Computer Display and Analysis of Urban Information Through Time

Computer Display and Analysis of
Urban Information Through
Time and Space*
I&O P. K A DA N OPP
JERROLD R. VOSS
and
WENQELL J. BOUKNIGWI’
[email protected]
The physical environment of our tit& is shaped by a somplex interplay between the
private sector--the real atatc’markrt ---and the public sector. Private investors. in a
myriad dindividual
decisions, attempt to maximize their cconom.is return or simply try
to find a good place to live, Ideally, puhlis policy attempts both ZCB
respond to rhe‘nceds
<fithis private sector and also to guide it in a mannrr whish will improve the life of all.
However, publir policymake~ have important 5ut limited tools. at their sommand: the
pktsrment and nature d transportation facilities; water and sewe~r; the sreative USCof
shiny and taxing authority; etc. Thcsc tnc& are supposed to aid and influrnsc the
private de&ions and thereby improve the workings and quality
the city.
I.‘nfortunatcl)- neither public policymakers, nor private investors, nor even students of
urban life understand very well the urban market place. As a result, man)- private real
csttntedecisions are imcarrcct and wasteful. Moreover, great public programs, sush as
wfm rcncwal or the intentatr highway program, have had very serious and unexpected
WW~VCside cffcrtn
2’0 avoid errors in future public and private de&ions, one needs both a better
-.. _____
of
theoretical understanding of urban growth and a mcanil ofpresenting
this knowledge in a
form which can be [email protected] to the &i&en or policymaker.
In recent years, considerable et&t has been directed toward developing an unticrstanding of the various public and private determinants of urban structure. hIatb
matical theories of the urban real estate market have been construeted by 1oe;rtion
theor& such as William Alonso or Lowdon Wingo. These theories have been simplified
and converted into “urban growth models” which have heen used to predict develop
ment in partieulai cities. Moreover, these mathematical models have been used extemively in the process of making decisions about the nature and plaecment of urlrPn
transportation facilities.
Yet, all this model-making haa not been fully suucc ful. First, it has been virtually impossible to focus on individual decision units such as households as this would greatl) increase the complexity and difliculty of the analysis. !3econd, most of the mod& have r’oucentrated on urban extension and expansion rather than renewal and change in ahead!:
developed areas. Third, the analyses have tended to be cross-see tional rather than
dynamic because of the difficulty of obtainling reliable time-set+ information and cotasequently they are unable to capture, reproduce, or simulate the. way in which an urtm
area actually evolves. Fourth. the models are built to serve a definite purpose in a 1”
tic&r city and in many cas(r have not been adequately tested and evahrated in the city
for which the model was prepared. And, finally, the other major reason far our lack of
progre~ has btqn our inability to analyze, examine, and arrange urban data in an interactive and dynamic fashion,
The complexity of a city is great and requires the most innovative use of our datamanipulation technology. It is no wonder that our response to this problem has ltr,en
more artistic tluan scientific. The techniques that are currently in use have done lirtk
toexpand the sj-nthais capabilitv of the researcher and the policymaker. There is n.ryY
a ubiquitous w of statistics, and especially correlation and factor analysis, and thot:$a
using these techniques is an improvement, they fail to capture either the complexity dr
dynamics of urban development. Indeed, thcac techniques are especially suited to CIV*
sectional analyses and in the absence ofgood time-series data there has heen little need to
apply spectral analysis in order to examine lags and hysteresis cfkcts. On the other ha td
WChave p
somewhat in our graphic representation of the city, but unfortunate y,
these developmoms have limited dynamic qualities which in turn limit the opportumt J
for analysis. Although there are other reasons why we have made such a modest imp -t
on coping with our urban problems, those mentioned seem to be the most importat t,
and therefore:have provided the basis for the development of this study.
The ultimata ~ 1 of our research is the construction and
nt of an urb: D
growth model which can he used to undewtand alternative urba
pattcrm in sue h
a way that they: may be cvahrated by policymakers. In order to do this it is nv
D
move forward along two paths: one which involves urban growth theory; and, the oth, r
whick attempts to make optimum use of our dtiia handling and display capabilities, I 1
the former case ithe residential development ofa city ofmanagtxrbk size (.W,Of.Nl)
popul; tion) was seItcted for an indcpth study of all real estate trvrsac tions jland and tot; 1
property value) and the construction and me histories of each structure. The time perio I
examined extends from 1854 (the date of the fimt real estate transaction) to 196!&Th:
ol&ct ir to construct a complete history ofall of the changes in these three areas so thar
the dynamic [email protected] of urban change can be anal&l.
In addition and through time,
data will be a# embkd on a wide variety of public and private variables which tnigh:
P~i~rrlSuia~2~1910),77-t0~
LooP.BbhtotTltd.
78
I
-.
i-
,;
u4
,.
k vc infktmd
the changes in these three areas. Substantively then. an effort is being
r rdc to model the physical cvotutim of a city with the emphasis on the residential en1 x.mmt.
lertcd. the prkr annge of O--1Ci.W
N’hcn the light pen contacts the
category of price irnd iantl use which
ts of light. These s p u l s a r c p !a d
of propr’y, and s,y roordinates of properfight pen, the cbmputcr looks through its
:k buffer and on the oc~pe until the sequence is completed.
‘I%e di+lays praduwcl in this manner can serve as a pow-erful tool in ubuahzing artd
~alyzing ptterns of urban gmwth. For eswnpie, Fig. 7 and 8 show the expansion in
bnc ofar
pattern set about a town c
. The basic diamond is prowed
iblc over-the-street
distance
pcapk wanted tlo live at a minimum
X2m r
nomaly such M the arrested
r of t&: town. The viewer can alau see
rttet n &growth in the lower left-hand cornet of the picture.
pc2 ~lW&. 71-m
$3
The
time
and not stron& influenced b y a n y nearby community. It. is an old town for Ittimk
situated on an interesting site, and has had a
urban life and today p Sa diver&y of incnme, ethnic, and racial
~~~t~t~~~ history is al,;o
quite diverse i$cluding railroads, street cars, a
rban trolkys, a cnrn~~ s t r r' %
system: and an iinterstate. highway. Motion throu
the u&an area has been modified 1:;
the jnxxnce of barriers such as the kw,
the
Ire&~ and the interstate highwa ,I
FinalIy, the new interstate highway (I-57) on
t may be expected to @QdlI~t:
significant chaees in the city in the mm Puture bemuse of the greatly incre
acceS *
much information
it& ahout Kankakee. W 2
ton since the founding of tb 1
data were collected from the WxorC.J
Recorder”s &ice b :
nfy of the stWx!tiin f
or$inal subdivision 4
aerial views, These ias.
city in 18% for every other block in ths city. 1
in the Kankak4ze County Title and Trust
examining sala prices and tax stamp affi
transptation
system
@cc Fig. 10) f+m the city’s engineer, &
sources are ak+ useful in helpinhs to s
whether or nd changa in land use t
CF. In Bdditiapl we have:
through time ion the amount and diitribtitiun of employment, far&I
uccupationai sbtus of the pop&&on. At thii time WC are gathering m
s informArm will permit a detailed examination of how various
uit and vacant land pricesand the rate of change o f ’ dwelling
‘%w!sentltiOPI e#m&nbke
Data
amic map of the town has heen conswuctcdfor CZR’Tdisplay
r.xxu, &&vi&n
f~ndarits, city limits, and the river. Each
to the map at the appropriate point in time and removed
urc chane
or disappears, Figure 15 preents the growth of
in rubdiviaion increments
fmmJunar if555 to January
1%9.
.liKttd
8?
p2
(19?Oj, 77403
I:
*:
‘Q
,. :
~.~
.w.
:
-,
:
:
.
:
at,
;r?r
.
,a,,
Cc~ovEQ
;iti
:
“1,
.B?Q
.
‘LI,L
rig. 1% A p&m
.
.
.
.
.
&z2$
d&c
~~~tiam
I
i
.
,;:;
: 94lp-J
.
:
.
.
200 R
:
printout tvfMr~& 2 r, s u t x t i v i r i m2 ~t s h o w i n gt y p i c a la n a l y s i n
s ~~t a t i o n a .
Uafortunatdy,
we have only a small computer (see Fig. 161, a CXK lW4 which has
%iK woxd memory that & not large enough I:Ocontain all the mapping information. ‘I’o
core spa&
WC hold most of the mapping data on magnetic tape arranged
:rrcording to the date the feature is added. Earh piece of mapping data is added to the
(WE 1~ it is ne&d
in the time-sequence presentation and arranged in core according
$7 the time it is to be removed fmm the map. Then when the feature disappears from the
+
and the data arc no longer nc&d
on tbc computer map, it is also automatically
~ yithdrawn liom COLC.
AS in th: &ttr&ii
[email protected], port&s
of the city may be examined with the aid of an
%ct devii
(ice Fig. 17). In addition, particular subdivisioti can he s&cted for special
udy. When the inset faturc is not used, the map may he *t to expand or coutract
‘.utonr~ tically to fill the entire screen (see Figs. 18 and 19).
l&my, spau of light am placed within blocks which contain
To depict the mic
WIWV~
economic f&u*.
Qnr can show 1 spot fm every block which contains, for
:qnxik
xampk: (a) rapid t&lding of new homes; (b) commercial land; (c) residential overrow&g;
fd) rapid deterioration
home v a h m ; or (c) sales in a given price range; as
~ 4 as may
other possibilities. Fram viewing all these diplays, we apcct to gain a
‘%&ng Ibr the kindsofdcvtkrpment
which cxcurrcd in the community and as we become
WJW c~ perienced, our ideas can be &c&d
against the visual image. For example, did
-and vplua go up or down in tbc neigbborbood of the new shopping cent-?
Did
af
dFvelopmmt in the eastern part of the city precede or follow the extension of sewers ’
kne ideas may require new programming: bF example, tn ohow a spot of tight fo.
cvcs-y blxk which has both deteriorating home values and new commcrciai land uses il3
the neighborhood. In this way, our computer program and our understanding of t.ht
community will grow together.
There are three problems which make de data-handling and presentatisn task fm
Kankakee very different from thr: one for Centralia. First, we have very many differeny
kinds of data for Kankakee; land use, structure inventory, price informatti,
streets.
schools, etc. Second, we bave much more data. It will obviously not fit into memory
Third, we cannot predict before WCbegin to know what kinds of relationships we shall
see and wish to q.xplorc.
To overcome the core-storage problem, we must hold data on magnetic tape, f&d it
into core as it is needed in presentation, and fixe the storage space as it is no longer
needed. As with the map9 the best w;ry to arrange for thii reading, use, and elimination is
to arrange o u r data aquentiai! y in time on the tape and in the sequence they arc to be
eliminated within core.
In order to s~ luiy the large vat&v of relationships which wili arise in our problem, it
is quite necessary that the tape records be both Rexiblc and compact. For diffrmt typea
ofstudies, tapesare required which contain various combination of the di&rent types of
data. It is undesirable to have blank spoccs 4
unntctcidar)’ data on the tapes bccaurt
they will slow down the presentation and take too long to read. Ther&res we have
Tshnt hgid
&mm&q m$ &C&I!
2 (t 9 7 o ).f ?-i a 3
90
L m [email protected] a & m d r u r c .
it--
1
..
olustrueted prdgrarns to permit a variable data format. By using relatively simple instructints, a bit analogom to Fortran, the programmer can select, recombine. read. and wnte
FfCltarecords whieh contain diffrrent data ofvariabte length and composition. Thus, after
ci WV minuta of prcwramming time and a few more minutes on the computer, the pror:fdmmer could make up a tape containing, for instance, onlv the price per square foot
f&r sales of vacant land with date and kration, together with the number of residences
1er aerc or Mock at the time ofsale, and a notatiun to indicate whether there were in&al
Iand t
on the block. Then WC could see for example whether industries on
ercrowdcd blacks ancrease or deercase the price of vacant land. This production af a
Gable data format and its associated read, write, sort, and merge routines is a complex
~ ogranttningI. problem. However, a result of the effort expended in this direction is a
-xible and quick data-handling
faeihty.
Using thcsc data-handling
methods wc have the.pntential for flexible and fast visual
rrsentation of the Kankakcc data. The expected result of this visual study is an appre4on of the variables which determine and limit urban growth. Following this stage of
w analysis WC will then attempt to dcvclop mathematical
relationships or “models”
,hich might provide a numcrieal explanation of that which we have seen. Por example.
be prcdictcd
uantitics might be prica for homes and vacant land as well as the extent
of rcridential a cvclopment.’ Thcsc variables or their rates of change in time might be
s4&= a probtcrrf usif~g
on-line compute1
tly. the d~monstratiotr simulates both a
an analyst would ititzrar3 with thr rom-
model which will spatially predict altd
ofchanges in accessibility ; and
ex&t which will permit an analyst to input inmation into the model by draying
th a light pen on the face of the CRT.
Forther we cOns&ructedthe followi
hypothetical situaaion :
‘a) a region with a relatively PCS- rfacc containing a distribrrlion of varying sized
ther by state and country raads and an interregional railroad
h the major urban ccnxr~~;
ttrat4y located city boundaries;
,:c) that the state highway department intends to construct an interstate highway
ihle locations Nor the facility; and,
rough the region, and that there arc two
:a&,
of the highway planners depends exclusively on the
jdj
the ultimate
iter
way will influence the spatial distribution of the populabow the new
3n in the region.
4
CRT U n i t
I
O-
l.J+t
Pen
system can be adapted to most computers. Since thia is the c&y portion of the disF a!
that is camput& dependent, the interfact unit has t x c n designed to operate with ur
CIX 1tXMtcomputer. Fifth, the CRT unit is a Fairchild Type 737h Large Screen ndicator containing an electrostatically dcfkcted I7-inch rectangular cathode ray tl br
and x andp deflection amplifiers with a full output band width of I megahertz. Slat X!
to this uniit is a magnetically deflected, electrostatically farused 5-inch CRT for hi :h
resolution ( 1000 lines per inch} photographic recording. Roth moving pictures and st lh
may be made with this cmer:a unit by either maoually or automatically operating t IC
shutter unit by the computer program. Sixth, the system is designed to operate at spec h
of 333,ooO points per second, W.(nlo characters per second, and IC;,QOO
iines per secon 1.
Seventh, and finally, the ayotem employs two means of communicating
with the car Iputer-by
typewriter or by liglht pen. The display typewriter can be used as a substitu .e