PANELUSE OF AERIAL PHOTOGRAPHS

FOR CADASTRAL SURVEYS10DERATOR

Mr. Daniel Kennedy

PARTICIPANTS

Mr. James M. Anderson

Dr. Clair V. Mann

Prof. C. G. Bryner

Mr. Noel M. Benson

Prof. Winfield M. Eldridge

Mr. Daniel McVay

DISCUSSANTS

Foreword

The A ccuracy of Planimetric Positions Determined from Large­Scale Photography

The Case for Adoption of Photogrammetric Methods in LandSurveying

Photogrammetry Used in Land Transfer of an Area that wasTransferredfrom the U. S. Government to the State of Utah

A Relationship of a State Coordinate System with Respect to FullControl Surveys with Subsequent Readout and the Use of aDigital Computer to Develop Geometric Layouts and Sub­divisions

Legal Aspects of Photogrammetry Used in Property Surveys

Use by the U. S. Forest Service of Photogrammetry in CadastralSurveying to Help Solve Existing Land Problems and to Pre­vent Further Problems from Developing

JOE STEAKLEY,

Convention Program Chairman

This Panel Meeting was held at the St.Louis Semi-Annual, Joint Meeting of theAmerican Society of Photogrammetry andthe American Congress on Surveying andMapping on 14 September 1962. Congratula­tions are due Mr. Jerome T. Berry, U. S.G. S., Central Area, Rolla, Missouri, for con­tributing his administrative and technicaltalent to all the necessary details before,during, and after this meeting. The meetingwas well received, and it provoked interest­ing new material for our profession.

INTRODUCTION OF MODERATOR BY JEROME T.BERRY

Introducing Mr. Daniel Kennedy theModerator is a pleasure.

Mr. Kennedy is the Central Region En­gineer of the Topographic Division,U. S. G. S., Rolla, Mo. He has been associated

'with that agency since 1926, except for fouryears when he served with Patton's Third

Army in World War II, a period as ChiefTopographic Engineer, Army Map Service,1946-47, and as Deputy Engineer, Task ForceSeven, the Eniwetok A-Bomb experiments, in1948. In May, 1948, he returned to theU. S. G. S. as Region Engineer, Central Area.From the Missouri School of Mines andMetallurgy Mr. Kennedy received the degreeof BSCE in 1932, Civil Engineering in 1935,and Doctor of Engineering (honoris causa)in 1949. He was given the Department of theInterior Distinguished Service Award in 1961.He is a Registered Professional Engineer inthe State of Missouri, and a member of theAmerican Society of Civil Engineers, Amer­ican Society of Photogrammetry, AmericanCongress on Surveying and Mapping, Amer­ican Military Engineers, National and Mis­souri Societies of Professional Engineers,Sigma Xi, and Chi Epsilon. He is author ofapproximately fifty Geological Survey Topo­graphic maps.

850

CADASTRAL SURVEYS 851

INTRODUCTION OF FIRST SPEAKER BY THE

MODERATOR

It is a pleasure to serve as Moderator ofthis Panel.

The first speaker is Mr. James McMurryAnderson.

Mr. Anderson is a part-time teacher, andoperator of the Wild A-7 Autograph, in theSurveying Department of Cornell University,where he is studying for his Ph.D., majoringin Photogrammetry. He received his B.S.

degree in Civil Engineering from the Univer­sity of Pittsburgh in 1949. From then untilJune, 1956, he was employed by H. A. Shope,Professional Engineer and Land Surveyor,Homestead, Pennsylvania, during whichperiod he became registered as a Surveyorand Professional Engineer in Pennsylvania.He worked with Richardson-Gordon & Asso­ciates, Consulting Engineers in Pittsburghuntil 1958, when he entered the GraduateSchool at Cornell University, receiving hisMaster's degree in 1959.

Accuracy of Planimetric Positions Determinedfrom Large-Scale Photography

JAMES M. ANDERSON,

School of Civil Engineering,Cornell Univ.,Ithaca, N. Y.

D URING recent years much attention hasbeen focussed on the use of photogram­

metric methods in connection with cadastralsurveys.

One of the chief questions concerns theobtainable accuracy.

That photogrammetric measurements,made with first-order instruments, are suf­ficiently accurate for cadastral surveys hasbeen well established. In experiments inGermany, Switzerland, Netherlands andCanada, ample evidence has been furnishedconcerning the accuracy of photogrammetricmeasurements for cadastral surveys.l.2,3 Inthe United States, such methods have beenapplied to practical cadastral problems bythe U. S. Forest Service and the U. S. Bureauof Land M anagemen t. 4 •5 However, each ap­plication of photogrammetry to property sur­veys had certain common characteristics:control points, property corners and testpoints were pre-marked with targets; thephotography scale varied from 1/6,000 to1/10,000; and the areas from 1,000 acres to2,000 acres.

Many land surveyors in the U. S. do mostof their work in developments containingfrom 30 to 200 acres. For their work a morerealistic evaluation of photogrammetry wouldbe provided by a test project through usinglow-altitude photography taken over a rela-

tively small area and containing no pre­marked con trol poin ts.

In conducting the test project, the instru­ment coordinates of 50 points were deter­mined in a single stereoscopic model covering'an area of approximately 60 acres. A first­order instrument-the v\'ild A-7 Autograph­was used to determine the instrument coor­dinates. The photography was taken from aheight of about 1,400 feet above groundelevation. The control points were not pre­marked. The objectives of the im'estigationwere as follows:

1. To make a comparison of the coor­dinated positions of points determinedby photogrammetric methods with theirlocations obtained by traditional groundsurvey methods.

2. To compare the lengths of lines deter­mined photogrammetrically with thelengths of the same lines measured onthe ground.

3. To evaluate the quality of the pointschosen for location and to select themost suitable type of point for this pur­pose.

4. To evaluate the accuracy possible indetermining the position, by photo­grammetric methods, of points in large­scale photography. The points so locatedwere not pre-marked with targets.

852 PHOTOGRAMMETRIC ENG! EERING

Emphasis should be placed on the fact thatnone of the points used is a property corner,nor was any attempt made to determine theproperty lines. The primary purpose was toevaluate the accuracy possible under theci ted condi tions.

TEST AREA

The test area is located on the CornellCampus at I thaca. I t was chosen because ofthe availability of aerial photography and theweal th of points well-defined on the groundand also identifiable in the photographs. Uni­versity facilities consisting of medium-sizedbuildings from two to six stories high makeup most of the structures in the area. Fourprimary streets pass through the area andnumerous sidewalks and utilities are visiblein the photographs. The photography wastaken late in May so the leaves had startedto be a problem. The land slopes uniformlyfrom east to west with the total relief amount­ing to abou t 100 feet. The area of the portionof the campus studied is approximately 60acres.

Photography

The photography was taken by the USC& G.S. using a Wild RC-8 Aerial Cameraequipped with an Aviogon lens (f= 152.29mm. 9" X9" exposure). The scale is approxi­mately 1 :2,800 or 1 inch to 240 feet. Stand­ard contact-printed Kelsh diapositive plates5 mm. thick were used for the photogrammet­ric operation.

PROCEDURE

The steps necessary to complete the projectmay be divided into four major groups:

1. Selection of points to be located.2. Acquisition of ground control points and

the location of test points.3. Determination of the point locations by

photogrammetric methods.4. Evaluation of the results.

Each of the above will be discussed sep­arately.

Selection of the Points to be Located

A preliminary examination was made ofthe stereo model in the A-7 to aid in selectinga more suitable group of test points. Althoughthis resulted in additional work on the A-7 itwas very beneficial. The examination of thepoints in the stereo model made possible amuch better evaluation of the type most suit­able for accurate pointing. The experience

gained from this project indicates that select­ing points from the photographic prints isextremely difficult. Examination of the dia­positive plates with the same amount of mag­nification as will be used for the instrumentalwork, provides the best selection of points.

Points were chosen that could be positivelyidentified in both the photograph and on theground. Examples of the selected types ofpoints are: the intersecting edges of side­walks, joints in the sidewalks; intersectingcurbs at the face of the curb, its back or itscenter line; corners of concrete platforms;corners of the cast iron grate inlets; and thecenter lines of intersecting painted stripesused to mark parking spaces. Although anattempt was made to use only points on theground level quite a few points were chosenon the corners of walls. An evaluation of thequality of the points selected will be madelater.

Acquisition of Ground Control Points

The primary con trol survey was estab­lished by transit and tape traverse with anaccuracy of one part in 10,000. To locatethe desired number of test points it wasnecessary to establish supplementary trav­erses. Such traverses were run with an ac­curacy of not less than one part in 8,000. Alltraverse computations were performed on theelectronic computer at the Cornell ComputingCenter.

In locating the test points, individualpoints were included either as a traversepoint in one of the supplementary traverses,or the point was located from the supple­mentary traverse by a closed traverse. Inthis way a check on every test point wasavailable on both field work and computa­tions.

Elevations of test points were determinedby closed-circuit differential leveling. Eachcontrol elevation was included as a turningpoint in the closed level circuit.

Photogrammet·ric Aleasurements

The photogrammetric phase of the problemmay be divided into three sections:

1. The relative and absolute orientation ofthe stereoscopic model.

2. Determination of the instrument coor­dinates of the points previously selected.

3. Transformation of instrument coor­dinates to ground coordinates.

All photogrammetric measuremen ts wereperformed on the Wild A-7 in the School of

ACCURACY OF PLANIMETRIC POSITIONS 853

Civil Engineering Surveying Department atCornell University. The computations neces­sary for coordinate transformation were madewith the Burroughs 220 Electronic Computerat the Cornell Computing Center.

Relative and Absolute Orientation

During the early stages of the project,correction plates to compensate for cameralens distortion were not available for the A-7at Cornell. After the first set of instrumentcoordinates had been recorded, a set ofAviogon correction plates was obtained. Sub­sequently, another absolute orientation wasmade, and a second set of instrument coor­dinates was recorded. Consequently, the re­sults evaluated in this report are based oninstrument coordinates obtained from twoseparate absolute orientations: the first, re­ferred to as Test 1, was made without correc­tion for camera lens distortion; the second,referred to as Test 2, was made using thecorrection plates to compensate for that dis­tortion. Although the lens distortion in theAviogon lens is very small (± 5 microns), usingthe correction plates was later found to beessential for the most accurate results.

In the first absolute orientation five ver­tical control points and four horizontalpoin ts were used for leveling and for scalingthe model. The same procedure was followedin the second orientation except that an ad­di tional horizon tal control poin t was used forscaling.

Instrument Coordinates

The A-7 at Cornell is equipped with anEK-3d Electrical Coordinate Printer. Thisequipment allows rapid, accurate tabulationof instrument coordinates with a minimumeffort.

In the initial test, each point was locatedby setting the floating dot on the point fiv.etimes in succession, with the mean value be­ing used for the transformation. This proce­dure seemed time-consuming without a com­mensurate increase in the accuracy. In thesecond test each point was located twice bytwo independen t observations. The meanvalue of the two observations was used forthe coordinate transformation.

where:X & Y = the ground coordinates of a

given partx & y=the instrument coordinates

of the same pointXo & Yo = translation in the X and Y

directionsA, B, C and D=transformation constants

representing a rotation anda change in scale.

Using two or more ground control points,the above equations are solved by the methodof least squares for the transformation con­stants A, B, C and D and the translationsX o and Yo. With these constants, the groundcoordinates for poin t i, instru men t coor­dinates Xi & Yi may be determined.

The transformation equations were pro­grammed for the electronic computer. Usingthis program in conjunction with the elec­trical coordinate printer, minimizes thechance of errors occurring in the computationof ground coordinates of points.

As an indication of the accuracy of theabsolute orientation, the standard error ofthe position of the ground points used for thetransformation was computed. These errorsare tabulated herein for the two absoluteorien ta tions.

Standard ErrorNumber of

Test ControlFt. On the Ground Points

1 0.30 42 0.23 5

The standard error in both cases is some­what larger than is desirable for this type ofwork. However, with only four or five controlpoin ts, a very small error in the photo orground coordinate of just one control pointcould cause the standard error to becomelarge and still not adversely affect the loca­tion of the majority of the points in the model.

EVALUATION OF RESULTS

The results of comparing coordinates de­termined by photogrammetric methods andground survey methods are tabulated inTable 1. For the evaluation the standard

TABLE I

. No. of dx,Test Points ft.

X Y

Maximum Errorsin Ft.

Coordinate Transformation

A linear transformation of instrument co­ordinates was performed by using the follow­ing equations:

X = Ax + By + Xo

Y = X o + By + Yo

12

4351

dy,ft.

0.25 0.470.32 0.23

0.700.67

0.960.43

854 PHOTOGRAMMETRIC ENGINEERING

errors were computed for the differences in Xand Y coordinates according to the followingformulae:

./---;;'i-Standard Error in the X direction = dx = 0/-­

(n-l)

vi V2

Standard Error in the Y direction = dy = --­(n-l)

n = number of test pointv = residual error between X ground and X

transformed

In Test 1, 43 points were located by bothmethods. The standard errors in X and Yrespectively, were 0.25 ft. and 0.47 ft. Maxi­mum errors were 0.70 ft. in X and 0.96 ft. inY. Correction plates to compensate for cam­era lens distortion were not used in this test.

Fifty-one points were compared in Test 2.Correction plates were employed to com­pensate for camera lens distortion in this test.The standard errors in X and Y were 0.32 ft.and 0.23 ft., respectively. The maximumerrors were 0.67 ft. in X and 0.43 ft. in Y.

Care should be exercised not to interpretthe results of Table 1 as an indication of theabsolute accuracy of the photogrammetricmethod. Errors occur in the coordinates de­termined by ground methods as well as inthe locations measured photogrammetrically.Since position determined by ground methodsis normally accepted as a standard if the workmeets the accuracy required for the job, thenthe results in Table 1 may be interpreted asan indication of correspondence.

Comparison of Length A1easured Photogram­metrically and on the Ground

Table II shows the standard errors result­ing from comparing lengths measured on theground with those calculated from coor­dinates determined by ground survey meth­ods. The standard error resulting from thesecomparisons varies from 0.13 ft. in 10 dis­tances of from 18 to 50 feet in length to 0.18ft. in distances from 150 to 200 feet in length.Since the errors in distances measured on theground tend to become larger with increasedlength of line, the distances which were com­pared were divided into groups of increasinglength. The comparison of lengths was madeonly for Test 2.

If one assumes that distances of 50 feet orless can be determined with near-absoluteaccuracy on the ground, then the above com­parison could be considered to be a morerealistic indication of the absolute accuracyof the photogrammetric method.

TABLE [[

Length Number Standard lvlaximumof of Error Error

Line Ft. Lines Feet Feet

0-50 10 0.13 0.2550-100 10 0.15 0.25

100-150 10 0.15 0.28150-500 12 0.18 0.38500-1,200 10 0.13 0.25

One would normally expect that discrep­ancies occurring in a comparison of distances,determined by the two methods, would be­come greater with increases in length of line.One reason for such an increase would be thelarger accumulation of errors in the measureddistance of the longer line on the ground.Another source of error would be due to theaccumulation of errors inherent in the photo­gram metric method. The latter group oferrors would be particularly evident in thephotogrammetric determination of very longlines occurring in a strip of aerial photography.On the other hand, in a single stereoscopicmodel, the discrepancies in the relative posi­tions of photogrammetrically determinedpoints should be fairly constant when usinglarge-scale photography in which compensa­tion for lens distortion has been made. Al­though the sample of comparisons obtained inthis study is small, the results as shown inTable II appear to verify the above con­clusion.

Evaluation of the Quality of the Points Selected

In order to arrive at some conclusion withregard to the type of points giving the bestresults, computation was made of the stand­ard errors of groups of the types of pointsoccurring most frequently. The points weredivided into three principal categories: (1)sidewalk in tersections and corners; (2) cornersof inlet grates; and (3) corners of walls. Theremainder of the points fell into groups toosmall to be useful for analysis. A summary ofthe results obtained in analyzing the abovethree groups is given in Table III.

Although the number of points in eachcategory provides a rather small sample foranalysis, certain conclusions can be drawn:(1) points having relief (the corners of walls)provided the least satisfactory resul ts; (2)well-defined points at ground elevation andhaving the darker tones of grey brought thebest results. For example, it was originallythought that the intersecting edges of side-

ACCURACY OF PLANIMETRIC POSITIONS

TABLE III

855

Standard Error (ft.) Maximum Error (ft.)Type of Point

Corners & rntersect­ing Edges of Side­walks

Corners of InletsCorners of \Valls

No. Points

191112

dx

0.190.220.31

dy

0.300.230.38

x

0.430.440.48

y

0.610.260.58

walks would furnish the best targets. How­ever, the light tone of the concrete createdglare that apparently reduced the pointingaccuracy. The corners of inlet grates havedarker tones and provided more accurateresults. On the basis of the tabulated resul tsin Table III, one can conclude that the mostdesirable points would have the darker tonesand be located at ground elevation on rela­tively level terrain.

CONCLUSlOKS

A comparison of coordinated pOSitIOns of51 points determined by photogrammetricalmethods and traditional ground surveymethods resulted in standard errors of 0.32feet in the X coordinates and 0.23 feet in theY coordinates. As indication of the accuracyof distance measurement, 50 distances rang­ing in length from 18 to 1,300 feet were cal­culated from photogrammetrically deter­mined coordinates and compared to the samedistances measured on the ground. This com­parison resulted in standard errors varyingfrom 0.13 to 0.18 feet.

To gain a picture of what order of accuracythe photogrammetric survey would yield,the size of the stereo model must be con­sidered. In the photography used for thisstudy, the dimensions of the area on theground which can be considered usable areapproximately 2,000 feet in the east-west (X)direction and about 1,300 feet in the north­south (Y) direction. Using the standard errorsgiven above as the basis for accuracy, thecoordinates of points could be determinedwith an accuracy of one part in 6,000 in theeast-west direction and one part in 5,600 inthe north-south direction.

Thus, the accuracy possible in determiningthe position of points by photograrnmetricmethods using large-scale photography, whenno points have been pre-marked, is between0.2 and 0.3 of a foot in position and between0.1 and 0.2 of a foot in distance.

In order to achieve this accuracy certainconditions must be met:

1. The area photographed must containsu fficien t well-defined points that maybe iden tified in the field and in the pho­tography.

2. Correction plates to compensate forresid ual lens distortion should be used.

3. Care should be exercised to establishthe primary survey used to locateground control points for absoluteorien tat ion, wi th an accuracy su fficien tto provide the desired accuracies in thephotogrammetric location.

4. Points chosen for location should becarefully chosen. Features having darkertones and those at ground elevation pro­vide the most accurate locations.

To increase the accuracy, targets shouldbe employed for point locations used for ab­solute orientation of the stereo model.Signalization of all utilized points wheneverpossible would certainly improve the ac­curacy. Howe\'er, the results obtained in thisproject using a single stereo model of large­scale photography, indicate that planimetricpositions of non-signalized points can ·bedetermined photogram metrically wi th ac­curacies of between one part in 5,000 andone part in 6,000.

REFERENCES

1. Trager, Herbert F., "Photogrammetry Appliedto Cadastral Surveys," Surveying and ]v[apping,Vol. 16. No.1, pp. 29-36.

2. Kasper, H. & Scholl, R., "Application of AerialPhotogral11metry to Land Registry Surveys."Special publication by Wild, Heerbrugg,Switzerland.

3. Schermerhorn, \V. & Witt, G. F., "Photogram­metry for Cadastral Survey," Photogrammetria,:\To. 2 1953-5..1,.

4. King, J. E., "Photogrammetry in CadastralSurveying," PHOTOGRAMMETRIC ENGINEERING,June, 1957, Vol. XXIII, No.3, pp. 493-505.

5. Van Zandt, F. K., "A PhotogrammetricCadastral Survey in Utah," PIfOTOGRAMMETRICENGINEERING, September, 1959, Vol. XXV, No.4, pp. 626-632.

6. Blachut, T. J., "Use of Photogrammetry in theLegal Survey Project at Alnwich," The Ca­nadian Surveyor, July 1959, Vol. 14, No.8, pp.336.

856 PHOTOGRAMMETRIC ENGINEERING

7. Dawe, H. G., "Cadastral Surveys in the MiddleEast," The Canadian Surveyor, Jan. 1957, p.634.

8. Siessor, T. J., "Use of Photogrammetry on aLegal Survey," The Canadian Surveyor, July1959, Vol. 14, No.8, p. 330.

9. Berry, Ralph M., "Technical Standards forProperty Surveys," Proceedings of Conferenceson Land Surveying, Purdue University, Janu­ary 1958, Series 93.

MODERATOR KENNEDY:

Dr. Clair V. Mann, Civil Engineer andeducator, is the next speaker. He has beenEngineer of Highways, Phelps County, Mis­souri, since 1951. He recei ved his B.S. degreefrom the University of Colorado, and aPh.D. in Engineering Education from IowaState University. After extensive experience

as a surveying and mining engineer, he joinedthe faculty of the Missouri School of Minesand Metallurgy, University of Missouri,where from 1920 to 1946 he was with Depart­ment of Engineering Drawing, serving as Headof the Department and Chairman of the Com­mittee on Engineering Education. He is theauthor of "Objective-Type Tests in En­gineering Education," published in 1930.

Dr. Mann has served as Chairman of theDrawing Division and on the committee onadministration of tests to engineering stu­dents of the American Society for Engineer­ing Education; he is also a member of theAmerican Congress on Surveying and Map­ping and the National Society of ProfessionalEngineers.

The Case for Adoption of PhotogrammetricMethods in Land Surveying

DR. CLAIR V. MANN,

Phelps County, Surveyor,Rolla, Mo.

T HE questions for discussion are (1)whether or not we endorse photogram­

metry for use in land surveying; (2) whetherconventional transit-tape methods, or photo­grammetric methods cost less and (3)whether photogrammetry will yield the requi­site measure of accuracy.

I presume my place on the panel grew outof my" recent connection as County Surveyorwi th the use of aerial survey methods on aDependen t Corner Restoration Survey of asix-mile square area-Township 36 North,Range 9 \Vest, in Phelps County, Missouri.

First of all, the topographic nature of thearea re-surveyed was typical Ozark hillcountry, consisting of rolling and often steephills lying between sea-level elevations of 750and 1,150 feet. The area is cut by two im­portant spring-fed trout streams, averaging20 to 30 feet in width, but of non-navigablenature. The area is also cut by numeroussteeply sloping ravines.

As described in papers by Mr. King* and

* PHOTOGRAMMETRIC ENGINEERING, June 1957.Also at 18th Annual Meeting of ACSM.

Mr. Fassett, t the Forest Service managesmany millions of acres of forest land withinthe United States. It became interested indeveloping a speedy, comprehensive, ac­curate method of re-establishing lost or de­stroyed Federal Land survey section corners,so that the boundaries of these huge posses­sions could be quickly, accurately andeconomically determined and could bemarked. In order to test aerial methods theForest Service set up trial projects-one inMississippi, a second in Minnesota, and athird in the Lake Tahoe area of Nevada andCalifornia. While general methods were de­veloped in these three projects, a further andmore definite and precise procedure was de­sired. After a very considerable search de­sired conditions were found in Phelps County.

The first contacts between Forest Serviceengineers and the author were on August 10,1956. By December 11 the definite supportand cooperation of Phelps county officialshad been pledged. The project planning took

t Delivered at June 14, 1958 meeting of MissouriAssoc. of Registered Land Surveyors.

LAND SURVEYING 85'7

considerable time to mature, so that actualwork was not started until November 13,1957. The decision had been made to do theprimary field work by using photogrammetricmethods-augmenting that by the necessaryground work of identifying corners and con­structing targets for aerial photographicfixation. The field work was finally finishedby May 18, 1959.

Those in charge of the project were fullyqualified for the work to be done. Mr. Kingfor many years had been engaged in executingextended cadastral surveys, and was expert inapplying photogrammetry to such work. Mr.Fassett had spent many years in precisetriangulation work for the Coast Survey andU.S.G.S. networks, and had been and is ChiefCartographer for the Forest Service withheadquarters at Milwaukee, Wisconsin. Theauthor had taught astronomy, geodesy andhigher surveying to senior engineering classesand for years (since 1903) had practisedgeneral civil and mining engineering, and hadspent years as a United States Mineral Sur­veyor in Colorado, Nevada and California.

The first step in the field work was positiveidentification of all possible corners of theoriginal Government Survey of the township(T.36N., R.9W); this survey was made inthe years 1822-23-or 135 years beforepresent work. To facilitate office and fieldoperations, each of the 133 governmentcorners was given a definite individual num­ber. Corner No.1 was the northeast corner ofthe township (and of Sec. 1). Corners 2 to 13included all full sectional and all quartercorners, numbered in succession to the westfrom corner No.1 to No. 13 which was thenorthwest corner of the township. CornersNos. 14 to 20 included the string of "quartercorners" from west to east, across the cen terlines of sections 6 to 1. Corners 21 to 33 ex­tended in successive order along the southlines of sections 1 to 6 inclusive. A somewhatsimilar pattern was adopted for identifyingeach of the 1/16 sectional corners whose loca­tion governed boundary lines of Forest Serv­ice lands. These corners were numbered from201 to 326.

I t seems appropriate to discuss the valueand use of previously made aerial photo­graphs, in the recovery of the corners that wewanted to find and authenticate. The area hadbeen "flown" in 1955 for the AgriculturalStabilization and Conservation Service U. S.Department of Agriculture. These aerialphotos were 9 X9 inches and on a scale of

1 :24,000. Before going into the field, Mr.King, by observing positions of swamped sec­tional lines, field corners, tree rows, and so on,was able to mark, on these photographs, andfor everyone of the 36 sections of the town­ship, the exact and approximate locations ofall of the full and quarter section corners, andalso of many of the 1/16 corners wanted.Quarter-inch squares were drawn around fullsections corner locations, and quarter-inchcircles for 1/16 corners.

These existing aerial photos proved to be ofthe utmost use and convenience in spottingthe general location of all corners. Whereroads led through forests to the corner vicin­ity, they could be followed either with ordi­nary passenger car or by jeep. Frequently wewalked, and followed positively identifiableravines or ridges to the corner point. Oftenthe photos showed open fields near corners.Mr.. King used these photos almost exclusivelyfor spotting every corner of the project-andin only two instances did he fail to get within50 yards of his goal.

The County Surveyor's prime part in thefield work was, first, to comb every record inhis office, or any other, that pertained to anydesired corner. This search included not onlythe County Surveyor's own records-but alsothose of the original Government Survey. Athird source of records was the file of theForest Service Rangers.

I am confident that the Moderator wiII notobject if I insert here a plug for his excellenttopographic maps. As an engineer, I was morefamiliar with using a topographic map thanaerial photos. So, in addition to the aerialpictures, I cut some of Dr. Kennedy'stopographic maps into such strips as wouldcover the area we were working in-and of asize that I could fold and paste inside one ofour small field notes.

These maps were to a scale of 1 :24,000, orroughly 3 inches per mile. The sectional lineswere printed in red-and of course the mapsshowed all roads, streams, houses and othertopographic features. To make spotting thevarious 1/16 corners possible, I drew on thestrip topographic map the necessary lines forsubdividing each section into its 1/16 (or 40acre) parts. The intersections of these lines metat the desired corners of the 40-acre tract;they were positively identifiable by close in­spection of the appropriate contour line,ridges, ravines and other features. Added tothe strip topographic maps were the adopted

858 PHOTOGRAMMETRIC ENGINEERING

numbers for all the corners. This made an ex­ceedingly convenient and readily usable fieldguide and record. To it the aerial photographs\\"ere added and carried in the field.

The field staff was equipped with small fieldcompasses,-each being mounted on a"Jacobs Staff"-and also with the customaryta pes and other accessories. The staff consistedbasically of Mr. King (project supervisor),Mr. Dodd (Forest Service recorder) and theauthor, together with George Bacon andEmmett Spencer (field aids and laborers, andtruck-jeep drivers). Field operations werebegun on November 13,1957. By use of truckor car, aerial photos, and topographic map­or perhaps by walking-we arrived at a cor­ner site. Diligent search was then made formarker or bearing trees. Only twice was itpossible to find vestiges of the originalGovernment Survey sufficient to make possi­ble corner location or restoration. For theother corners positively identified, that identi­fication was made possible only because thecounty surveyors in past years had visitedand perpetuated the corner, added new bear­ing trees, and made official record of suchdata.

vVhen properly authenticated corners werefound, existing markers were replaced withstandard reinforced concrete posts, bearingbronze caps. For full section corners, theposts were 6X6X36 inches and for 1/16 cor­ners 5 X 5 X 30 inches. Caps for the sectioncorners were bronze plates three inches indiameter and stamped by the manufacturerto read "Phelps County Survey Marker." Inappropriate spaces the plate stamping wascompleted, so as to read (for example) "Cor­ner to Secs. 1-2-11-12, T. 36 N., R. 9" Set1957 by C. V. Mann." For every corner, notless than two-usually three or four-newbearing trees were tied to the corner by bear­ing and distance. The county surveyor has aspecial tree blaze used in Phelps County­this was used on this project-two side­wise blazes on the tree about 5 feet aboveground and three deep notches below theblazes, so as to directly face the corner.

\Vhere no sufficient corner evidence war­ranted present monument setting, as well aswhere corners were actually found, targets forspotting from the aerial flight were built.Usually the target was offset from the corner,in a nearby spot open to the sky or with thebrush cleared enough to make it visible. Themost effective target center was an auto­mobile tire, painted white. Out from it were

placed four white-painted boards 1"X12"X6ft., laid at right angles, thus forming a veryconspicuous target cross. At times, the targetcenters were sheets of iron 3 feet square, witha bull's eye painted thereon. These sheetswere usually raised five feet or so from theground, and nailed to the tops of oak saplingscut off to that height. Light wire stays werepassed from sheet corners to other saplings,for anchorage. The targets were secure againstdisturbance by roaming cattle. \iVhere cornershad been found and monumented, they weretied to these targets by bearing and distance(usually very close), using the small field com­passes set to allow for local magnetic varia­tion. Where corners were not presently found,they were later located by bearing and tapeddistance from targets and known position fromtargets.

Out of the 133 originally set governmentcorners, as perpetuated by coun ty surveyors,56 were positively identified, authenticatedand approved by the county surveyor. Thirty­two of the 125 desired 1/16 corners werefinally so located, authenticated and ap­proved. The bronze plates were stamped inthe field, aside from the lettering the manu­facturer had already placed.

Having by meticulous search and consider­ation of all available evidence so set monu­ments and targets for and at every positivelyidentifiable corner, and having placed similartargets as close to the probable location of allthe other missing corners, several days werespent in running special third order traverseswithin the township, in open and advanta­geous places where targets and corners could betied to the traverse. This was done to pro­vide an "on-the-ground" measuring stickwith which to check aerial location of thesesame targets. In addition, several secondarytriangulation stations, also tied to targets,were then tied, with precise triangulation, in­to several Coast and Geodetic Survey tri­angulation stations, both inside and outsidethe township area. These stations were also"targeted," so as to be covered when theaerial flights were made. This work wasfinished on Feb. 25, 1958.

\iVi th the foregoing work completed, thetownshi p area, plus area covering the ou tsideCoast and Geodetic Survey stations, were"flown" on March 16, 1958, with a two­engine monoplane made by Beech AircraftCo. The engines were of 450 H.P. each. Thecamera carried was a Zeiss RMK 21/18 Toparlens, shutter speeds 1/100 to 1/1,000 sec.

LAND SURVEYING 859

Flight over control was at altitude of 7,800feet-and for strip runs 9,800 feet. The lenshad an 8.25 inch focal-length.

The control flights extended sufficientlyeast into Township 36 North, Range 8 West, totake in U. S. Mo. Highway 63, as an aid incontrol. Flight northward took in the CoastSurvey triangulation station at the Friscobridge across Gasconade river. The "strip"flights were along center lines of each east­west tier of sections-first west over sections1 to 6, then back over 7 to 12-and so on. Thephotographic films were processed in thePhotogrammetric Unit of the Forest Serviceusing a Zeiss C-8 Stereoplaneograph. At­tached electronic equipment made up atyped record whereby each of the approxi­mately 360 set field targets was definitely andaccurately positioned and recorded with their"x" (east-west) and "y" (north-south)coordinates, in FLEET with reference to theMissouri (Mercator) Plane Coordinate Net­which is itself tied into the precise triangula­tion net of the USC &GS. The prime meridianfor the south-central Missouri coordinate netis the meridian of 92°-30' west longi tudepassing through the center of that part of theMissouri net.

These target coordinates were then sentto the Milwaukee office of the Forest Service,where a staff headed by Ray Fassett and Vic­tor Hedman computed the X and Y coordi­nates for each of the recovered section cor­ners. The same staff, conforming to acceptedFederal and State requiremen ts governingrestoration of lost and destroyed corners, thenestablished the "X" and" Y" coordinates forall the corners now missing. I n doing this,they used accepted principles of "single" or"double" interpolation of corners, whicheverwas properly applicable.

\Vith these computations completed, thefield party was ready to monument each ofthe missing corners. From each respectivetarget the corner was located by taped dis­tance and bearing. This field work was com­pleted May 18, 1959. The County Surveyorhad been in the field, hunting and authenticat­ing corners, supervising corner setting, and inthe office digging up records-for 66 days.The other members of the field force spen t atleast twice that time in the local field.

The final work consisted in making the re­vised township plat, placing on it the bear­ings, distances and the corner numbers for theentire township. There followed typing the

detailed record for each indi"idual corner­how found and authenticated, how monu­men ted and witnessed, and what else wasdone. I ncl uded were the "X" and "Y"Missouri Coordinate Net values for each cor­ner. The plat measured 18 X 24 inches. Thedetail notes were first written up in terms offeet, tenths and hundredths. The originalplats with detail notes, have been officiallyfiled with the Phelps County Recorder and inthe office of County Surveyor. This, I believe,makes me the first official in the United Statesto file such a photogrammetric plat.

The original fields notes, expressed interms of feet, were changed to chains and linkin the final detail notes-this change wasmade at the urgent request of the U. S.Bureau of Land Management and to the keenregret of those who executed the survey.

An Appraisal of the Effectiveness of Photo­grammetric Method

The author has been through a completerestoration survey of a Federal land surveytownship. Also since 1903 he has been closelyassociated with the conventional transit-tape­stadia methods of surveying, including merid­ian establishment from solar and Polaris ob­servations. He has practiced in the steepmountain areas of Colorado, Nevada andCalifornia, both in general practice and as aUnited States Mineral Surveyor. Because ofthis experience he feels fully qualified toassert that this photogrammetric method ofrestoring sectional corners is entirely superiorto the old conventional transit-tape methods.

Photogrammetry offers a degree of unityover the entire surveyed area, hard to achieveotherwise. As computed for this project, andon the basis of comparing ten targets located(a) from ground work, by traverse; and (b)from aerial computation, the apparent errorsfor nine of the targets, ranged from 0.00ft. to 1.6 ft., the arithmetic mean being 0.6 ft.The tenth target was off by 7 feet-thiserror or difference was run down and cor­rected.

In two separate areas in the steep mounainsof Colorado the author was associated withcorner location and identification made bytransit-tape-triangulation methods. Based onthis knowledge he regards the presen t projectmethods as being very far in advance in re­spect to completeness, unity, and detail overthe older conven tional methods. I n one area,along Clear Creek Valley, at Georgetown,Colorado-a region of unusually steep moun­tain slopes-the several U. S. MineralSurveyors operating there laid out a precise

860 PHOTOGRAMMETRIC ENGINEERING

triangulation network extending up anddown the valley for six or more miles, with awidth of some four miles back on valleyslopes. The triangulation stations were usuallynot over 1,500 feet apart. Following the tri­angulation, all known section corners weretied to the net, and the coordinates of eachstation determined and also the coordinatesof the sectional corners, and many corners ofold patented mining claims. Complete fieldand office notes were then made up and filedwith each participating surveyor. That jobcould have been handled far easier throughusing photogrammetric methods.

As to relative costs, no final summation ofcosts for the Phelps County project has yetbeen made. For actual "Leg-Tiring Work"that was and is involved, the author would athousand times prefer the project as handledthan by the transit and tape method.

MODERATOR KENNEDY:

I know of nobody who is a more ardentsupporter than Dr. Mann in the field of cadas-

tral survey of the individual practitioners. Hehas described in detail a method that withinthe near future, I am sure, will be acceptedpractice throughout the State.

The next speaker according to the programis Professor Bryner. Because of illness he willnot be with us. This is much regretted.Instead, Mr. Jerome Berry will read the pre­pared paper.

Professor Bryner is a Professor in the De­partment of Civil Engineering of the Uni­versity of Utah. He is a member of the RockyMountain Region, American Society ofPhotogrammetry, and immediate Past Chair­man of the Utah Section, American Congresson Surveying and Mapping. He is on theBoard of Directors, Utah Chapter of theAmerican Public \Norks Association. He wasconference director and proceedings editor forthe Annual Surveying and Mapping Confer­ence held at the University of Utah, devotedto Land Surveying, Photogrammetry andrelated subjects.

Photogrammetry Used in Land Transferof an Area that Was Transferredfrom the United States Governmentto the State of Utah*

PROFESSOR C. C. BRYNER,

Dept. of Civil Engineering,Univ. of Utah,Salt Lake City

T HE proposal for this panel is in threeparts: recognition, cost and accuracies.

Regarding recognition, the Bureau of LandManagement in 1958 made a survey of tentownships in southeastern Utah. This projectdiffered from other similar tests in that itspurpose was to survey only the Township ex­terior and the four school sections that wereto pass into State ownership. This project wasin rough terrain and was not intended as acorner reclaiming project. Since the principalobjective was to survey specific sections theplanned work did not include effort in all partsof the Township. In this particular case, the

* The paper was read by Mr. Jerry Berry.

remote control characteristic of the photo­gram metric procedure made possible pin­pointing any specific need or a location for aparticular point or corner, without extensivetraversing into the interior of the Township.This is a critical contribution to the generalcadastral surveying problem.

This survey was made to facilitate transferof title from Federal to State government.This certainly indicates recognition of themethod as acceptable in the administrationof the Federal lands.

A problem does exist, however, in the areaof acceptance by local government agenciesand title companies. One contribution to this

PHOTOGRAMMETRY USED IN LAND TRANSFER 861

problem is frank and open analysis of theerrors that have been found in the photo­gram metric procedures. I personally objectto the implication in the reports of some teststhat the conventional steel tape procedureswere used to locate the errors in the photo­grammetric procedure, without specificallythrowing out the unknown factors and errorsthat exist in the taped method. This state­ment is necessary because of those whoseconfidence in the older method, is based onlyon continued use of that method.

Item 2: The records are available for somecost experiences. A large spread and averageof unit cost data have been obtained. Typicalof this spread is $180 per mile for one project,and $240 for another.

The importance here is not that the costsvary or that the photogrammetry is expensivebut rather that the detailed circumstancesof field conditions and engineering proce­dures have been and will remain varied untilsome specific routine becomes accepted as astandard. Until this happens it will probablybe difficult to obtain cost figures for compari­son to standard procedures for which we havea wealth of historical record.

The individuals who read these test resultswith an eye to the factual data can evaluateeach circumstance. The real problem is, thatmany a personal conclusion has been drawnfrom average figures only. This is all right initself, but all too often this person lends nosupport to photogrammetry or refuses to sup­port further use of a modern tool or methoddue to stubbornness or prejudice based oncasual observations.

I personally believe that the question is notonly: Can photogrammetry compete, butrather can we afford to properly design amethod which will accomplish any desiredresult? History has proven only one thing,and, that is, that we have a method availablethat has been satisfying our needs in thepast. There is no doubt about the validity ofthe equation accuracy equals money. Butthere are those who seem to doubt that theuse of photogrammetry will accomplish ac­curate results. Doubtfulness as to the ac­curacy of the end product often stems fromthe application of a design test program thatincludes such things as available photogra-

phy, existing data, and other trial items whichresult in the presence of many errors.

As to the question: Can a properly designedphotogrammetric engineering procedure meetour specified accuracy requirements? Theremust be an unreserved "Yes." The designbalance procedure just mentioned includesthe necessary premise that each step in theprocedure must meet every aspect of accuracyrequirements for that particular step, andthat no sacrifice is made in any step.

It is proven that first-order plotters, suchas the Zeiss C-8 Stereoplanograph, can supplythe foundation for a net of the photographicimage to an accuracy of one and one-halffeet with reference to the control networkupon which the photogrammetric solutionhas been based. This can be accomplishedwhen all of the preceding steps taken in thetotal project meet all the accuracy needs ofthat particular step.

The accuracy of each step must follow thegeneral criterion that any following step can­not observe or detect the preceding error asgreater than the smallest quantity that par­ticular step can measure.

This suggests that such elementary factorsas scale selection becomes critical. That onone to six thousand photography, the floatingmark of the C-8 covers a circle approximatelytwo-thirds of a foot in diameter. The em­ployment of any graphical solution in anystep must be such that at the scale of the par­ticular work a human mechanical error mustnot bring about an error that can be meas­ured, and thus accumulate through followingsteps, if it is to be certain that the final an­swer is to be near the accuracy of the abilityof the final machine step.

With relatively flawless control of secondorder, at all points and a proper mathe­matical solution to the bridging problems withcontrol on every fourth model, the adjust­ment problem will produce a random error ofnot more than one and a half feet.

Now if we are willing to program a sequenceof steps to conform to this general criterionand if the final accuracy specification is inaccordance with realistic needs, then thephotogrammetric engineering procedure canserve in solving any such measuring problem.

862 PHOTOGRAMMETRIC ENGINEERING

MODERATOR KENNEDY:

The preceding papers have shown theapplication of photogrammetric methods tocadastral surveying. The next paper goesone step further. It shows the subsequentextension to digital computations.

The paper will be presented by Noel M.Benson. He is a partner and vice president inthe firm of M. J. Harden Associates, inKansas City, Missouri, PhotogrammetricEngineers. He attended Kansas State Uni-

versity. For two and a half years he wasAssistant County Engineer of DickinsonCounty, Kansas. He served in the U. S. AirForce for four years. Subsequently he wasemployed by Howard, Needles, Tammen &Bergendoff; in 1955, he was first assistant incharge of electronic computations for a 650IBM computer, and later in obtaining con­tracts for the Photogrammetric Section. He isa member of the American Society of Photo­grammetry and the American Society of CivilEngineers.

A' Relationship of a State Coordinate Systemwith Respect to Full Control Surveys withSubsequent Readout and the Use of a DigitalComputer to Develop GeometricLayouts and Subdivisions

NOEL M. BENSON,

Vice Pres. M~. J. Harden Assoc., Inc.,Kansas City, Mo.

O UR company is primarily a participant inthis panel because the opinions of a

small organization are highly important indetermining the relationship between photo­grammetry and cadastral surveying. It wasthrough the development of photogrammetrythat our firm originally started growing. Theincreasing development in new instrumenta­tion and in advanced techniques in the com­puter field have created the belief that it isnecessary to undertake a closer relationshipof these two specialized fields. Accordingly wehave undertaken the development of a groupof computer programs, that could be used inconjunction with a photogrammetric basemap.

A new subdivision development of any sizeis taken as an example.

\Ne must first realize the importance of theState coordinate grid system. Through usinga coordinated group of points, or of any otherpoints, we are able to undertake almost anytype of geometrical manipulations and toobtain both a complete horizontal boundary

survey and the final computations for anytype of subdivision which could be laid out.

The first step in undertaking a new devel­opment is to obtain aerial photography at alow enough altitude to permit photogram­metric mapping by stero plotting equipmentat a scale of 1/1 equals 50' and with a contourin terval of one or two feet. Next, controlplanning for the entire flight of photographsis undertaken. 'liVe tie our horizontal controlinto a government coordinated control markof at least second-order accuracy. From thispoin t a second-order boundary traverse isrun around the property limits of the tract ofground to be subdivided, picking up enoughhorizontal points to accurately scale themodel or models in the plotting equipment.

After the horizontal boundary survey hasbeen completed and all of the property cor­ners and the horizontal points have beencalculated and adjusted for the error ofclosure within second-order accuracy, thephotogrammetric base map is plotted withstereo plotting equipment. The job is thengiven to either the consulting engineer or to

USE OF A DIGITAL COMPUTER IN SURVEYI G 863

the developer who is actually determining thesubdivided arrangement of lots. With thebase map he will try to determine the mostfeasible subdivision layout that can be pro­duced, taking into account the topographicfeatures of the land and also, the zoning regu­lations. After completion of his scale drawing,he is ready for undertaking the job of com­pu ting all of the lots.

To facilitate compu tation of his informationwe have developed an electronic computerprogram. Although the basic operations insuch a calculation are simple-for example,computation of the intersection point of twolines or a line and a curve-the key to efficien tcomputation is the organization of many in­dividual calculations into an orderly arraywhereby each individual problem is inter­related to problems previously solved as wellas for use in answering anticipated questions.

I t was our in ten tion to develop an internal­referencing program capable of solving anygeometric figure inherent in subdivision de­sign-in only one pass through the computer.Now, after two and one-half years of develop­ment, the program is operative, fully testedand is available for general application.

Specifically, our program will: 1) calculatebearings and distances of all sides of theboundary traverse, check the error or closureagainst the specified quality of the surveyedtraverse, and distribute errors to balance thetraverse; 2) compute bearings and distancesof all lines in the street centerline traverses;establish points of curvature, tangent inter­section, tangency and curve center and com­pu te curve lengths, radii and deflection angles,or degree of curvature; 3) compute bearingsand lengths of all lot lines; 4) compute lotareas in acres and square feet; 5) computecoordina tes of all bou ndary traverse poi n ts,street centerline points and corners.

All computations are in sufficient detailand accuracy to permit field stake-out di­rectly from the com pu ter ou tpu t, and topermit description for plat and abstractpreparation. The engineer need furnish only:1) a dimensioned, preliminary layout of thesubdivision; 2) field information on theboundary survey; 3) any special criteria per­tinent to design.

As a resul t of the first two projects recen tlycompleted, it is our belief that our programswill result in important economies whiledeveloping more complete accurate informa­tion.

Another area of development in our organi­zation, which has somewhat of the same typeof problems previously mentioned in respectto our subdivision work, is in the field of pipe­line engineering.

The major photogrammetric difference be­tween a pipeline project and a normal sub­division project is primarily due to the lengthof the pipeline. This in most cases will bemany miles in length. For this reason we havebeen using a computer program for bridging alarge group of stereo models between govern­ment bench marks in a triangulation system.The first job that was undertaken was in veryadverse terrain because it was very heavilywooded and with miles of swamp land.

vVe planned the horizontal and verticalground control from the horizontal and verti­cal bench marks over to the pipeline with atleast two bands of control in each flight line.By usi ng a first-order stereo plotter the proj­ect was then bridged between the terminalmodels at each end of the flight strip.

The selected horizon tal and vertical poin ts-the output from the plotter-were thentaken and by using a computer adjustmentprogram, the points were adj listed to a sui tabletolerance of error for the pipeline company.Each individual adjusted model was set intoKelsh plotting equipment; all of the apparentproperty lines, construction obstacles, andpipeline profiles were extracted. The co­ordinate location of all pipeine P.I. pointsand property corners were then extracted,and the rights-of-way descriptions for eachparcel of ground crossed by the pipeline werecomputed. All of the pipeline alignment,property descriptions, and the areas of eachparcel of ground were obtained by again usingthe series of computer programs that wereused in the subdivision computations.

A rectified photo mosaic was made of theentire pipeline project. This was divided intoplan and profile sheets, each of which con­tained all of the property data and engineer­ing data necessary for staking the pipeline inthe field. With the exception of obtaining ourbands of ground control, all of this was donewithout putting any survey parties on theproperty owners ground prior to obtainingpurchase of rights-of-way.

Also by using stereo plotting equipment wewere able to prepare in the office all of thecrossing permits for highways and railroads.In this particular project we also made nu-

864 PHOTOGRAMMETRIC ENGINEERING

merous alignment changes without putting asurveying party in the field.

The pipeline company staked our line in thefield in order to check our method of obtainingalignment. The company had very high praisefor our accuracy. With this procedure it alsowas able to win condemnation proceedingsagainst four property owners.

By using the described procedures, we are

MODERATOR KENNEDY:

The next speaker is Professor Eldridge. Heis an Associate Professor, in the surveyingareas of instruction, in the Department ofCivil Engineering, University of Illinois atUrbana. Previous employers include the AeroService Corporation and the U.S.G.S. His 24years of active surveying include propertysurveying, topographic mapping, construc­tion surveying and surveying teaching. Hereceived a B.S.C.E. degree from the Univer­sity of Tennessee, and a M.S. in Civil Engi­neering from the University of Illinois. He isa member of the American Society of CivilEngineers, American Congress on Surveying

able to eliminate a large amount of man-hourcost and expense.

We believe that there is a very close unionof photogrammetry in the field of cadastralsurveying. To date we have only a smallamount of this work but we hope to under­take a lot more, and also to develop more pro­cedures to reduce the cost and to increase ac­curacy tolerances.

and Mapping, American Society of Photogram­metry, American Society for EngineeringEducation, and Canadian Institute of Sur­veying. He is Secretary of the IllinoisRegistered Land Surveyors Association. Hewas ACSM representative to CommissionI-A at the Tenth Congress of the FederationInternationale des Geometres, in Vienna, in1962. Prof. Eldridge was co-author with Mr.Curtis M. Brown of the book "Evidence andProcedures for Boundary Location," pub­lished by John Wiley & Sons in 1962. I under­stand there is planned a later edition oranother book on that same subject. Also com­piled is the "Bibliography of Property Sur­veying Literature."

Legal Aspects of PhotogrammetryUsed in Property Surveys

PROFESSOR WINFIELD H. ELDRIDGE,

Dept. of Civil Engineering,Univ. of Illinois,

Urbana, Ill.

T HE principal responsibility of the propertysurveyor is that of locating on the ground

the physical limits of property ownershipfrom evidence, or that of describing suchproperty from a physical location. This ac­tivity depends upon metrical data of length,direction, and position. Many devices, fromrods and chains, through invar metal tapes toelectro-magnetic emissions have been used tomeasure these quantities.

Why not use photogrammetry?

As property locations vitally depend uponthe physical evidence of measuremen ts, thesurveyor is often faced wi th the problem ofqualifying his measurements so that theycan be admissable in the courts of law. There-

fore, it might be appropriate to consider threeaspects of legality to decide how photogram­metry may serve in this capacity. First thequestion of evidence, next, the quasi-legalspecifications and contractual agreemen ts,and lastly, professional registration.

MEASUREMENT EVIDENCE

In many of the court decisions, the judgehas ruled that a measurement is merely anestimate. This estimate, with sufficient proof,may be admitted as proper evidence. AnIllinois judge has stated:

"Objection to a question intended to measure awitness' capacity to estimate distance by statingthe length of the court room should have beenoverruled"-Heidler Hardwood v Wilson &Bennet, 243 III app 89)

LEGAL ASPECTS OF PHOTOGRAMMETRY 865

In a North Carolina case:

"Where in an action of trespass, a surveyor hastestified that he knows the boundaries of theland in dispute as described in the complaint,he may testify that they are within the naturalboundaries set out in a grant from the State....being of substantial fact."-(Berry v CedarWorks, 184 N C 187).

And in Alabama:

"In a boundary dispute, an expert surveyor, whohad the description of the line before him andhad surveyed and located it, was properly per­mitted to testify as to where the true line ran."-(Burleson v Gillam, 205 Ala 673).

These expert opinions must be based on thebest available evidence. In a Wisconsin case,the court has ruled:

"In surveying a tract of land according to aformer plat or survey, the surveyor's only func­tion or right is to relocate, upon the best avail­able evidence, the corners and lines at the sameplace originally located. Any departure fromsuch purpose and effort is unprofessional, and,so far as any effect is claimed for it, is unlawful."-(Pereless v Gross, 126 Wis 122).

Can the surveyor testify to measurementswhen he has not performed the act on theground?

The courts have favored actual meaure-ments, made on the ground. For example:

"Actual measurements might be more con­vincing"-(Reynolds v Stimson, 121 Kan 431).

"Lines actually run and marked on the groundare the best evidence of the true location of asurvey. "-(Matson v Poncin, 152 Iowa 569.)

"Facts held to show that surveys were made onthe ground, and that surveyor acclaimed posi­tion of unmarked lines and corners called forwith certainty from identified positions."­(Anderson v Schaefer, 205 SW 300.)

I can find no cases in the reports involvingthe admission of photogrammetric evidence,bu t there are many cases referring to photo­graphs. In a Federal Court case:

"1 t is not necessary that a witness would takethe photography himself, or see it made, inorder to testify to its correctness."-(MissouriRy of Texas v Magee, 49 SW 928.)

The Illinois Statutes, Ch. 51, Sec. 3, statethat photographs may be admitted as originalevidence if properly qualified. But there is alittle more concern when quantitative dataare taken from photographs. For example:

"In determining value of a farm taken for air­port, exclusion of photographs of property onground that they did not give accurate descrip­tion of terrain with slopes, elevation, andvalleys, was not abuse of discretion."-(Lutz vAllegheny County, 327 Pa 589.)

And:

"Where the photograph is offered, not as a meregeneral representation of the locus in quo, butto show distances, relative sizes, or locations ofobjects, it may be very deceptive and mis­leading."-(Ligon v Allen, 162 SW 536.)

I present these references to cases to illu­strate that in a lawsuit, especially the first toinvolve photogrammetric measurements, itwill take considerable proof to qualify thecomplied data, particularly if they are to over­come "actual" ground measurements. li1etri­cal photography has not been a science longenough in this country for the courts to takejudicial notice, and the surveyor who depends onthese data to prove property location, may needto educate the judge and jury on the entire proc­ess from initial control to final adjustment.

REGULATlOKS AND SPECIFICATIOXS

Can photogrammetric measuremen ts beused for property surveys that are controlledby one of the many sets of requirements andspecifications? The ew York, the Penn­sylvania, and the Florida Title Associationeach have "Minimum Standards" for titleinsurance policy surveys. One paragraph isidentical in each of these three:

"The surveyor's field work must be performedwith transit and steel tape and its accuracyproved by a closed traverse."

The Pennsylvania Title Association "Mini-mum Standards" has added:

"In rugged terrain it is recommended that verti­cal angles be measured on excessive slopes. Allmeasurements made with steel tape should becorrected to a temperature of 68 degreesFahrenheit."

Apparen tly there is no place for a cameraand plotter here.

The Technical Standards for Property Sur­veys approved by ACSM in 1946 was a littlemore foresighted with the statement:

"Measurements shall be made with instrumentscapable of attaining the required accuracy forthe particular problcm involved."

The more recent Jl1inimum Standard DetailRequirements for Land Title Surveys, ap­proved by both ACSN and ALTA earlierthis year is even more flexible with the state­ment:

"The surveyor's field work must be performedto locatc the property corners accurately."

The question might arise however: What isfield work?

Many specifications such as those of theInterstate Highway Act, are based upon the

866 PHOTOGRAMMETRIC ENGINEERING

government specifications for Triangulationand Traverse. These orders of precision arebased upon geometrical conditions that canbe satisfied with traverse or triangulationbut cannot be readily adapted to photogram­metric compilation.

In Massachusetts, the Torrens System oftitle registration is more in force than inany other state. The Land Court is ratherspecific in requiring that all surveys be madewith an engineer's transit fitted with a vernier,reading to one minute or a smaller division,and a steel tape with graduations to feet,tenths, and hundredths of a foot. Surveysmust be recent actual surveys, made upon theground with a closed traverse, and preferablyaround the periphery of the property.

The Land Court has this to say regardingphotogrammetry:

"Nearly all surveys and plans for the LandCourt are most economically made by using thestandard survey procedures. However, theCourt recognizes that in relatively few instances,other methods of surveying and preparing plansare highly satisfactory. Generally these othermethods, such as triangulation and photogram­metry, must be done in conjunction with theusual methods and must be limited to particularportions of the area covered by the surveyorplan. In such instances, consultation with theEngineer for the Court before performing thefield work is recommended, since additionalinformation about the work will be required forthe Land Court records. Such information mightconsist of a statement of the type of instrumentor other equipment used, a copy of the actualfield notes and calculations, and a statement onthe precautions taken to assure that the finalplan is a true report of the facts existing on theground."1

QUESTIONS OF REGISTRATION

N early all of our States require registration,either as land surveyor or a professional engi­neer, before a person can perform propertylocation services. In most of these registrationacts, there is no hint of photogrammetry. InIllinois, Chapter 133, land surveying is meantto include:

"Surveying and measuring the area of any por­

I From 1959 ~Manual of Instructions for the Sur­t'ey of Lands and Preparation of Plans. XXII, (1),p. 27, Massachusetts Land Court, Boston, 1959.

MODERATOR KENNEDY:

There are many points Professor Eldridgehas raised that are new to the audience. Wewill probably require more explanation duringthe Question and Answer Period.

Our last speaker, Mr. Daniel McVay has

tion of the earth's surface, the lengths and direc­tions of the boundary lines, and the contour ofthe surface for their correct determination anddescription, and for conveyancing or for record­ing or for establishing or reestablishing, locating,defining and making or monumenting landboundaries or lines and the platting of lands andsubdivisions thereof;"

And in Missouri, Ch. 344:

"'Land Surveying' means the surveying andmeasuring of the area of any portion of theearth's surface; locating and measuring thelengths and directions of the boundaries of land,and the contour of the surface thereof; and theplotting of lands, and subdivision thereof in thisstate, for compensation in any form or manner."

Maya registered land surveyor make hismeasuremen ts photogrammetrically in eitherof these states? Can a non-registered photo­grammetric engineer make measurements ofreal property boundaries without violatingthese statutes?

The picture is somewhat clearer in Cali­fornia where the Statute (Ch. 41, Sec. 8726)states:

"A person practices land surveying within themeaning of this chapter who, either in a publicor private capacity, does or offers to do anyoneor more of the following:

(including)d. Determines the configuration or contour ofthe earth's surface or the position of fixed ob­jects thereon or related thereto, by means ofmeasuring lines and angles, and applying theprinciples of trigonometry or photogrammetry."

If land surveyors are to employ photogram-metry, or if photogrammetrists are to makemeasurements for property location, there willarise several questions of registration. Mustthe land surveyor be fully qualified to makethe high-order measurements employing theul timate of photogrammetric cameras andplotters? If photogrammetry comes under thestate registration acts, who of the team shouldbe registered, the pilot? the photographer?the plotter operator? the compiler?

The state registration examination in Cali­formia has included some questions on photo­grammetry. How can such examinations beadministered to test the qualification of aphotogrammetrist for the serious and respon­sible job of property location?

been engaged in photogrammetric, geodeticsurveying, and cadastral surveying work since1935, when he started with the V. S. ForestService Division of Engineering in Missoula,Montana, constructing maps from aerialphotographs. Except for five years with theV.S.D.A. Agricultural Adjustment Admin-

LEGAL ASPECTS OF PHOTOGRAMMETRY 867

istration Photogrammetric Laboratory inSalt Lake City, Utah, and World War IIArmy service with the 29th Engineer Topo­graphic Battalion in the Pacific, all of hisemployment has been with the U. S. ForestService. Assignments have included surveyingand mapping projects in six states, and thePhilippine Islands. In 1958, he was assignedto and was in charge of setting up a PropertyLine Location program in the Forest Service

Northern Region until his transfer to \\"ash­ington, in 1961. He is a member of the Ameri­can Society of Photogrammetry, the Ameri­can Congress on Surveying and Mapping. Heis a Registered Land Surveyor in Montana,and presently is a Civil Engineer in the Alex­andria Unit of the Surveys and Map Branch,Division of Engineering, U. S. Forest Service.Washington, D. C.

The Use by the Forest Service of Photogrammetryin Cadastral Surveying to Help Solve ExistingLand Problems and to Prevent FurtherProblems from Developing

DA TIEL MCVAY,

U. S. Forest Service,A lexandria, Va.

D URIl\"G the last few years, various paperson cadastral surveying by photogram­

metric methods have been presented at meet­ings such as this. Also interesting and informa­tive articles have appeared in various publica­tions.

Common sense indicates that legal accept­ance of a cadastral survey should be based onits compliance with required standards ofaccuracy and adequacy, instead of the toolsused to do the work.

Aerial photographs (and instruments fortheir use) should be as much a working toolof the modern cadastral surveyor as thetransit, chain, subtense bar, theodolite, elec­tronic measuring device, desk calculator, orautomatic data processing.

I t is no longer a question of whether or notsatisfactory cadastral work can be done byphotogrammetric methods but, rather, aquestion of which working tool, or combina­tion of tools, is best sui ted to the particularjob to be done, the capabilities of personnelassigned to the work, and the facilities avail­able for their use.

It must be remembered that in using aerialphotos to obtain angles and distance measure­ments-sufficiently accurate for cadastralsurvey purposes-the very best photography,

aerial cameras, com para tors, grou nd con trol,and work techniques are required.

Photogrammetry can be highly eA'ective inother phases of cadastral work, such as cornersearch and recovery, reconnaissance, plan­ning, etc. Moreover, required work processesare simple; necessary equipment is inexpen­sive; and photography adequate for this pur­pose is often already available. Forest Serviceneed for, and use of, these processes may beof in terest.

The U. S. Forest Service is responsible forthe administration and managemen t of 186million acres of Federally owned land. Muchof this land is intermingled with land ownedby others in complex ownership patterns re­quiring abou t 820,000 property corners to con­trol over 208 thousand miles of ownershipboundaries between this land and the land ofnearly t of a million adjoining landowners.About 5% of this land is within the metes­and bounds survey areas, the remaining 95%being in the rectangular survey system.

The relati ve meri ts (or lack) of these twosystems of land surveying, are well-knownand need no elaboration. However, one or twocomments may be in order; the first is thatthese two systems have at least one thing incommon-no effective steps have ever been

868 PHOTOGRAMMETRIC ENGINEERING

taken to preserve the surveyed lines and cor­ner monuments. Consequently, their physicalevidence has been and is disappearing at analarming rate; secondly, in the rectangularsurvey system, the common (but short­sighted) practice is to grant title to ruralparcels of land by legal subdivision-of-selec­tion description, without requiring a surveyto subdivide the section and to mark the linesand corner points on the ground. This prac­tice has created countless miles of owner­ship boundaries that have never been sur­veyed. Furthermore, many of the corners re­quired to control these subdivision surveyshave now disappeared. Hence, many rurallandowners have clear ti tie and excellentrecords showing exactly what they own, onpaper, but nothing to show their propertylines and corner-point locations on theground.

As land and natural resource values rise,management activities become more inten­sive and boundary problems more acute. Tomeet these problems, the Forest Service hasset up a Land Line Location Program. Ac­tivities of this program fall logically into threemain parts:

(1) To recover what remains of theboundary con trolling corners and topreserve their locations with enduringmarkers.

(2) To plan and to execute the surveys re­quired to reestablish the lines and cor­ners that are lost, and to establish suchnew lines and corners as are needed.

(3) To mark these property boundaries sothat their location is readily apparenton the ground, and to set up a con­tinuing maintenance program to insurethat these lines and corners will not beobliterated.

Considering the magnitude of the job,funds available for this work have beenlimited in amount. The need to recover theremaining property corners before all evidencecompletely disappears is urgent. Therefore,the most activity is being directed towardcorner recovery and official remonumentation.Also, regardless of whether subsequent cadas­tral surveying is to be done by photogramme­try, ground methods, or some system not yetinvented, controlling corners of previous of­ficial surveys must first be recovered. The useof aerial photos in corner search enables thesearcher to find corners· that he otherwisewouldn't find, and to do it in less time. To

accomplish this the approximate corner loca­tions are spotted on the photos. Then, theyare used in selecting the best route to thecorner, and in making a search in the immedi­ate vicinity of the point, shown on the photo,as the approximate corner location.

To spot these corner locations on the photosfor this purpose, we "work backward"through the stereoplotter, and project the sec­tion corner location shown on a good basemap to its corresponding position on thephoto.

With practice this action can also be takenwithout a stereoplotter. By matching photofeatures with the corresponding features on areliable large-scale map, and using the surveynotes and plats and any evidence of land linesthat appear on the photos, the approximatecorner locations can be satisfactorily posi­tioned on the photos.

In addition to the photos, the field men arefurnished copies of the notes, and plats of allprevious official surveys in the project area, arecord of the land status and ownership,suitable maps and all available informationon the known condition of the corners to berecovered.

A 2-i'nch-to-the-mile (or larger) scale plani­metric map is used to show the land status,ownership, and boundary lines. By usingsuitable symbols, it also serves as a progressrecord to show corners found, cornerssearched for and not found, condition ofboundary lines, etc. Furnished with theseitems, plus a few simple tools and suitabletransportation, the search party operateswith a minimum of lost motion. At each re­covered corner, bearing tree, signs are placedon remaining authentic bearing trees. Aprecise photo identificatiol1 of the corner isobtained. Paint, flag cloth, etc., are used tomark the corner so.it can 'be found again.Also, the search party makes a completewritten record of conditions found at eachcorner point visited. After the corner searchwe know which corners are lost, which needremonumenting to prevent their being lost,and which, if any, are stilI in good condition.

The next step is to remonument the cornersthat need it. To be of any value, an enduringcorner monument must be set and the workmust be done by, or under, the direction ofthose who have the authority to do the workand to prepare an official record of work done.The Bureau of Land Management has such

PHOTOGRAMMETRY IN THE FOREST SERVICE 869

[.lower to survey land that has at any timebeen patented. Either the BLM or the Stateau thorized surveyor is au thorized to surveyboundary lines between these two categoriesof land.

The remonumentation party, therefore, in­cludes either a Bureau of Land Managementsurveyor, or a land surveyor, authorized byState law, depending on the status of the landinvolved. vVhen the surveyor is a Bureau ofLand Management employee, that Bureausupplies its official corner monument. Also,the remonumentation notes become a part ofits official survey records.

\iVhen the services of a registered land sur­veyor are used, the surveyor's name and regis­tration number are stamped on the cap, alongwith the corner designation letters and num­bers, at the time the corner monument is set.Due to the present lack of State law governingthe official preparation and filing of surveyrecords for work of this type done in ruralareas, the Forest Service in several Statesprovides the forms for recording the work; itretains this record until arrangements can bemade for its official recording and filing.

After the corner search and remonumenta­tion has been completed in an area, it isknown what additional surveys are requiredand what information is needed for soundwork planning. Arrangements can then bemade to have the work done under the properauthority and by the survey method bestsuited to existing conditions. If photogram­metric methods are selected, the procedureou tlined below is used:

(1) The photos, and photo identificationsof remaining corners (recovered by theproced ures described above) are usedto construct an accurate large-scaleplanimetric map showing previouslysurveyed land lines and, by protrac­tion, the new lines that are required.

(2) The locations of missing corners, andnew corners to be established, are pro­jected to the photos. The photos arethen used to locate these approximatepositions on the ground.

(3) All existing usable horizon tal and ver­tical control is recovered, and any re­quired additional control is established.These positions are converted to StatePlane Coordinates.

(4) Ground targets are placed on all pointsto be used for horizontal and verticalcontrol and on all recovered propertycorners. Also, placed on the approxi-

mate locations of property corners thathave been lost and new corners thatare to be established.

(5) After the ground targets are in place,the area is photographed with a precisemapping camera.

(6) These photos are used in a first-orderstereoplotter to bridge between hori­zontal control to obtain machine co­ordinate values for each property cor­ner target set (Item 4). Bridge adjust­ment and State Plan Coordinate valuesare then obtained by electronic dataprocessing.

(7) With the coordinate values obtainedfor the targets on the known cornerpoints and the original survey notes,the coordinate positions are computedfor the corners that are lost, then forthe positions of the new subdivision ofsection corners that are to be estab­lished.

(8) With the coordinate positions of thelost corner points, the new cornerpoin ts that are to be established, andthe ground targets set near them in thefield, the distance and direction is com­puted from the target to its respectivecorner point location.

(9) The distance and direction from thetarget to the corner point is laid off onthe ground and suitable corner monu­ments and accessories are set.

(10) Official notes and plats are prepared.Photogrammetry has attained an impor­

tant place in cadastral surveying just as it hasin mapping, road location and design, andmany other engineering functions.

The use of photogrammetry includes recon­naissance, survey planning, guidance in groundsearch for survey evidence, obtaining surveymeasurements and preparing official surveyplats and notes. One or more of these photo­grammetric processes can be effectively used onmost cadastral survey jobs.

Testsl conducted in the United States andin other countries have established the factthat measurements adequate for cadastralwork can be obtained by photogrammetricmethods.

1 King, J. E. "Photogrammetry in CadastralSurveying," PHOTOGRAMMETRIC ENGINEERING,June 1957.

King, J. E. "Advantages of Photogrammetry inCadastral Surveying," Surveying and Mapping,March, 1962.

Van Zandt, F. K., "A Photogrammetric Ca­dastral Survey in Utah," PHOTOGRAMMETRIC ENGI­NEERING, September, 1959.

870 PHOTOGRAMMETRIC ENGI EERING

The developmen t of more stable-base aerialfilms, more precise aerial cameras, electronicinstruments for rapid and accurate grounddistance measurements, and comparatorscapable of measuring photo-point coordinatesto an accuracy of a few microns are increasingthe survey potential of photogrammetry.

I should add that any failure in the use ofphotogrammetry for the processes describedby me, could I think, be attributed primarilyto the weakness of the system that is used. Iam not qualified to speak on the legal aspects;however, I believe that a survey that is cer­tified by a Registered Land Surveyor or anofficial who is operating in that capacityshould be judged to be legal. I t may not be

MODERATOR KEl\NEDY:

I thank you, Mr. McVay, for your excel­lent presentation of the uses of photogram­metry in cadastral surveys by your organiza­tion.

At the QUESTION AN 0 A SWERPERIOD which follows, there is only oneground rule: Those having a question shoulddirect it to the panel member concerned. Usethis microphone and give identification as toname and the organization represented. Thiswill help the other members of this group tounderstand both the question and the back­ground of the questioner.

MR. BROWN of San Diego: Mr. Anderson,you gave some figures on the experiment thatyou ran. As I recall them, when the cornerswere 50 feet apart, you had about a 0.13error, and when the corners were 1,800 feeta part, you had an error of abou to. 23.

The change in distance is 52 feet up to1,800 feet. That is about a thirty-six timeslarger distance and the error is changed fromapproximately 0.13 up to approximately60% more.

What is the linear relationship betweendistance between points on photography as tothe error, or is there any relationship betweenthe points, or is the error dependent uponjust the constant, no matter how far apartthe points are?

MR. ANDERSOl\: I will run through thosefigures again. There were ten distances offrom zero to fifty feet. The standard devia­tion there was thirteen-hundredths of a foot.

any good as are a lot of ground surveys.Nevertheless [ believe it is legal and it standsas such until somebody proves that it isn'tany good.

~ ow as far as getting confusion betweenthe photogrammetrist and the land surveyor,I can see no conflict or reason for any. If theland surveyor wants to use photogrammetricmethods, he can either become a photogram~

metrist himself or he can go to someone who isand get the services that he wants. A photo~

grammetrist can't certify to a survey unlesshe is registered as a land surveyor or anengineer. So there should be no conflict ofinterest; I think such conflict is needless.

For the distances from fifty to 100 feet it wasfifteen-hundredths of a foot; from 100 to 150feet, fifteen-hundredths of a foot; from 150feet to 500 feet, it was eighteen-hundredths ofa foot; and from 500 to 1,200 feet it was thir­teen-hundredths of a foot.

The standard error in the distances whichwere compared-eighteen to fifty feet-give amuch more realistic indication of the accu­racy, because you can measure distances fromzero to fifty feet very accurately on theground. The comparison would be less reliablefor the longer lines because errors in distancesmeasured on the ground will increase with theincrease in length of line. If we disregard thedeformations in the stereo model caused byfilm shrinkage-lens distortion, of course, wastaken into account here-the photogram­metric measurement should be fairly consist­ent, regardless of length of line. If I have amodel where I can measure lines that are 200feet long, the standard deviation should bethe same for the group of long lines as for agroup of short lines.

MR. BROW:-I (San Diego, California): Thepoint is that surveyors must change theirmode of thinking when they go to photogram­metry. There is no such thing as an error inproportion to distance, in proportion to twopoints. You get the same error whether it isfive foot apart or a hundred foot apart.

MR. ANDERSON: That is about right for thephotogrammetric method.

MR. MAYER: I am in private practice. Jwould like to question Dr. Mann on the

PHOTOGRAMMETRY IN THE FOREST SERVICE 871

quality of corners. r understand that somesearched for were internal corners rather thanon the periphery. \~'ere those corners origi­nally set when the Government surveyed thearea? \~That is the basis for searching for acorner at that location?

DR. MANN: Our County Engineers or CountySurveyors through the years ha ve establishedthose corners according to State law. That is,by statute.

MODERATOR KEl\l\EDY: How about the panelmembers. Possibly M r. McVay would like toask a question.

MR. MCVAY: I ask M r. Benson to tcll us whatsystem he used to get the photogrammetriccorner location back onto the ground to iden­tify the property lines on the ground itself?

MR. BENSOl\: When we undertook this bridg­ing on the pipe line project, the rights-of-waywere actually described by apparent propertylines. As J mentioned, this particular line wasin a very rough terrain, the wooded area \\'asvery heavy and there was a lot of s\\'amp areainvolved. \Vhat few property lines were avail­able were held to be accurate and were usedas the base of other property parcels. Only afew corners were available even though wehunted on the ground with our horizontalsurvey party, and with the use of a RegisteredLand Surveyor. We tried to find all theproperty corners that were available andmarked these. However, we had to go byapparent properties from there on.

MODERATOR KENl\EDY: There is time for onemore question.

MR. SIME: At the last Board Meeting of theAmerican Society of Photogrammetry we

established a ne\\' Committee on the subjectof this Panel-the Application of Photogram­metry to Cadastral Surveying. I make thisannouncement so that anyone that would liketo work on this committee may make arrange­ments for this by contacting the CommitteeChairman. I want to make it very clear and tosay Amen to \I'hat Mr. McVay said, thatthere certainly isn't any conflict of interest.\~ie believe that photogrammetry can be atool to cadastral surveying, and that the pur­pose of this Committee is to impro\'e theopportunity for service.

:vIR. JAMES P. WEBB (Army Map Sen'ice):This comment may not apply directly to theland survey measurement. However, there isa film now developed and being used that asfar as we can tell has the same stability asglass. In fact, we are considering using thisfilm on the stereoplanigraph and A-8 and tosandwich it onto an optical flat rather thanusing a glass plate.

Second, so far as distances are concerned, Ican't speak regarding very short distances.\\'e have found through many tests and ex­periments, that there are systematic errorsthat creep in when you triangulate manymodels in a series. However, there are alsoaccidental errors in a single model. The sys­tematic errors we cannot accurately balanceunless we know exactly what is causing themand usually we don't, but we find that thcadjustment of the random error in a strip offive models give results with about the sameaverage that you would get in a single model.

MODERATOR KENNEDY: \\'ith Mr. Webb'sremark we will close this panel, and againexpress appreciation to the panel membersand to the audience for their attendance andparticipation.

ATTENTION MEMBERS

The Nominating Committee will meet in October to nominate candidates for Presi­dent, First Vice-President, Second Vice-President and five members of the Board ofDirection. Members who wish to suggest Members for consideration as candidatesfor these offices are welcome to do so. The committee will give serious considerationto such suggestions provided received by the Committee Chairman before October 15.Address suggestions to Charles W. Culkin, 403 Orleans Circle, Vienna, Virginia.