566 PHOTOGRAMMETRIC ENGINEERING

USES AND LIMITATIONS OF AERIAL PHOTOGRAPHYIN URBAN ANALYSIS AND PLANNING*

Matthew M. Witenstein, Army Map Service, Washington, D. C.

ABSTRACT

This report presents an item-by-item tabulation, for the major components of UrbanAnalysis, of the uses and limitations of aerial photography. Its development shouldprovide a guide to city and county administrators and photogrammetric engineers alikein planning surveys based on aerial photography and should encourage the more exten­sive use of aerial survey methods in day-to-day administration.

Illustrated are various civilian experiences with major problems in city and countyadministration and planning, and with civil defense preparation in Montgomery County,Maryland, a suburban county in the Metropolitan Area of Washington, D. C.

U RBAN planning, in textbook terms,is an orderly process of organizing or

altering the physical plant of the city tomeet changing economic and social needsof the community, and guiding its futuregrowth and development.

In practice however, urban planning hasbecome a race to catch up with the postWorld War II phenomenon of rapidgrowth and spreading out of cities; anexpansion motivated primarily by thedesire for individual home owership anda place in the sun, and by the more generaluse of the automobile. In the wake of thisconstruction activity has come a tremen­dous increase in demand for urban servicesand utilities.

The expansion of facilities to meet thisdemand has generally been slower than thegrowth of new housing, commercial, andindustrial areas. In turn, each expansionof facilities has encouraged new growthand still greater demands for communityservices and utilities in a constantly ex­panding cycle.

In meeting these situations the aerialsurvey has demonstrated a remarkablecapability for assembling the area basisfor evaluation of planning problems, withsimplicity, speed and economy. Such asurvey provides a means for overcomingthe high cost and slowness of conventionalground survey and statistical sampling,for developing simple methods of integrat­ing a great volume of statistical data, andfor presenting information quickly so as tomaintain public interest in planning proj­ects.

Only slightly additional cost is entailedin projecting the use of aerial photography

from mapping to analysis and planning.In each of three main aspects of urban

study indicated in Figure 1 the applicationof aerial survey permits:

In mapping-A complete inventoryof the physical features of the urbanarea; the topography of its site and sur­rounding area, the drainage pattern,built-up and open areas, the road pat­tern and other means of communication,and the location and identification ofindividual installations.

In analysis-The classification of thefeatures of the urban area as land-usepatterns of function, structure-type, andbuilt-up area density, so as to charac­terize the population distribution, in­dustry, commerce, community servicesand facilities, and the movement of peo­ple and goods.

In planning-The relating of en­gineering standards to land-use features,in order to plan the size, capacity andlocation of facilities, and to estimate theamount of work entailed in preparationof site.In the mapping aspect, the speed, ease,

and economy of aerial survey needs noelaboration. I

In the planning aspect too, the prac­ticability of aerial survey is being recog­nized in a greater degree, particularly withregard to large-scale survey. An Englishfirm reports aerial surveys of railroadyards at scales of approximately 1 :500 forrelocation of track and other facilities. Afirm in New York reports its use of large­scale aerial survey in planning a compre­hensive county-wide sewellage system inrecently "Suburbanized" Nassau County,

* This is a part of the Panel on Photo Interpretation held on March 8, 1955 during theSociety's Annual Meeting.

PANEL--PHOTOINTERPRETATION

AERIAL SURVEY APPLICATION TO CITYAND COUNTY ADMINISTRATION

567

MAPPING

Location Size and

Distribution of Features

"Topography

'Drainage

Buift-up and open areas

Streets and transportation

Principal buildings

ANALYSIS

Pattern of

Area Characterization of Social and

Economic Differences in Settlement

Land usePopulation distributionReal estate valuation

Tax assessment

Traffic drainage areas

Trade areasUtility service areas

FIG,l

PLANNING

Layout and Work Estimate

Industrial, commercial, residential

and other subdivisions

Hi,hway, street and utility plannin,

SchO!lI and recreationallacilities

an area of rapid outgrowth from NewYork City. Another firm performs slopestudies in the layout of housing subdivi­sions, for location of streets, house" lots,water, sewerage, and gas lines, and forestimating the quantities of earth to bemoved in preparation of home sites.

On the other hand, the application ofaerial survey to the analytic aspects whichspan the broad range between mappingand planning, has lagged far behind be­cause adequate techniques have not beendeveloped for studying aerial photographyof social and economic factors which un­derly engineering considerations.

The development and testing of suchmethods, the object of this investigation,indicate that techniques for measuringsocial and economic factors by theirexpression in the use of the land affordssimplicity and economy equal to that ofmapping, for purposes of urban planningand administration.

A case in point is that of a well-knownand highly esteemed economic analyst who

"is currently having difficulty in gettingthe report of his survey accepted by alocal county government, in part becausethe survey took two years to complete,and many of the specific issues it sought to

solve have since become merged withnewer problems. For his part, this analyst,in a recent newspaper interview, has con­tended that he lost money on the survey.

In contrast, consider the case of thecity of Rockville, a suburban town in theMetropolitan Area of Washington, D. C.,which agreed to act as a test area for var­ious applications of aerial survey to urbanplanning and administration problems.These were outlined at the last AnnualMeeting"of this Society.* During the fol­lowing year many surprising events havetaken place.

The City of Rockville won a highly­coveted All-American City award for itseffective planning program and the pro­gressive attitude of its citizens. During thispast year, the issue of a small nucleus ofold inhabitants versus the large mass ofnewcomers was decisively settled in anelection; the "reform" group, largely thenewcomers, won by 83% of the total vote.The results of the election were predictableas can be readily determined in the photo­mosaic of the town, Figure 3, page 426,June 1954 issue of this JOURNAL.

* Published in the J1I ne 1954 issue of PHOTO­GRAMMETRIC ENGINEERING.

568 PHOTOGRAMMETRIC ENGINEERING

• Three choices for future development ofthe city were presented in that report:

1) becoming submerged as a dormitorysatellite of Washington, D. c.,

2) developing as a regional shoppingcenter,

3) developing as an area of light manu-facturing industry.

The choice decided upon was that of aregional shopping center with some lightindustry.

The development plan for Rockvillewhich has been unfolding since this choice,has acted as a catalyst to fuse variousfactions together and to form an attractivecommunity with an aggressive civic spirit.New petitions for residential, commercialand industrial construction permits in­dicate that the population which grew 400per cent between 1946 and 1954 can beexpected to double again by 1960, withconsequent increase in land values, busi­ness activities and job opportunity.

The heart of this development programis an annual aerial survey which will servea variety of planning needs and keep pacewith the rapid growth of the city.

In general, urban analytic studies forcivilian purposes are undertaken to con­sider such problems as the determinationof service areas in relation to residential,industrial, commercial and other require­ments for utilities; trading areas for exist­ing and potential shopping areas; trafficdrainage sheds for planning roads; popula­tion density and differentiation for plan­ning school districts and location of rec­reational facilities.

The basis of the analytic study is theassembly of statistical data which are con­sidered characteristic of the areas to whichthey apply; such characteristics may bepopulation per acre, automobiles per block,income per household, and many others.Considerable information is obtainablefrom tax and utility receipts and otherregularly organized data.

Sampling surveys are generally em­ployed for data not regularly collected.The traditional method is the randomsampling over a wide area in order to de­termine differences between areas, to marktheir limits, and to establish a body ofstatistical data for each area. Because ofthe high cost, large samplings cannot beundertaken at frequent intervals. A recentsurvey of a large regional shopping centerwas based on a random sampling of 5,000

households taken over a wide area somefour years earlier, in a locality which hassince doubled in population and is nowlargely urban in character. This type ofsurvey may be likened to the reconstruc­tion of the size, shape, and use of a build­ing from piecemeal exploration of itsmany individual parts. Even more im­portant than the question of recency is thefact that a random sampling procedure ismainly a check of existing buying habitsand is not of itself a measure of the po­tential buying power of an area, the pri­mary object of the trade area determina­tion.

On the other hand, an aerial survey inconjunction with sampling techniques,provides a means for directing a concen­trated· house-to-house canvass of selectedsample areas for data applicable to similarareas throughout the city. Statistical dataof great accuracy and ready application to.engineering planning can be obtained withfewer canvasses, less complicated statisti­cal exercises, and fewer extrapolations andassumptions than with random sampling.

Aerial photographs pro"vide first, anover-all view of the city, from which dif­ferentiation of the land-use pattern ismade. -Then in conjunction with availablepublished information, similar areas areclassified. From each classification, repre­sentative areas are selected for detailedground survey.

In the case of trade-area study, thelimits of area are established on the basisof accessibility by automotive travel to thepoint of interest. Usually 5, 10, 15 minutesof driving time represent the limits ofattraction to neighborhood, community orregional shopping centers. This is deter­mined by grou'nd check. Within these timezones, sample areas are selected from aerialphotographs for house-to-house canvass.Other areas of like type are then spotchecked for consistency with the sampleareas. Details of this application werepublished in the June 1954 issue of PHOTO­

GRAMMETRIC ENGINEERING.

A similar approach is being tested for astudy of many problems of civil defense inconjunction with Civil Defense officials inMontgomery County, a suburban areaadjacent to Washington, D. C. Currentcivil defense policy considers the massevacuation of urban populations fromprobable target areas to widely dispersedreception centers safe from blast, fire, and

PANEL-PHOTO INTERPRETATION 569

MONTGOMERY COUNTY

ROUTES AND MAIN BUILT·UP AREAS

DISTRICT OF COLUMBIA

/I

I

/

FIG. 2

radio-active fall-out. As the technologyof atomic warfare advances, the problemsof evacuation grow more and more com­plex. Neither precedent, excepting a fewlimited evacuation exercises, nor appro­priate statistical data are available uponwhich a concrete course of action can bebased.

In Montgomery County, civil defenseplanning has had to change three times inthe past six months. Until recently thecounty (see Figure 2) was considered rela­tively safe from atomic bombs, and plan­ning was concerned with reception of pos­sible evacuees from Washington. With newdata on the Hydrogen Bomb it becamenecessary to plan the evacuation of built­up "areas adjoining Washington. As a resultof more recent data on radio-active fall­out, drastic modifications in the evacua­tion plan are now required, dependent onweather and wind conditions.

In such situations the versatility of theaerial survey offers means of developingdata of reasonable accuracy and directapplication to evacuation planning, andprovides the flexibility required to meetchanges in policy.

Two basic considerations for evacuationare readily developed from aerial photog­raphy:

1. Determination of areas capable ofevacuation by private vehicles, and areasrequiring support of public transportation.

2. Analysis of route location and capac­ity for handling expected traffic loads.

The method developed for determiningthe capability of areas for evacuationwithout aid of public transport com­prises first, a land-use analysis to deter­mine areas of like type to be used in select­ing sample areas, Figure 3.

In Montgomery County single dwellingspredominate. In areas of single dwellingsthe family automobile forms the principalmeans of transportation to work. Theproblem has been to determine the num­ber of vehicles which remain in each areafor use of daytime home population. In aselected sample area comprising 700homes (Sample Area 1, Figure 4), a house­to-house canvass was made with the co­operation of the local citizens association.The number of invalids, school and pre­school children, and the number of daysthe family car remained at home was de­termined. The survey indicated that dueto car pooling arrangments, sufficientvehicles remain in the area for evacua­tion of the daytime non-school population,but that no great surplus of car capacity isavailable. With careful planning the en-

570 PHOTOGRAMMETRIC ENGINEERING

APPliCATION Of AERiAL PHOTOGRAPHY TO AREA SAMPLING'FOR SOCIAL AND ECONOMIC DATA RELATED TO URBAN PLANNING

BASIS-DF

SELECTED AREA SAM_PLING.

CHARACTERISTIC

HABITATION

PllOCEDURE

1. Q!fFERE.NTlAII'Of' SElTLEMENT

PATIE.rt;IN AIRPHOTO"

2. SELECTION OF

SAMPlE AREAS

3. D£TlttUD .

HOUSE TO HOUSE

CAHVM.SIN SAMPLE AREA

LAND·USE

~rcWandll'ld~

...--­4-1 perlCn'i\-3,.'KN

//

HABITATION: A JUXTAPOSITION OF HUMAN AND AREA RELATIONS GENE~LLAN[)'USE 'PATTERN AND AREAS SELECTED FOR·

SAMPLING

FIG. 3

tire non-school home population could beevacuated. Spot checking in like areasshowed substantially the same situation.In view of these results detailed samplingof Sample Area 2-an area averagingbetter than two cars per house-was notconsidered necessary.

This program dovetails with the presentpolicy of evacuating school childrendirectly from schools under teacher super­vision. Because many children are nor­mally transported over long distances itwould be impracticable to send themhome first. Instead, the home populationwould proceed to predesignated centerswhere they could be reunited with theirchildren. Applying this sampling on acounty-wide basis, the expected trafficload can be determined from counting thehouses in aerial photographs.

Route location and capacity for han­dling expected traffic loads are criticalfactors in evacuation planning. As canbe seen on the map (Figure 2), the numberof main roads suitable for evacuation islimited. The problem is mainly to evenout the flow of traffic and to secure maxi­mum use of side streets and secondaryroads.

Initially, evacuation planning con­sidered the movement out from the targetarea in all directions. With problems of

fall-out, evacuation planning has had toconsider movement only in those direc­tions affording maximum safety, as in­dicated by wind and weather forecasts.

To determine how such traffic wouldflow we proceed from the viewpoint thatin general each individual will seek theshortest time-route to a main road leadingaway from the center of the city.

A pattern of movement is assembled bystudying the road system in aerial photo­graphs. Between diverging main routes adividing line is drawn representing thetraffic drainage shed for each main route.

Within each shed the potential trafficload is related to the capacity of the mainroad, and the adequacy of the route isassessed. Further study of the street pat­tern indicates potential bottleneck poi;tsand supplementary routes, as well aswhere emergency traffic controls would berequired.

Using aerial photographs, varying com­binations of traffic flow can be studied inrelation to the several evacuation planswhich may be necessary on the basis ofweather conditions.

New aerial photography is being pres­ently obtained for the main built-up areasadjoining Washington. This will servemany urban planning needs including thatof Civil Defense, and provide the means

PANEL-PHOTO INTERPRETATION 571

for keeping abreast of rapid growth andeven more rapid change.

DISCUSSION OF MR. \i\TITENSTEI 'sPAPER

QUESTION: How can you predict damageresulting from an attack?

MR. WITENSTEIN: The damage dependsupon the wind as well as the area. So thereis no way to predetermine it. For instance,regarding the fire susceptibility of build­ings, there would be the question of fuelcontent of the building, the density ofconstruction, the type of occupancy of atract; all of these determine fire susceptibil­ity. There are many unknown factorswhich enter into the determination of fire­storm areas; these make it a most difficultproblem in the estimation of damage.

QUESTION: In this business of photo­graphing Rockville, were you able to someextent to predict what the vote would beon reorganizing the county?

MR. WITENSTEIN: No. But I will saythat in this case the vote coincided withthe delineation of the area that had takenplace prior to the election. Actually ourestimate of what would be the attitude ofpeople in the area was the basis upon whichone party in the campaign fought theissue; apparently it worked.

bAMPlE AREA I·MASS PRODUCED SUBDIVISION

QUESTION: Do you mean that the peoplein the newer houses were the ones thatwould vote for it?

MR. 'WITENSTEIN: Yes. The implicationis that some psychological and physicalaspects come together in the housing type,the settlement type that can be seen in thephotograph. There is an expression both ofthe sociological and economic factorswhich developed in that particular settle­ment in the first place; by reasoning backfrom the settlement type, we can arriveat many estimates and conclusions for thissort of planning.

QUESTION: Outside of the city limits ofRockville, what advantage has the aerialphotograph over the road map, in yourplanning for "civil defense evacuation?

MR. WITENSTEIN: This civil defenseevacuation to which I refer is a county­wide activity in which Rockville is justone small portion. As regards using roadmaps and the aerial photographs, theadvantage of the photo is that all of thedetails are in the photo, while the map is aselection of details usually somewhat outof date. Events are moving fast in areaslike Montgomery County. The maps arefine, because they present informationwhich is not always obtained from aerialphotos but the aerial photo is also neces-

SAMPLE AREA 2·"COUNTRY CLUB" DWELLINGS

FIG. 4

572

~::t

PHOTOGRAMMETRIC ENGINEERING

sary in conjunction with the maps. Theaerial photo is needed because it gives thelatest possible information. Naturally, theaerial photography has to be of recent date.It will then clear up details and perhapsindicate the number of vehicles which canbe expected on any particular road.

QUESTION: Do you recommend mosaicsfor this purpose?

MR. WITENSTEIN: Yes sir. The mosaic isthe means of getting the over-all view ofthe area.

QUESTIO ': Suppose that in a given areayou have a drainage pattern for traffic forcivil defense and for some reason it isblocked. Can a plan be prepared auto­matically which will be followed immedi­ately, without the destruction you speakof?

MR. WITENSTEIN: The automatic planwas discussed by me with the Ch;ef ofPolice for the State of Massachusetts, ata recent civil defense school conference.He said the only kind of planning that thepolice can expect to carry out during a civildefense emergency is to take care of fel­lows that are smaller than themselves.All that can be done in the way of civildefense planning is an attempt to estimatewhat will be the proba ble course of actionof the individual without supervision. Thatwould indicate points at which maximumbottlenecks might occur. I think you arereferring to potential bottlenecks whichcan be studied thoroughly from the air.Other factors may come into account

which can not be indicated by interpreta­tion, such as the number of vehicles break­ing down .and causing a jam at any givenpoint. I don't know what one can do aboutthis, but in planning one might try to un­derstand the general or potential flow oftraffic under any given set of circum­stances. The aerial photograph can be usedfor that purpose.

QUESTION: Has civil defense a detailedplan for evacuating Montgomery County?It so, has it taken into consideration theflow of traffic, say from Baltimore if thatwas also a target site, and merging withthe north-going traffic from MontgomeryCounty?

MR. WITENSTEIN: There are, of course,many intangibles. While it would be help­ful to have a real situation presented forstudy, no one wants such a situation tooccur if only to provide an opportunity totest the plan. It is possible we will have nowarning at all. Assuming there is some sortof warning-and that is the only basis onwhich any kind of plan can be developed­it is thought that we will have three tofour hours' warning. The idea is to evac­uate all the cities, because we don't knowwhich city will be the target. Generalevacuation is the only precaution that canbe taken. There would be traffic from Bal­timore and Washington merging in Mont­gomery County an hour after the evacua­tion starts. This would take place awayfrom the populated areas and would be astate police problem rather than an urbancivilian defense problem.