphically. cess, however, only the
One can also compare data element of road ac-
sites that are attractive cess needs to be altered.
with their potential im- All the rest of the data can
pacts; by overlaying the remain unchanged. The
green attractiveness ana- analysis process then simply
lyses on each of the red gets repeated (by using the
vulnerability summaries. duplicating technology)
Thus, “greener" = more and the results can be visu-
attractive and less vulner- ally compared.
able; while "redder” = less Suppose you have some
attractive and more vul- doubts about choosing and
nerable. Ultimately, the de- weighting the criteria them-
sign involves explicit deci- selves. It can hardly be
sions that combine "attrac- easier to discuss several
tiveness" and "vulnera- alternatives and their spa-
bility" in specific (and by tial and qualitative con-
this approach, not neces- sequences. Just "pull" the
sarily quantifiable) judg- appropriate trial "sand-
ments. As a general and wiches" on a light table.
simple rule, however, one There's no drawing to be
tries to keep to the done.
"greener" areas ("go")
rather than the "redder" Some additional advantages
("stop!"). Suppose that the project
Suppose that spatial lo- is to result in a publica-
cation criteria are involved tion, in color. I have often
— for example, to find seen projects in which all
locations a given distance the working maps had to
from a highway, or down- be completely redrawn just
Forest I ;y| 'Mt-4CKJ m wind from a point pollu- for publication purposes.
tion source. If you can Note that the data file ele-
identify proximities as ini- ment maps are just like
tial data needs, you can photo-offset plates. Thus,
treat proximity as a data with the added work of
element, and draw an ap- contacting a good printing
propriate set of maps as shop, they can be trans-
was done for highway ac- formed directly into display
cess in Fig. 2.5. These can data maps for publication,
then be used in the hand in any color, without re-
overlay process just like drawing. Some of the fig-
any other data. ures in this article are an
Open and Agriculture [20] 401 + <n.
Suppose that in develop- example. For several years
ing the design you wish to we have been using com-
"test" the effects of a major puter graphics as the

l a n d s c a p e a r c h it e c t u r e —SEPTEMBER 1976 453

 "plates" (instead of hand bered, and now is the time FIGURE 2.5
drawings) and have con- to apply these new and in-
verted them to color-on- expensive methods to de-
plastic for use in publica- sign. ®
tions and presentations. It
is effective and cheap.
Another helpful hint: do 1. In preparing this section, I
all this as small as pos- thank Lawrie Jordan, Bruce
sible. It saves time and Rowland, Charles Killpack,
money. Why use bigger Henry Foster, and Paul Parker
base maps if they are not for their many diverse contribu-
tions to this work, and David
more detailed? Why not Stinton for his helpful com-
use slides or overhead pro- ments. I thank Bill Johnson and Flat [0-6%) Medmri [7-15 °/c
jectors to fill those walls? JJR for their risk-taking. Some
(How often have we done of the following work has been
this after reducing bigger furthered by a grant from the
maps?) Also remember that Ford Faculty Research Fund,
Department of Landscape Arch-
most publications and re- itecture, Harvard Graduate
ports are SVi" x 11", There> School of Design. (C.S.)
are photographic services 2. There are several compet-
that can make poster size ing commercial products that
effectively perform this step.
blow-ups if they are really We have tested many. Examples
necessary. in this paper have been made
In answer to the ques- by the 3M brand Color Key
tion with which we began, Contact Imaging Material and Fine-Medium Fine-Medium
the Xerox 6400 color or black
I suspect there is no end to and white reproduction pro-
it, once you experiment cesses.
with computer-like data 3. This project is reported in
handling and follow the Johnson, Johnson and Roy,
simple helpful hint — don’t Michigan Natural Environment
Study: Survey Summary, Ann
make hand-drawn data Arbor, Michigan, 1973.
maps — make data files 4. The following is also the
instead. "bare bones" outline of an ap-
Our experience in teach- proach used by my colleagues
and me when working with
ing and in professional computers.
work, and that of the 5. A very useful distinction
people who are already us- developed by Prof. Richard Open and Agriculture Forest
ing this simple method^ Toth, Utah State University.
shows that it is flexible, 6. The reader should not con-
fuse the structure of this ex-
efficient and useful. It is ample with its contents. The
easy to explain on a light contents are no more than ex-
table or with an overhead amples. These are undoubtedly
projector; and it is easy to too simple.
7. Johnson Johnson and Roy,
change. It brings across Inc., Michigan Natural En-
many important aspects of vironmental Study: Survey
design; among them the Summary, January, 1973; John-
direct relationships between son Johnson and Roy, Inc.,
criteria, analyses and de- Boardman River Study, Sept-
ember, 1973; Office of Planning
sign product, the appro- and Research, Georgia Depart- Elevation 0-200 m. [2X]
priate requirements for ment of Natural Resources, At-
data, the need to pre-de- lanta Region Prototype, Feb-
fine how designs will be ruary 1974, 1975; Several stu-
dent projects at Harvard Grad-
evaluated, and the possi- uate School of Design, 1974,
bility of using more form- and many current projects in
alized synthesis ap- various places.
proaches.
. . . And it only gets BIBLIOGRAPHY
drawn once; thus substan-
tially altering the time im- After examining approximate-
balance between designing ly 200 different sources, the au-
and drudgery. I believe thors have selected the follow-
ing references as examples of Commercial Attractiveness Residential
that the days of drawing- the development of overlay pro-
board drudgery are num- cesses.

454
 Abercrombie, Patrick, and Regional Plan of New York and
T.H. Johnson, The Doncaster Its Environs, Vol. 1, prepared
Regional Planning Scheme, The by the staff of the Regional
University Press of Liverpool Plan, New York, 1929.
Ltd., Hodder & Stoughton Ltd., Tyrwhitt, Jacqueline, "Sur-
London, 1922. veys for Planning," Town and
Adams, Thomas, H. Lewis Country Planning Textbook,
and M. Crosley, Population, Edited by APRR, The Archi-
Land Values and Government, tectural Press, London, 1950,
Regional Survey of New Tork pp. 146-196.
and Its Environs, Vol. 11, New Wallace, McHarg, Roberts
York, 1929. and Todd, A Comprehensive
Alexander, Christopher and Highway Route Selection
Marvin L. Manheim, The Use Method, prepared for the Dela-
of Diagrams in Highway Route ware-Raritan Committee on 1-
Location: An Experiment, Pub- 95, 1966.
lication No. 161, Department of Whittle, Jack, The Prepara-
Civil Engineering, Civil Engin- tion of Planning Maps," Town
eering Systems Laboratory, and Country Planning Text-
Massachusetts Institute of Tech- book, edited by APRR, The
nology, Cambridge, Massa- Architectural Press, London,
chusetts, March, 1962. 1950, pp. 540-559.
Forshaw, J.H. and Abercrom-
bie, Patrick, County of London
Plan, MacMillan and Company
Ltd., London, 1943.
Manning, Warren, "The Bil-
lerica Town Plan,” Landscape
Architecture, Vol. HI, No. 3,
April 1913, pp. 108-118.
McHarg, Ian, Design with
Nature, The Natural History
Press, Garden City, New York,
1969.
Recreation in Wisconsin, State
of Wisconsin, Department of
Resource Development, David
Carley, Director, Philip H.
Lewis, Jr., Consultant, Madi-
son, Wisconsin, 1962.
Sender-Katalog Fur die
Cruppe Stadtebau der Stadle-
ausstellung zu Dusseldorf, 1912,
Druck von A. Bagel, Dussel-
dorf, 1912. FIGURE 2.6
The Graphic Regional Plan, Commercial/Erosion