ALTIMETERS AS USED BY THE 30TH ENGINEERSFOR MAPPING*

Col. Robert R. Robertson, Commanding Officer, 30th EngineerBase Topographic Battalion

ABSTRACT

The 30th Engineers successfully used altimeters transported by helicopter for vertical controlof multiplex mapping. The two-base method proved best and bases were separated as much as 30miles horizontally and 7,000 feet vertically. Altimeter runs included 69 points of known elevationwhich were checked within 8 feet. The new metal cased altimeter was more accurate than thewood cased model. The best results were obtained in flat or undulating areas when the wind speedwas under 18 m.p.h. Nighttime is best for altimetry and overcast days are satisfactory. Lines weredouble run. Field calibration of altimeters at the start of the survey, as well as comparison readingsat the bases for each run, improved results. Criteria for satisfactory altimetry conditions and foraccepting altimetry elevations are given.

L. E. DEMLER

I NEARLY 1950, the 30th Engineer mentation had been undertaken by theBase Topographic Battalion, under Corps of Engineers, the U. S. Geological

operational control of the Army Map Survey, U. S. altimeter manufacturers andService, was assigned the long range the survey agencies of the Canadianmission of mapping certain critical areas Government. The results obtained werein Alaska. This project entails establishing quite varied, but in summary led to thethe ground control required for the multi- conclusion that accuracies of somewhereplex compilation of large scale maps. between ±5 feet to 15 feet could be ob-

A preliminary reconnaissance of the tained during periods of stable weather ifassigned areas disclosed that field opera- operations were confined to elevationstions would have to be conducted over under 2,000 feet. It appeared that thesome of the most isolated and rugged so-called "two base" method yielded theterrain on the North American Continent. best results, but that the horizontal separa-It was evident that the survey could not tion of the bases should not exceed 10be carried on in accordance with the miles.practices commonly used in the compara- These apparent limitations were too re-tively developed and accessible areas of strictive to permit beneficial use of thethe United States. The establishment of altimeter in the Alaskan operation. Thevertical control particularly was prob- tests as conducted, however, did not ap-lematicai. Differential leveling was en- pear conclusive. Accordingly, the Battaliontirely impracticable and even trigonomet- developed and executed its own series ofric leveling was not readily adaptable to tests to more definitely determine therapid, accurate progress in terrain which capability of the altimeter. These testsvaried from great lake-studded tundra showed that under the adopted procedure,plains to precipitous ice-capped mountain the altimeter would consistently give ac-ranges. curacies of ± 8 feet up to elevations of at

To meet the problem of establishing least 4,000 feet and over courses in whichvertical control, consideration was given the horizontal separation of the two basesto the use of the altimeter. Available was as much as 20 miles. Even better re-publications on this instrument were re- suits were subsequently obtained in actualviewed, but the results were not hearten- operatioJl as will be shown later in thising. Apparently, very little had been done discussion. The tests also emphasized ato exploit its use to the extent that would previously anticipated problem. To mini-

. be required in Alaska. Concurrently, but mize the possibility of major atmosphericpractically independently, some experi- changes, a particular run over an altimeter

* Cleared for publication from a military security standpoint, by Corps of Engineers, U. S.Army and by Office of Public Information, Department of Defense.

839

840 PHOTOGRAMMETRIC ENGINEERING

course should be completed in a reasonableperiod of time, preferably not more than

. two to three hours. How to maintain thistiming in an area where even foot travelis slow and laborious posed a considerable

. problem. Only one solution was apparent­the helicopter. Additional tests were runand it was enthusiastically concluded thatby employing the team of altimeter, heli­copter and dauntless men, barriers of timeand terrain would become pregnable and

. many areas previously considered inac­cessible could be quickly and accuratelymapped.

On the bases of the Battalion helicopter­altimeter tests, Army Map Service revisedthe Alaskan project specifica..tions to callfor vertical picture point control accuraciesof ±4 feet below the 200-foot contour and± 8 feet above the 200-foot contour. This

is a relaxation from the normally requiredaccuracies of 1/10 the contour intervalwhich in this case would be 2.5 feet and5.0 feet respectively, 'based on the 25-footand 50-foot contour intervals to be usedin· compilation. With the anticipatedachievable accuracies accepted by AMS,the Battalion was ready to put to test inthe field the conclusions drawn and theprocedures evolved from its experiments.

The considerable success of the firstseason's work demonstrated that accu­racies obtained during experimentationcould not only be equalled in an actualoperation under field conditions, but couldbe bettered. Specified accuracies were ob­tained by altimeter on points up to 7,000feet in elevation and in some cases overcourses 30 miles long. ow, after addi­tional tests and with two seasons fieldwork completed, the Battalion has ac­cumulated a considerable amount ofvaluable experience in altimetry. Fromthis experience, an operating procedurehas been established which, by checks inthe field as well as by utilization of fieldwork in actual compilation, has proventhe great potential of the altimeter for thetopographic mapping of large areas ofdifficult terrain. The remainder of thisdiscussion is devoted to the use and pecu­liarities of the altimeter as experienced bythe 30th Engineers.

The Battalion has found that, withcertain incorporated refinements. the two­base method consistently produces thebest results under the field conditions en-

countered. It is not intended to imply thatthe two-base method is unquestionablythe best. Preliminary experiments haveindicated that the "leap-frog" method hasconsiderable potential and with refine­ments may equal the results obtained withthe two-base method. The two-basemethod has a particular advantage in thattemperature and humidity observationsare not required. It has a disadvantage inthat the elevation of the upper base mustbe known. In areas of sparse control suchas Alaska, bench marks are generally rareand in the high mountainous areas arepractically nonexistent. This necessitatesthe use of trigonometric methods to estab­lish upper base elevations when required.

The two-base method is comparativelysimple. In essence, it is a metho'd to deter­mine by altimeter (air pressure) the eleva­tions of a series· or line of points lying be­tween two base points (an upper and alower) whose elevations are known. Thismethod assumes that atmospheric pres­sure has a definite relationship to altitude..Th'is, of course, is true only under certainstandard conditions. Thus, to reducevariations due to atmospheric instabili tyto a minimum, as well as to offset frictionand imperfections in the mechanism of thealtimeter, a number of planning considera­tions, operational procedures and com­putation refinements have been incor­porated into the two-base method. Thesecan best be described by considering theplanning, execution and computation ofan altimeter run.

Altimeter runs are planned in the officeand the points to be included are pre­selected from a study of aerial photo­graphs. In planning a course or run, ex­perience has shown that better results willbe obtained if certain basic principles arefollowed. Acceptable results can be ob­tained over runs where the horizontaldistance between bases is as m.uch as 30miles. However, better results (fewerpoints rejected) are obtained if the hori­zontal separation is limited to 15 to 20miles. The vertical separation betweenbases apparently is not critical insofar asthe altimeter is concerned. In operations,just as good results have been obtainedwith a 5,000 foot or 6,000 foot difference'between base elevations as with a 1,000foot difference. ormally the upper andlower bases should" be higher and lower

ALTIMETERS AS USED BY THE 30TH ENGINEERS FOR MAPPING 841

respectively than any of the verticalpicture points in the course. If a particularsituation requires i't, elevations on pointsoutside the vertical range of the bases canbe obtained. In such cases, ho~ever, alow or high point should not be below orabove a low base or high base respectivelyby more than 10% of the difference inelevation between the two bases. Similarly,an ideal course would have all the altimeterpoints lying in a straight line between thetwo bases. In practice, terrain featuresdictate a compromise with this ideal.Experience has indicated that better re­sults will be obtained if the horizontardeviations from a straight line are limitedto 1/10 the horizontal distance betweenthe bases when in the immediate vicinitvof a base and i the horizontal distanc~between the bases when operating near themid-point between bases.

Terrain is closely associated with in­stability of atmospheric conditions andshould therefore be carefully consideredin planning altimeter courses. In moun­tainous areas, temperature differences be­tween high and low areas tend to generatewinds and local disturbances. Similarly,the temperature differences between areasin sunlight and those in shadows have thesame effect. It has been found that bestresults are obtained from the altimeter inflat or undulating areas. ext best arerolling hilly areas. Least best, but stillacceptable, results are obtained in pre­cipitous mou.ntain terrain. When workingin precipitous terrain, points should be

located on peaks or tops of ridges. Whennecessary, side hill points can be used, butaccuracies are decreased. Poorest resultsare obtained in deep valleys or canyonsand over large glacial areas. Ice fields andglacial areas present a special problem.On many clear, sunny days when otherareas are calm and ideal altimeter condi­tions prevail, erratic winds of 50 and 60miles an hour have been encounteredwhipping across the surface of ice fields.During overcast days, the weather oversuch areas appears to conform more to theaverage prevailing conditions. It appearsthat on sunny days the generally warmair, upon coming in contact with the coldair of the glaciers, produces the chaoticwind conditions observed.

While the above planning considera­tions, if followed, contribute to the greateraccuracy of an altimeter run, the majorburden for success falls upon the chief ofthe altimeter party. IIi addition to assuringthe proper handling and reading of instru­ments, he must be something of a clima­tologist since accuracies are directly re­lated to atmospheric stability. As windforce is an excellent indicator of 'atmos­spheric conditions, altimeter parties shouldutilize it as a guide as to wh ther or notaltimeter operations should be conductedon a particular day. An easy way ofjudging in the field is utilization of theBeaufort Scale as a guide to wind speed.Operations should not be conducted be­yond Force 4 on this Scale, which is givenas follows:

BEAUFORT SCALE

Force Description Specification of Beaufort Scale M.P.H.

0 Calm Smoke rises vertically 0-11 Light Air Wind direction shown by smoke drift but not by 1-3

wind vanes2 Light Breeze Wind felt on face, leaves rustle, wind vanes move 4-73 Gentle Breeze Leaves and small twigs in constant motion, wind 8-12

extends small flag4 Moderate Breeze Raises dust and loose paper, small branches sway 13-185 Fresh Breeze Small trees in leaf begin to sway 19-246 Strong Breeze Large branches in motion, whistling wires 25-317 Moderate Gale Whole trees in motion 32-388 Fresh Gale Break twigs off trees, progress impeded 39-469 Strong Gale Slight structural damage occurs 47-54

10 Whole Gale Trees uprooted, considerable structural damage 55-6311 Storm Rarely experienced, widespread damage 64-7512 Hurricane 75

'.

842 PHOTOGRAMMETRIC ENGINEERING

The altimeter itself is also a good guideto atmospheric conditions, A rising readingon a stationary altimeter indicates a fallingbarometer and thus an unstable atmos­phere, If such a reading reaches a 300 to400-foot variation from a normal reading,it is unwise to initiate altimeter operationsfor the day unless there are definite in­dications that a return to normalcy is inprocess. Party chiefs should make it ahabit, while in camp, to frequently observea stationary altimeter to better judge theadvisability of undertaking operations,

Acceptable results can general1y be ob­tained during any period of the day, butthe degree of accuracy is related to timeof day. During midday when the sun'srays are hottest, the air thins and pressurefal1s. The resulting fluctuation of thealtimeter reduces accuracy. Nighttime isthe most stable and thus suitable periodfor altimeter work, though it is not alwaysfeasible to conduct operations at this time,Next best in order of preference are lateafternoon and twilight" early to mid­morning, and midmorning to late after­noon, Overcast days are general1y suitablethroughout since the blanketing out of thesun's rays brings on conditions equivalentto twilight.

Once the atmospheric. conditions areconsidered satisfactory, the next step pre­paratory to physical1yrunning thealtimetercourse can be taken. This step is in threeparts and includes adjusting, calibratingand comparing altimeters. After arrivingin the field, but prior to use, all altimetersin each group (a set of 6 altimeters) areadjusted to a reading on a mercury column.This can be done by setting an arbitrarilyselected instrument at the nearest weatherstation and then using it as a master to setthe other instruments. Such adjustingwas, of course, done by the manufacturerof the instruments, but should be repeatedin the field to assure that the originalsetting has not been disturbed.

It should be noted that a set or groupconsists of 6 altimeters. It has been foundthat by always using altimeters in pairsand taking the average of the two readings,accuracies are improved. Thus, during arun, 6 instruments are used. Two remainstationary at the lower base, two remainstationary at the upper base and two rovebetween the bases and are read at eachpoint on which an elevation is to be deter­mined.

After adjustment, the set of instrumentsare calibrated by taking simultaneousreadings at each 100-foot change in eleva­tion from or near sea level to the maximumelevation at which operations will beconducted. This is most easily accom­plished by connecting the instruments to apressure-vacuum pump and using onealtimeter as a guide for reducing pressuresin 'stages equivalent to a 100-foot rise inelevation. When the maximum simulatedelevation is reached, the process is re­versed by taking similar readings backdown to sea level. This cycle is completedtwice and the results recorded, Great careis exercised in taking the calibrationreadings, A magnifying glass should beused, and if the light is not good, a flash­light for each instrument reader is anasset. If a vacuum pump is not availableor the type of al timeter being used cannotbe made air tight, calibration can beeffected by physical1y transporting theinstruments up and down a mountain sidethroughout the elevation range in whichthey are to be used. If the range is extreme,increments of 200 feet or more can be usedto simplify calibration, In either case, ateach stop, a mean reading is determinedby taking the sum of the read'ings of al1instruments and dividing by the numberof instruments. Each individual instru­ment reading is then compared with thismean reading. Any variation, either plusor minus, is the correction or "C" factorfor that particular instrument. The aver­age of the four "C" factors (two cycles ofup and down the elevation range) for eachinstrument at each lOO-foot change inelevation is more conveniently used ifconstructed in the form of a graph inwhich the ordinate is correction in feet andthe abscissa is altimeter reading. Any sub­sequent r~ading on a particular instrumentis adjusted (plus or minus as the case maybe) by the amount indicated on its graphfor the elevation at which the reading istaken. Recalibration is indicated if any ofthe following circumstances prevail: (1) if30 days elapse since previous calibration;(2) if instruments are shipped over a longdistance; (3) if an instrument is substitutedin a group in which it has not been previ­ously adjusted; (4) if any instrument in agroup has been jarred.

An additional refinement which hasbeen shown by Battalion tests to improveelevation accuracies by an average of 3,3'

ALTIMETERS AS USED BY THE 30TH ENGINEERS FOR MAPPING 843

per altimeter point is a system of "Com­parison Readings." Prior to and uponcompleting an altimeter run, comparisonreadings between the instr\lments aremade at the upper- and lower bases. Thisis accomplished by transporting the sixadjusted and calibrated instruments tothe lower base where all are read simul­taneously and the readings recorded. Twoaltimeters are left at the lower base wherethey are read every 5 minutes throughoutthe operation and the reading recorded.The remaining four altimeters are carriedto the upper base where they are readsimultaneously. Two instruments remainhere and are read at 5 minute intervalswhile the other two become the rovingaltimeters and start down the course byhelicopter, visiting each preselected picturepoint. All readings on the roving altimetersare made on even 5 minutes of time so asto conform to the base altimetei readings.Upon completion of the usual two runsover the course (preferably one down andone up), the roving altimeters are againcompared with the upper base altimeters.These four altimeters are then transportedto the lower base where all six are com­pared. All comparison teadings are re­corded together with the time they weretaken. These data are used by computersto obtain a correction factor which isseparate and distinct from the calibrationor "C" factor. This factor, denoted as"X," is obtained by taking the mean ofthe readings of the six instruments at thefirst lower base comparison and again forthe last lower base comparison. As incalibration, each individual instrumentreading is compared with the mean of eachcomparison and the difference constitutesa correction factor for each instrument,based on time. From these data, graphsare prepared for each instrument in whichthe ordinate on the left represents correc­tion in feet based on the first lower basecomparison, while the ordinate on the rightis the correction based on the last lowerbase comparison. The abscissa denotestime ftom left (first comparison) to right(last comparison). From this graph it iseasy to determine the correction to beapplied to an instrument reading taken atany time during the run. The same pro­cedure is followed for the four instrumentscompared at the upper base.

Between calibration figures, comparisonreadings and readings obtained during the

running of the altimeter course, the com­puter is faced with a comparatively simple,but somewhat lengthy process to arrive atelevation determinations of the selec.tedaltimeter points. The various field dataare entered on specially devised forms bythe compnter. The computer also has aform on which his computations areentered. To obtain the elevation of analtimeter point, the two runs over acourse are computed separately and amean taken of the final results. The com­puter starts with two field readings sincethe instruments are always used in pairs.To each of these readings, the appropriate"C" factor correction, as obtained fromthe individual calibration graphs, is added(plus or minus as the case may be). Thenext step is to apply the comparison cor­rection factor "X." This is done by ref­erence to the comparison factor graphs todetermine the amount of correction to beused for the time the reading was taken.Each altimeter has two graphs, one fromthe lower and one from the upper basecomparison. To the reading of altimeter"A" (with the "C" factor already added)the lower base comparison graph factor isadded. This is repeated using the upperbase comparison graph to obtain the upperbase factor. The whole process is repeatedfor altimeter "B." Thus there are fourreadings for the altimeter point. The meanof these is taken and the result becomesthe "field reading" used in the final com­putation. Similarly, the appropriate factor"X" correction is applied to both thelower and upper base readings. Since eachpoint in a run is computed separately, the"X" factor corrections, as applied to thebase readings, are for the same time as thereadings were taken on the point beingcompUted.

The final computation utilizes fiveknown or basic figures: the actual eleva­tion of the lower base, the actual elevationof the upper base, the corrected lower basealtimeter reading, the corrected upperbase altimeter reading and the correctedroving altimeter reading on the pointwhose elevation is to be determined. Thelower base reading is subtracted from theroving reading and the difference en teredon the computation form as a plus value.The roving reading is subtracted from theupper base reading and the differenceentered on the form as a minus value.Assuming, of course, that the altimeter

844 PHOTOGRAMMETRIC ENGINEERING

point is higher than the . lower base andlower than the upper base. Next, the lowerbase reading is subtracted from the upperbase reading to obtain the altimeter in­dicated difference in elevation between thetwo bases. This figure is divided into theknown elevation difference between .thetwo bases to obtain a factor based on theratio of the true elevation difference to theindicated elevation difference. This factortimes the difference between the lowerbase reading and the roving altimeterreading (the + value referred to above)plus the known elevation of the lower base-equals the computed elevation of thealtimeter point. A similar computation,using the upper base, provides a check onthe computations. This value averagedwith a similarly computed value as a re­sult of the second run over the courseequals the final accepted elevation of thepoint.

When possible, it is most desirable toinclude one or more check points (points,other than the upper and lower base, whoseelevations are known) in an altimeter run.Also, when lines are so located that it isfeasible, a new run should include one or.more points of a previously completedrun. Such check points are not used toadjust an altimeter run, but they do pro­vide an indication of the reliability of thework. No specific cd terion has yet. beenestablished as to how far a check pointmust be missed before a run is discarded.If a check point appears "wild," the runwill be under suspicion. If a second runconfirms the first, then the check pointwill be suspect if it is a trigonometric oraltimeter elevation. If it is a spirit levelelevation, consideration is given to thepossibility that it is poorly located in therun from either a terrain standpoint or

with respect to the bases. If such is thecase the run is temporarily accepted, butone or more of the altimeter points shouldbe incorporated into the next run forconfirmation.

Certain criteria have been established asto acceptance or rejection of an altimeterrun. No single run over a course is accept:able. At least two runs from which a meancan be extracted are required. These runsshould be in opposite directions. If thefirst run is down a particular course, thenthe .second run should be up the course.The runs should agree within specifiedlimits. If an accuracy of E feet is required,then the two elevations obtained from thetwo runs should agree within 1t E. Wheretwo successive elevations on a point varyby more than this amount, a third runshould be made over the course and theallowable spread would be 3 E. If two ofthe three elevations disagree by more than3 E, but the third elevation agrees witheither of the other two within It E, themean of the two which so agree is accepted.

Altimeter results are improved by exer­cising care in handling and observinginstruments. In reading, the observershould shift his liead until he can see thepupil of his eye' centered with the needlepointer in the annular mirror. The instru­ment should be level and firmly supported.Before reading, at least two minutes'should elapse to permit the indicator tosettle. A light tap on the glass panel withthe finger will overcome any lag due to thepossible presence of static electricity. Theinstrument should be protected from sunand strong winds both during observingand transporting. Watches of all observersshould be thoroughly reliable and shouldbe ·synchronized before each new run. Oneof the most important factors is tempera-

ANALYSIS OF ALTIMETER CHECKS ON 69 POINTS OF KNOWN ELEVATIONS

(1) Number of plus errors(2) Number of minus errors(3) Number of no errors(4) Average error on all points(5) Ayerage error between sea level and 2,000 ft. elevation(6) Average error between 2,000 ft. and 4,000 ft. elevation(7) Average error between 4,000 ft. and 5,000 ft. elevation(8) Average error on points 0 to 10 miles from nearest base(9) Average error on points 10 to 15 miles from nearest base

(10) Average error using old wood case type altimeters(11) Average error using new metal case type altimeters

2937

38 ft.·9 ft.

9! ft.4 ft.

8! ft.10 ft.9 ft.5 ft.

ALTIMETERS AS USED BY THE 30TH ENGINEERS FOR MAPPING 845

ment of personnel utilized. Use of thealtimeter is inclined to be routine and thusboring. However, to obtain results, per­sonnel must be extremely reliable, con­scientious, industrious and well groundedin altimetry.

During an actual field operation, 69'points of known elevations were incor­porated in various altimeter runs underwidely varying atmospheric and terrainconditions. An analysis of the results is anexcellent indication of the potential of thealtimeter when used in the manner out­lined in this discussion. (See tabulation.)

The Battalion has converted to all

metal case type altimeters, which is con­siderably improving results.

Personnel of the 30th Engineers recog­nize that comprehensive altimetry is stilla comparatively new endeavor. While ithas been successfully employed by theBattalion, it is apparent that much is yetto be learned. I t is anticipated thatpresently planned experiments will resultin further refinements. Particularly soughtare influencing factors, corrections thereforand simplification of procedures and com­putations. In short, increased efficiencyand accuracy.

Chicago Aerial Survey CompanyFounded 1924

332 South Michigan Avenue-Chicago 4, Illinois

AERIAL PHOTOGRAPHIC EQUIPMENTResearch and Development

ENGINEERINGDESIGNING

MANUFACTURING

Aerial Cameras-Continuous Printers-Stereoscopic Viewers-Photographic Ap­

paratus-Electronic Controls-Optical Navigation Instruments-Sonne Continuous

Strip Cameras.

Contractors to U. S. Air Force-U. S. Navy and Airframe Companies