-EEMIhhhEEMMhE -EMEhEEEhE -EEomohEEMNOEE · 2 1. SECURITV CL.ASI2PICATI@OO*@ NO PAGE (wm Do. ... accurate method of position determination for onshore and airborne navigation

AD-Alll 511 NAVAL POSTGRADUATE SCHOOL MONTEREY CA F/6 17/7APPLICATION OF THE GLOBAL POSITIONING SYSTEM TO NEARSHORE NYORO--ETC(U)SEP 61 V E NEWELL, 0 0 WINTER

UNCLASSIFIED N

MAIEN hEEEE-EEMIhhhEEMMhE-EMEhEEEhE-EEomohEEMNOEE

11.*2 2.2

13611111 40 .2Q

1 .25 1.4'*~ III 1.6

MICROCOPY RESOLUTION TEST CHART

NAVAL POSTGRADUATE SCHOOL- Monterey, California

DTICS MAR0 l1982

THESIS

I,

APPLICATIOI OF THE GLOBAL POSITIONING SYSTEMTO ;EARSHORE HYDROGRAPHIC SURVEYS "

by

Virginia E. N.1ewellDonald D. Winter

September, 1981

Thesis Advisors: R. W. Garwood, J, Von Schwind

Approved for public release, distribution unlimited

2 1

SECURITV CL.ASI2PICATI@OO*@ NO PAGE (wm Do. anfe_ _ _

RZAD INSTRUCT~!fNSREPORT DOCUMENTATION PAGE '" RP0 COMPL'r.NF

,. REPORT NUM691 ".'GOVT ACCESSU ONO, S. RECIPtITS CATAL.OG NUlRSER

4. TT L It (ad SU641ie) V* Type Or REPORT 6 or.0IOO COVEREO

Master's ThesisApplication of the Global Positioning System September, 1981to learshore Hydrographic Surveys . PERFORMNG OXG. REPORT NURA99ER

7. AUTmOtRe) . OW AC ONAuT NMU Rtej

Virginia Elisabeth NewellDonald Dane Winter

S. PIRPORING Oft"MIZAION MAMIE AND A0001666 A ANWR I)_fTwRj[Py TUNSE

::aval Postgraduate School

Monterey, California 939401 ONO6MOjC"U AAOOE2S 'a. REPORT CATE,, CONR~OLNG oC[ NAME .AR -'September, 1981

f!aval Postgraduate School, Monterey, CA 93940 24 NU5ERo9PAEsW I4. NONITORING A4ENCY NAME 6 AOICRSS(It dIt"IfrNI (FOM CR.1 mA4e1 Ofice) t. SWURITY CLASS. 'eei hs roten)

tie. O9CL. ASiIr1CATIONi OOWNGRAINGSCHIOULE

14. DISTNV19UTION STATEMENT NO We. Rep u)

* Approved for public release, distributionunlimited

I?. OISTRIeUTION STAT[MENT (tel t o bagt Erdli1f4 0810 20. IS teene Cm P" s qj

IS. SUPPI-EMENTARY NOTES

to. KEY WO6OS rCapOlu on o"@ eaeo of £Goe9 am aemefr or 61000 N10mb)

Global Positioning System, hydrography, translocation, ARTEMIS, theodolite

20. A66"ACT (COiRHO GR alga if ft....... Ifitly W neaj

Z Translocation of the Global Positioning System has proved to be a highlyaccurate method of position determination for onshore and airborne navigationbut it had not been previously evaluated for nearshore hydrographic surveys.The technique of translocation for hydrographic operations involves thesimultaneous reception of signals from the GPS satellites by two independentreceivers; one receiver onboard the survey vessel and one located at the

Do 1473 DImoN ot I Nov Go is O EOI -

S/N ft a0 to 0a 14. 6001 ICUDTITY CLAMPICAIOMN OPTO IMS 0AGE ftm tWO 0teep)

known stationary position. A position correction is obtained at the stationary receiver, which is then applied to the shipboard receiver online orduring postprocessing. An accuracy determination of translocated GPS wasconducted at the Naval Postgraduate School, Monterey, CA, May, 1981. Twomethods of positioning were used for comparison with GPS: '1) a least square!solution of three lines of position observed from three Wild T-2 theodolitesjand 2) a position determined from ARTEMIS, a range-azimuth short rangemicrowave positioning system. Translocated GPS accuracies of 10 meters weredetermined. It is anticipated that greater accuracies will be obtained byusing a more sophisticated receiver and more advanced processing methods.

AcceSsion For

D?'EC T.AB

1<- _ - ,?~:° ~ e _ - -

t

jil 2orl'r 14G3S,.'2.4- , g102S014-660 9 1.

Approved for public release, distribution unlimited

Application of the Global Positioning System toNearshore Hydrographic Surveys

by

Virginia Elisabeth lewell Donald Dane Winter

Lieutenant, NOAA Lieutenant Commiander, 1O0AAIB.A., University of Maine, 1971 B.S., University of Michigan,M.S., Virginia Polytechnic Institute 1973

and State University, 1975

Submitted in partial fulfillment of therequirements for the degree of

MASTER OF SCIENCE IN OCEANOGRAPHY (HYDROGRAPHY)

from the

NAVAL POSTGRADUATE SCHOOLSeptember, 1981

Authors:

Approved by:

Co-Advisor

Chai n, t c f O anoqrar'v

Dean of Science and Engineering

3

ABSTRACT

Translocation of the Global Positioning System has proved to be a highly

accurate method of position determination for onshore and airborne navi-

gation, but it had not been previously evaluated for nearshore hydro-

graohic surveys. The technique of translocation for hydrographic opera-

tions involves the simultaneous reception of signals from the GPS

satellites by two independent receivers; one receiver onboard the survey

vessel and one located at a known stationary position. A position

correction is obtained at the stationary receiver, which is then applied

to the shipboard receiver online or during postprocessing. An accuracy

determination of translocated GPS was conducted at the Naval Postgraduate

School, Monterey, CA, May, 1981. Two methods of positioning were used

for comparison with GPS: 1) a least squares solution of three lines of

position observed from three Wild T-2 theodolites, and 2) a position

determined from ARTEMIS, a range-azimuth short range microwave position-

ing system. Translocated GPS accuracies of 10 meters were determined.

It is anticipated that greater accuracies will be obtained by using a

more sophisticated receiver and more advanced processing methods.

4

I. INTRODUCTION ]

II. EQUIPMENT DESCRIPTION AND OPERATIONAL CAPABILITIES..22

1. ARTEMIS POSITIONING SYSTEM .................... 23

B. THEODOLITE NETWORK ............................. 26

C. TRANSIT POSITIONING SYSTEM ..................... 30

III. GPS SYSTEM DESCRIPTION AND OPERATIONALCAPABILITIES ........................ ..... 38

A. SYSTEM DESCRIPTION . .................. . 38

B. MANPACK RECEIVER ............................. 58

C. DIFFERENTIAL GPS * ........... . a .... *...... 7

IV. TEST PROCEDURE ................. ...... . . ... 070

A. POSITIONING NETUORK ...... 70

B. POSITIONING EQUIPMENT .......................... 72

751. Theodolites ...... 0-0. 7

2. GPS and ARTEMIS ............................ 77

C. LOG OF EVENTS ....................... q....0... 083

1 . S t a b i l i t y T s . . . . . . . . . . . . .. . . 8 4

2. Static Test ........................... 84

3. Shipboard Test .......................... 84

V. DATA PROCESSING METHODS ..........................88

A. BASIC SURVEY COMPUTATIONS ...................... 88

Li

B. THEODOLITE AND ARTEHIS SHIP POSITIONDETERMINATION ...........................

C. DIFFERENTIAL GPS SHIP POSITION DETERMINATION ...95

Do STATISTICAL DATA ANALYSIS ...................... 99

VI. ERROR ANALYSIS .......................... 103

A. ARTEMIS POSITIONIN3 SYSTEM ................... *103

1. Microvave System Errors ................... 103

2. ARTEMIS SysteU Errors ..................... 10

B. CONTROL NETWORK ............................... *112 jC. THEODOLITE NETWORK .................. 114

D. TRANSIT POSITIONING SYSTEM .................. 120

E. GPS ERROR SOURCES ........... ... .... *. ... .... 121

7 1 I I . R E S U L T S . . . . . . . . . . . . . . . . . . . . . . . . .] 3

A. STATIONARY GPS EVALUATION .. o................ 135

B. EVALUATION OF NONDIFFERENTIAL GPS ............. 158

C. DIFFERENTIAL GPS EVALUATION WITH STATICCORRECTIONS ...... ........... ..... 174

D. EVALUATION OF DIFFERENTIAL GPS WITHOUT STATICCORRECTIONS ......................... . 181

VIII. CONCLUSIONS ...............................

APPENDIX A ... .... ..... ........ ....... ...... .224

LIST OF REFERENCES .- ........ ... .o.o.-..f.. 228

BIBLIOGRAPHlY ...... . . ... . .. .. .. . . ... . .. .232 I

INITIAL DISTRIBUTION LIST .... .......................... 236

6

I. Ionospheric Group Delay as a Functionof Frequency .............................. 49

II. SANPACK Interface Wiring Code ...................... 66

III. Range Error Budget ............................. 68

IV. Satellite Observation Times ....................... 74_U,

V. Statistical Definitions ............................. 101

VI. Theodolite Instrument Errors ...................... 116

VII. TRANSIT System Errors ............................. 12

VIII. Ephemeris Update Times ........................... 127

7

2.1. TRANSIT Satellite ............................... 32

2.2. TRANSIT Satellite Orbital Network ............... 33

2.3. AN/PER 14 Geoceiver Antenna ..................... 36

- 3.1. The Nonuniform 18 Satellite Constellation ....... 45

3.2. Malker 18/3/0 Constellation Outages ............. 46

3.3. NAVSTAR Satellite ......................... ... 48

3.4. Pseudo-Range Description ........................ 57

3.5. Global Positioning System MVUE Receiver ......... 59

3.6. MVUE Receiver Face ................ .ee... 60

3.7. Cutaway of ,VUE Receiver .............. 61

3.8. NVUE CDU Keyboard/Display Layout ................ 63

3.9. Circuit Diagram of Receiver Interface ........... 65

4.1. Test Area with Traverse Sketch .................. 73

4.2. Antenna Tower on Coast Guard Pier ............... 78

4.3. R/V ACANIA Equipment Layout ..................... 81

4.4. Equipment Installation on the R/V ACANIA ........ 82

4.5. survey Area and Location of Sets by Day ......... 86

5.1. Error Ellipse ...... ....... 93

5.2. Differential GPS Processing ................... 98

6.1. ARTEMIS Range Holes ............................ 106

J8

6.2. ABTEMIS Systeu Error .......................... 109

6.3. Representative Theodolite Intersections for

Error Elipse Computations .............. oo 119

7.1. Stability Test Results 9,10 May ............... 136

7.2. Stability Test Results 9, 10 May ...-............ 137

7.3. Stability Test Results 9, 10 day ................ 138

7o4 Total User Range Error 9 May ...................139

7.5. Total User Range Error 10 May ................. 140

7.6. Static Test Results 11 May *.................... 142

7.7. Static Test Results 11 May ..................... 143

7.8. Static Test Results 11 May ..................... 44

7.9. Satellite Elevation &ngles..................... 1 4 5

7.10. Satellite Azimuth Angles .......... o........... 146

7.11. Total User Range Error 11 May .................. 47

7.12. Geoceiver vs. Gps Position Computation ......... 148

7.13. Distance Between Geoceiver and Stationary GPSPosition 12 ay ....enGe ...r .......d St.o... 150

7.14. Distance Between Geoceiver and Stationary GPSPosition 13 may ................................ 152

7.15. Distance Between Geoceiver and Stationary GPSP o s i t io n 15 M a y . . . . . . . . . . . . . .. . . . . . . 152

7.16. Total User Range Error 13 May o................. 153


7.18. Azimuth Between Geoceiver and Stationary GPSPosition 12 May ........................ 155

I

7.19. Azimuth Between Geoceiver and Stationary GPSPosition 13 May ............................ 156

7.20. Aziauth Between Geoceiver and Stationary GPSPosition 15 May .............. .................. 157

7.21. Distance Between T- $hip Position andUncorrected GPS Position (4 Satellites) Set 1 .. 159

7.22. Distance Between T-? Ship Position andUncorrected GPS Position (4 Satellites) Set 2 .. 160

7.23. Distance Between T-2 Ship Position andUncorrected GPS Position (4 Satellites) Set 3 .. 161

7.24. Distance Between T-2 Ship Position andUncorrected GPS Posit lon (4 Satellites) Set 4 .. 162

7.25. Distance Between T-2 Ship Position andUncorrected GPS Position (4 Satellites) set 5 .. 163

7.26. Distance Between T-?.Ship Position andUncorrected GPS Posit on (4 Satellites) Set 6 .. 16d



7.29. Distance Between T-? 5hip Position andUncorrected GPS Position (4 Satellites)Sets 9 and 10 .................................. 167

7.30. Distance Between T-2 Ship Position andUncorrected GPS Position (4 Satellites) Set 11.. 168

7.31. Distance Between T-; $hip Position andUncorrected GPS Position (4 Satellites)Sets 12 and 13 ................................. 169


7.33. Distance Between T-2 Ship Position andUncorrected GPS Position (5 Satellites) 12 [ay.. 171

7.34. Distance Between T-2 Ship Position andUncorrected GPS Position (5 Satellites) 13 lay.. 172

7.35. Distance Between T-2 Ship Position andUncorrected GPS Pos t on (5 Satellites) 15 [ay.. 173

10

7.36. Distance Between T-2 Ship Position andDifferential GPS Position With Static Correction1 75Set 1 ......... .............

7.37. Distance Between T-2 Ship Position andDifferential GPS Position With Static CorrectionSet 2 ..... .......... ........................ 176

7.38. Distance Between T-2 .Ship Positioa andDifferential GPS Position With Static CorrectionSet 3 ............... ....................... .. 177

7.39. Distance Between T-2 Ship Position andDifferential GPS Position Without StaticCorrection Set 1 ........................... 78

7.40. Distance petween T-2.Sbip Position andDifferential GPS Position Without StaticCorrection Set 2 ................... .... ...... 179

7.41. Distance Between T-2 Ship Position andDifferential GPS Position Without StaticCorrection Set 3 ........... .. 180

7.42. Azimuth Between T-2 Ship Position andDifferential GPS Position Without StaticCorrection Set 1 .............. ............. . 192

1 7.43. Azimuth Between T-2 Ship Position andDifferential GPS Position Without StaticCorrection Set 2 .. ... .... .............. 83

7.44. Azimuth Between T-2 Ship Position andDifferential GPS Position Without Static* Correction Set 3 ............................... 185

7.45. Distance Between ?-2 Ship Position andDifferential GPS Position Without StaticCorrection Set 4 ....... ..................... . 186

7.46. Azimuth Between T-2 Ship Position andDifferential GPS Position Without StaticCorrection Set 4 ............................... .187

7.47. Distance Between T-2 Ship Position andDifferential GPS Position Without StaticCorrection Set 5 .. .. ... .. ...................... 189

7.48. Azimuth Between T-2 Ship Position andDifferenLtial GPS Position Without StaticCorrection Set 5 ............................ 190

11

7.49. Distance Between T-2 Ship Position andDiffereiotial GPS Position, Without StxtiCcorrection Set 6 ...... .... .. 9.*.... 191

7.50. Azimuth Between T-2 $h'~ Poiion andDifferential GPS Posit ig 11 thloat StaticCorrection Set 6 ** *. ............* ........ 192

7.51. Distance Between T-2 Ship Position andDiffere~itial GPS Position Without StaticCorrection Set 8 ..... ..... ................... 193

7.52. Azimuth Bqtween. T-2 Ship Position andDifferential GPS Position Without StaticCorrection set 8 .................... ........ .194

7.53. Distance Between T-2 Ship Position andDiffereotial GPS Position Without StaticCorrection Sets 9 & 10 .*... ........ *............ 196

7.54. A:;iiiuth Between T-2 h4.p Position andDiffere~tial GPS Position Withouit StaticCorrection Sets 9 & 10.....................19

7.55. Distance Between T-2,Shiip Position and*Differential GPS Position. Without StaticCorrection Set 11 .. ......... ............. 198

7.56. A;imuth Bqtveen T-2 $h~.p Position andDifferential GPS Position. Without StaticCorrection Set 11 ........ .. ... ... .. .. .. . .. . . ... 199

7.57. Di.stance Between ?-2.Ship Position and*Differential GPS Pos-tion Without StaticCorrection Sets 12 & 13 . . . .......................... 201

7.58. A;imuthBptween T-2 ;h~ Position andDi fere tial GPS,3osi.ti n Without StaticCorrection Sets 12& 13 000. . . . . . . .... . . . ... .202

7.59. DitneBetween T-2 Ship Position andDifferential GPStPosition Without StaticCorrection (5 Satellites) 12 M~ay ............... 203

7.60. Distance Between T-2 Ship Position andDifferential GPS ?osition Without StaticCorrection (5 Satellites) 15 day ............... 204

7.61. Latitude Comparison Sets 1, 2 ........... 206

7.62. Longitude Comparison Sets 1, 2 ............. 207

12

7.63. Latitude Comparison Set 3 ...................... 208

7.64. Longitude Comparison Set 3 ..................... 209

7.65. Latitude Comparison Sets 4, 5 .................. 210

7.66. Longitude Comparison Sets 4, 5 ................. 211


7.68. Longitude Comparison Sets 6, 7 ................ 213


7.70. Longitude Comparison Sets 8, 9 ................ 215

7.71. Latitude Comparison Sets 10, 11 ................ 216

7.72. Longitude Comparison Sets 10, 11 .............. 217

7.73. Latitude Comparison Sets 12, 13 ................ 218

7.74. Longitude Comparison Sets 12, 13 ............... 219

1i

gO ABRREVTATIONS

BPS Bits Per Second

C/A Coarse Acquisition Code

CDU Control Data Unit

CHAS Circular Map Accuracy Standard

DNA Defense Mapping Agency

GMT Greenwich Mean Time

GPS Global Positioning System

HTC Hydrographic/Topographic Center at DNA

LMT Local Mean Time

3CS GPS Master Control Station

MIS NVUE Instrumentation System3VUE MANPACK Vehicular User Equipment

NAVSAT Navy Navigational Satellite System

NAVSTAR Global Positioning System

OTES NOAA Office of Ocean Technology and EngineeringSupport

P Precision Code

PVC Polyvinyl Chloride Pipe

SAMSO Space and Missile Systems Organization

UR1 User Range Error

USAF United States Air Force

?JGS 1972 world Geodetic System of 1972

ja4

ACKrIO!ILEDGEMENTS

We would like to thank all the NPS students and other individuals who

contributed their time and expertise during the long nights at the T-2

locations--without their support this thesis would not have been possible.

Thankts go to: Marjorie Amazeen, Bill Anderson, Gene Brown, Bob and Hac

Clar!k, Ron Dickerman, Barry Donovan, Pat Eaton, Mike Ellett, Jerry and

B.J. Mills, Dave tinkel, Roger Parsons and Annette Parsons, Alan Shaffer,

Keith Toepfer, Gerry Wheaton, and Jon Winter.

Special thanks also to Ted Calhoon for his cooperation and to the Captain

of the ACANIA, Woody Reynolds; and the crew members who contributed

their ideas for the equipment installation onboard the ship.

'!e would like to thank Mr. Knute Berstis and Mr. Gary bhitsell from OTES

for their time and efforts in making the thesis possible, and Cdr. A.

Durkee, of the Joint Programs Office, for his interest and bipport in

providing the GPS receivers.

15

I. TO ZO

Shipboard operations for the purpose of oceanographic

research, commercial transport, tactical operations, or

hydrographic surveying require accurate and dependable

positioning systems. The operation of most electronic

positioning systems is hampered by atmospheric effects on

the propagated signal and design limitations which restrict

operating ranges, positional accuracy and dependability. A

universal positioning system that provides accurate

information to the user regardless of atmospheric conditions

and that is free of range limitations is desirable. with

the advent of satellite positioning techniques, such a

system is now available.

The Navy Navigation Satellite System (NAVSAT), or

TRANSIT, has been operational since 1964 and has proved

dependable and sufficiently accurate for certain users.

Although developed for military uses, the application of

TRANSIT in the civilian community has continued to expand at

an exponential rate. The system, however, is limited by the

time interval between fixes, which is 90 minutes on the

average, and the requirement for precise knowledge of the

ship's velocity. Although the system does not provide

16

One method of improving system accuracy is by using two

receivers simultaneously in a technique termed

translocation, or differential mode. The method is based on

the assumption that signal propagation errors to two

proximate receivers are nearly the same. One receiver is

placed over a known position, and the variation of the

observed position with time is determined relative to the

absolute position. This variation is applied to the

observed positions determined by a simultaneously operating

receiver located at another location--on a survey vessel,

for example. The assumption underlying this technique is

that the distance between the two units is not sufficient to

introduce a significant additional propagation error. The

resulting position of the translocated receiver should be

more accurate than the position obtained using a single

receiver.

For hydrographic operations, the translocated Global

Positioning System (GPS) provides an alternative to

presently used positioning systems. The satellites orbit at

20,000 ki, which theoretically would allow receiver

separation of as much as 500 km without causing significant

additional system error. With this flexibility, one

17

sufficient positional accuracy for hydrographic surveying,

it has been used for geodetic application by the Defense

Mapping Agency (DNA) using fixed shore-based receivers to

establish positioning networks in inaccessible areas

worldwide. CRef. 1]

The Global Positioning System, or NAYSTAR, is a more

sophisticated system designed to replace TRANSIT in the

mid-1980s. Six satellites are currently in orbit with

twelve more to be launched before the operational date of

1987. The 18-satellite orbit configuration will provide a

multi-user, passive system with 24-hour availability, global

coverage, and possible fix intervals of less than one

second. The operational network consists of three

elements--the satellites, the ground-based control segment,

operated by the Space and Missile Systems Organization

(SIMSO), and the user's receiver. Prototype receivers have

been tested under optimum conditions by SAMSO at the Yuma

Proving Ground, Arizona, on both mobile and stationary

platforms. Results show excellent system stability and

accuracy when compared to currently used positioning

systems. [Ref. 2]

18

shorebased receiver could be placed at a central location

and a survey of several hundred kilometers of coastline or

offshore area could be completed with only one additional

receiver on the survey vessel. [Ref. 3] The value of this

flexibility would be realized in the significant decrease in

survey planning, cost, and time compared to that spent

presently in conducting hydrographic operations.

The accuracy of the system would be sufficient for all

survey scales. The National Ocean Survey (OS) defines the

required accuracy of a positioning system to be no more than

one-third of the survey accuracy, which is 1.5 millimeters

at the scale of the survey. (Ref. 4] Non-translocated GPS

has been shown to be of sufficient accuracy for 1:80,000

scale surveys and smaller. [Ref. 5]

The object of our research is to determine whether GPS

at the present operational level is adequate for 1:10,000

scale surveys. To do this, a position determined by two GPS

receivers in the translocated mode is compared to the same

position determined by two other methods. If translocated

GPS can be shown to achieve specified inshore survey

accuracies, the system would provide sufficient accuracies

at all smaller scale surveys.

19

L itlr TII

Part of the difficulty in determining this level of

accuracy concerns the reference system used. most geodetic

positions published by the National Geodetic Survey (IGS)

are based on the North American Reference Datum of 1927

(NAD-27) and the Clarke 1866 Ellipsoid, the best

approximation to the shape of the earvk f-- the continental

United States. The satellite syst* ,. ,ever, use a

reference system which is not bast.. rn z portion of the

earth's surface. A mass-centered *#!.- seid referenced to

the earth's center of mass has b.4P. zcoputed using observed

gravity and astronomical data. The best fit to this global

ellipsoid is the World Geodetic System of 1972 (WGS-72),

which is utilized by both ,TRANSIT and GPS. Transformation

equations exist which convert NAD-27 coordinates to WGS-72

coordinates, but uncertainty in the gravity model produces

errors as large as 10 meters due to the transformation

prccess alone. (Ref. 6]

To establish all positions on the same reference

ellipsoid for this research, a TRANSIT receiver was acquired

from DNA and a first order Doppler station was established

at the central location of the test area. A third order

geodetic survey was conducted from the TRANSIT station to

21)!

establish two additional stations on the UGS-72 ellipsoid.

Two single channel aanpack GPS receivers were obtained from

SANSO. Stability testing and a receiver comparison test

were conducted prior to shipboard testing to letermine a

quantitative measure of receiver performance. One of the

receivers was then placed over the TRANSIT station mark and

the other aboard a survey vessel which maneuvered along

j predetermined track lines. Data were recorded via an

interface to an HP-9825 computer and printer at each site.

Simultaneously, two additional independent measurements were

made of ship position--one using a range-azimuth short

range microwave positioning system (ARTEMIS) and the other a

position determined by a least squares solution of three

lines of position observed from three 1-second theodolites

set on the three pre-established VGS-72 locations. The

comparison of the ARTEMIS and theodolite-derived ship

positions with the translocated GPS ship position provided

the test results.

II. SOUIMM DESCRI PTIO ZLP BILITIES

Two positioning systems were chosen for comparison with

the Global Positioning System: a theodolite network and a

microwave positioning system. The theodolite network

consisted of three shore-based units so as to provide three

lines of position. The microwave positioning system

prcvided a range and an azimuth from a known location on

shore. Both systems are accurate at short ranges and

represent two contrasting methods of positioning--visual and

electronic. The theodolite method of positioning at short

range has been and continues to be used for hydrographic

operations as an alternative to electronic systems when

gecmetry and system accuracy limitations impose the need for

a more flexible positioning system. The requirement for

experienced theodolite observers as well as the

communication and logistics difficulties involved limit the

use of this method to special cases.

This chapter describes the ARTEMIS short-range

electronic positioning system, the operation of the

theodolite network as used during the comparison test, and

the TRANSIT satellite system.

22

A. ARTEMIS POSITIONING SYSTEM

The ARTEMIS position fixing system is a short-range

microwave system (to 30 km), built by Christiaan

Huygenslaboratorium B. V., in the Netherlands. It has been

used operationally (primarily in Europe) since 1972, when

the system was introduced at the International Hydrographic

Conference in Monaco. At the present time. fifty systems

are in use. The system is unique in design, is highly

accurate at short ranges, and is particularly useful for

small harbor surveys, river surveys and relative positioning

uses. For this study, the ARTEMIS system, along with

technical assistance, was obtained from the American

distributor, MARINAT Corporation of Houston, TEXAS.

ARTEMIS is a range-azimuth positioning system which

employs a unique tracking method. A stationary or fixed

unit is placed over a known position on shore. Using a

telescope that is mounted on the antenna, the antenna is

sighted onto another known point and the known azimuth

between the two locations Is entered into the system via a

digital display on the fixed unit. The unit is then locked

into place for operation. I mobile unit is placed on the

vessel whose position is desired.

23k4

The mobile and fixed antennas resemble radar antennas

which, while operating, track ,tach other so that the two

antennas are always parallel to each other and perpendicular

to the line of position between them. A microwave link is

established between the two antennas during periods of

measurement and information is relayed between the mobile

and fixed units. (A voice link is also available during

operation.) The fixed unit provides the direction and the

azimuth of the mobile unit with respect to the point sighted

on by the telescope. At the same time, the mobile unit

transmits a coded interruption of the signal to the fixed

unit, which replies with the same interruption. The time

difference between the transmission and reception at the

mobile unit determines the distance between the two

a ntennas.

Distance accuracy is determined by the number of

measurements averaged over time and is available in two

modes--a dynamic mode which averages 1000 measurements or a

static mode which measures 10,000 measurements per displayed

distance. The azimuth and distance information is displayed

by the mobile unit so that personnel are not required at the

stationary unit once operation has begun. The system can be

24

operated in long or short-range modes, which determine the

amount of power output for the system.

The microwave frequencies used are 9.2 and 9.3 GHz.

Problems arising with signal interference, observed with all

microwave positioning systems, therefore can occur with

ARTEMIS. The regions of range holes, or multipath

interference, can be computed for up to 10 ka range by prior

knowledge of the antenna elevations above sea level and the

operating distance from the shore-based antenna. [Ref. 7]

Each antenna is designed with a 22' vertical beam width

and a 20 horizontal beam width. A power source of 22 volts

DC per unit is required, with an average current load of 2.5

amps.

The accuracy of the system depends on both the distance

and the angle measureC. The angle measurement has a 2 a

probability of measuring within 2' of arc (0.0330 . The

distance measurement has a 2 G probability that ..he mean

distance value will vary within + 1.5 meters.

One of the advantages of ARTEMIS over other electronic

systems is that the two lines of position defining the fix

always intersect at 9 0 . A unique advantage is that the

coverage around the fixed unit can be circular, with a

2E4

maximum range of operation to 30 ka, dependent on antenna

heights. The short range applications of the ARTEMIS fill a

need not provided by many currently used positioning

systems.

For this study, the ARTEMIS was used in conjunction with

the positions derived from the theodolite network to

determine the best estimate of the shipis position. The

accuracy of the ARTEHIS was more than adequate for the

ranges encountered during the test, and the ease of use and

operational versatility of the system made its use very

advantageous.

B. THEODOLITE NETWORK

A network of three Wild T-2 theodolites was used as the

control positioning system. The theodolite is widely used

for surveying as a means of measuring angles between two

observational locations--either in a horizontal plane for

traverse or triangulation surveys, or a vertical plane for

astronomical observations.

The use of three theodolites as a short-range

positioning system was an accepted method of positioning

prior to development of electronic distance measuring

systems. There remain disadvantages when using electronic

26

positioning systems for hydrographic operations at short

ranges (1-2 k). Sost short-range electronic systems are

designed for measurement of position at line-of-sight

ranges. This region includes a large portion of inshore

* hydrographic surveys. Inshore positioning involves more

thorough planning to obtain the required accuracy for the

scale of the survey. For example, in a harbor or cove it

may be difficult to place two range-range electronic units

so that the area to be surveyed is not within the region of

positioning system uncertainty, or 300 (radially) on each

side of the baseline between the two units. To reposition

the electronic units as often as is required to obtain

-optimum geometry is time-consuming and requires manpower and

extensive planning, and can result in operational delays.

Interference of the signal due to reflection from a calm sea

surface, or sultipathing, is also a problem at short ranges

for microwave systems.

Theodolites can be set at any intervisible location with

preplanned geometry and can be used for the entire region to

be surveyed. This method requires less expensive equipment

(three theodolites) , more manpower (a recorder- and an

observer at each s.ite), communication between the three

27

locations (via portable radio, for example) and reasonable

weather conditions. The basis of this procedure is that

three angles observed from known locations to a

predetermined point on the vessel constitute three lines of

position. After removing all possible systematic errors and

through a least squares solution of the three lines of

position, the position of the vessel is determined. k major

disadvantage is that observational errors are not determined

in real time because the data cannot be processed

immediately.

The basic procedure for the method requires

predetermined sites which are known relative to each other.

This may be achieved by using published geographic positions

or by conducting a local plane survey to establish new

positions using preestablished stations nearby. Either way,

all three locations must be based on the same datum and

preferably permanently positioned by a disk or a station

mark.

Each theodolite was aligned using a plumb bob or optical

plummet over one of the known station marks or disks. By a

predetermined scheme, each site was marked so as to be

clearly visible to the other sites (lights were used for

28

this study). The telescope of each theodolite was sighted

onto one of the other sites and the horizontal circle set to

read an arbitrary value close to 00. The crosshairs in the

telescope were set to intersect the mark. When the

theodolite was exactly sighted, the horizontal circle was

read and recorded. (At this time the lights were turned

off.) The operation began when all three sites completed

this step, called the "initialling procedure", or "initial

pointing". The instrument was sighted on a point on the

survey vessel that was visible to all operators. The ship

was tracked by the operators by using the micrometer wheels

on the instrument until a "mark" was called over the radio

by the recorder at the control theodolite location. At that

instant, the operator stopped tracking and the horizontal

circle was read and recorded. This observation constituted

a fix. After a series of fixes were recorded, or when the

survey vessel reached a predetermined location, a set of

fixes was completed. The lights were then reactivated at

each site, each observer intersected the original mark, the

horizontal circle was again read and the "final pointing"

was recorded. If necessary, the instrument was relevelled.

The process was then repeated by starting with the

initialling procedure.

29

It was assumed that the theodolites were in good

operating condition. Collimation error had previously been

determined. The observers were kept at the same locations

throughout the test to minimize the error due to individual

observer variations. Since only one portion of the

hcrizontal circle was used, and no reverse readings

observed, a systematic error was introduced. This was

assumed to be minimized during computation of the vessel

position.

The observations were corrected for collimation error

and for pointing error by evenly distributing the difference

between the initial and final pointing values. The

resulting angles were processed using an intersection

computation with a least squares fit applied to optimize the

result. The computations for this study were processed

using an HP-9815 computer and software prepared by Cdr

Ludvig Pfeifer, YOlA. The resulting position was used as

the best estimate of the position of the ship.

C, TRANSIT POSITIONING SYSTEN

The TRANSIT, or Navy Navigation Satellite System, is the

most widely utilized satellite positioning system at the

present time. The network of satellites and ground tracking

30

stations has been in operation since 1964 and is becoming

widely used by both the military and civilian communities.

It is expected to continue operation until at least 1995,

when the Global Positioning System is expected to replace

it.

The system consists of i series of TRANSIT satellites

(fig. 2.1.) in fixed polar orbits of 1095 km elevation,

fcrming a network within which the earth rotates (fig.2.2.).

The orbits are spaced so that satellite passes occur, for a

given ground location, at intervals of between 35 and 100

minutes. Each satellite travels an observable distance of

between 4400 and 7000 km per pass, providing a sufficient

time interval and baseline over which the user can record

data.

The idea of positioning an observer on the earth's

surface by using the satellite as a reference system

originated from observation of Sputnik 1, launched by the

Soviet Union in 1957. By recording the change in frequency

of the satellite signal as the satellite passed overhead, a

Doppler shift could be measured. The Doppler shift gives a

spatial relationship as a function of time between the

observer and the satellite. If the position of the

31i

Fig. 2.1. 7RANS17 Satellite

32

:73

ig. 22. Tar'sit Satellite oroital Nlet'Mor

33

satellite is known at the time of observation, the

observer's position can be determined.

In the TRANSIT system, the satellite positions are

precisely determined by tracking stations located at

Prospect Harbor, Maine; Rosemount, Minnesota; and Wahiawa,

Hawaii. The tracking information (in the form of the

Doppler frequency shift as a function of time) is relayed to

the computing center at the Naval Astronautics Group

Headquarters, Pt. Mugu, California. Here the actual

satellite orbits are computed. This information is used to

predict orbit information. Parameters of this predicted

orbit are relayed to two injection stations at Rosemount and

Pt. Mugu. During the next satellite pass, the information is

relayed to the satellite as part of the navigation message,

with new updates occurring every 12 hours.

The satellites therefore contain predicted orbit

information which is received by the user via a satellite

receiver. Several receivers are available with complexity

and positional accuracy dependent upon the needs of the

user. A computer is normally required to determine the

user's mobile or fixed position in real time from a single

satellite pass.

34

For many users having a dual channel receiver and knowL

velocity, this method of positioning provides sufficient

accuracy with typical error of 27 to 37 meters root sum

square. Optimum accuracy is obtained, however, if the

receiver position is fixed, and several consecutive passes

of the same satellite are recorded. During each pass,

lasting from 10 to 16 minutes as the satellite travels from

one horizon to the other, many Doppler counts (observations

of shift in frequency as a function of time) are recorded.

After a predetermined number of passes are stored, the data

can be forwarded to the Defense Mapping Agency in Brookmont,

Maryland. There the precise ephemeris data from the ground

tracking system is applied to the data from the recorded

passes to provide the optimum solution. The resulting

position has a typical error of 6.3 meters root mean square

for a series of passes, with a 3-dimensional result of 1.5

meters per axis repeatability after 25 precise ephemeris

passes. This method is used to determine positions in

remote areas by DA geodetic survey groups using the

Sagnavox AN/PRR-i Geoceiver (fig.2.3.|. (Ref. 8]

The TRANSIT system uses the World Geodetic System of

1972 (UGS-72) reference ellipsoid which is based on the

35

,AN/IPqR-14 GEocE:'/ERANTENNA

F ig. 2.

~36

Universal Space Rectangular Coordinate System--an earth,

mass-centered coordinate system. The VGS-72 ellipsoid

describes the best fit of the entire earth to a mathematical

model. The satelliie orbits are computed using this

geopotential model, the accuracy of which is based on

4. gravitational and astronomical data.

For the present study, a Magnavox kN/PRR-14 Geoceiver

was acquired from the Department of Satellite Geophysics,

Satellite Tracking Branch, DNA. The receiver was placed

over a fixed point and 4O passes of one satellite were

observed. The procedures described previously for an

optimum solution of point positioning were followed. The

resulting latitude, longitude, and height relative to the

WGS-72 ellipsoid are coordinates of a first order TRANSIT

position as defined by DNA standards. This position was

used as the basis for horizontal control for the study (see

test procedures).

37

III. S SYSTEM DESCRIPTION N OPERATIONAL CAPABILITIES

A. SYSTEM DESCRIPTION

NAVSTAR, or the Global Positioning System, was designed

for maximum operational capability--high accuracy,

twenty-four hour access, passive user, and world-wide

coverage. A brief description of the GPS, and the

satellites (NAVSTARs), will serve as a background for the

more in-depth discussion immediately following.

GPS navigation is based on the measurement of four

ranges, or pseudo-ranges, from each of four satellites with

known, or predicted, positions. Each range determination is

computed using the propagation velocity of the signal

multiplied by the satellite-to-user travel time. Satellite

transmission time is included as part of the navigation

signal message. From the four known ranges, the user

position and user clock error are computed on a three-axis,

orthogonal coordinate system. This computation is then used

to transform the user position onto the WGS-72 ellipsoid.

The remainder of this chapter will provide the

background for GPS development and the three major system

segments of the GPS: the space system segment, the control

38

segment, and the user system segment. Included in the user

system segment will be a discussion of the nanpack receiver

as developed by Texas Instruments Corporation used for this

study.

With the successful operation of the TRANSIT system in

the early 1960's, the Department of Defense (DOD) began to

consider the specifications of, and the technology required

for, the development and implementation of a second

generation system. On 17 April, 1973, the Deputy Secretary

of Defense stated that a Defense Navigation Satellite System

was to be developed. [Ref. 93 Using the new synthetic

oscillator technology, and the Department of Defense surface

and nearsurface navigation objectives and requirements, a

DOD interdepartmental task force developed the GPS as the

second generation satellite navigation system. With the

United States Air Force (USAF) as the managing agency, the

GPS became a joint service program, with representatives

from the DNA, Army, Navy, Maiines, Coast Guard, and the

North Atlantic Treaty Organization. [Ref. 10] The Joint

Programs Office was established within the Air Force Space

and Sissile Systems Organization to design, develop, and

implement the GPS program.

39

GPS was conceived and designed to provide continuous,

real-time, world-wide navigation coverage for subsurface,

surface and nearsurface operational vessels, vehicles, and

aircraft. Accuracy of better than 10 meters was desired

SIl under all conditions, with navigational capabilities to

update and interface with other forms of existing and

,Al projected navigational systems. knother requirement of the

GPS satellites was nearly trouble-free and simple user

operation, even in low signal-to-noise environments. An

encoded signal structure was required for the satellite

system for security reasons. In addition, the signal

transmitted by the NXVSTAR satellites had to include other

operational information in conjunction with the navigational

message. To provide for navigational requirements in event

of hostilities, the GPS was designed to permit the use of

redundant satellites.

To achieve these objectives and requirements, the GPS

program was organized according to three developmental

tasks: concept validation, full-scale engineering

development and systems tests, and production and operation.

Starting in 1974, the concept validation task was initiated

with the merger of the TIMTION series satellite program and

40

the USAF System 21B (a highly accurate three-dimensional

navigation system) program. Tests with the Navigation

Technology Satellite, NTS-1, utilizing two rubidium clocks

(accuracy 5-10 parts per 10 ) and an orbit 7500 nautical

miles (nm) above the surface and three ground based

receivers (called inverted range) provided concept

verification. [Ref. 11] Additional tests were conducted

with an improved NTS-2 satellite positioned at an altitude

of 10,980 nm and containing two cesium clocks (accuracies on

the order of 1 part per 10 ). [Ref. 12] Orbital height of

10,980 nam permitted the evaluation of the effect of the

planned GPS constellation altitude on navigational

accuracies using the improved cesium frequency standards.

In 1980, NTS-3 was launched and the Advanced Developmental

Model (ADR) hydrogen maser frequency standard evaluated for

use on the GPS (providing an expected frequency accuracy of

I part per 1013). [Ref. 13] The results obtained from the

RTS-3 satellite have not yet been released. Since 1978,

additional satellites have been orbited to provide the

six-satellite constellation required to fully evaluate the

first development task--concept validation, Phase I.

41

During Phase I, user requirements were consolidated to

allow the development and construction of four basic types

of test receivers: continuous tracking receivers,

sequential tracking receivers, icw cost receivers, and the

Manpack receivers. [Ref. 14] These receiver types were

then constructed for Phase r testing by various contractors

and their performances evaluated aboard helicopters,

aircraft, ships, vehicles and personnel at the Joint

Programs Office's testing facilities. Phase I, concept

validation, was completed in late 1979 and Phase II testing

was begun in early 1980 with full scale engineering,

developmental, and systems tests.

Phase II tasks, finishing the operational and control

segments of the testing, are scheduled for completion in

aid-1983. The operational control station will be relocated

at a central continental United States site during this

phase. Additionally, all satellite monitor receivers at the

remote monitoring stations will be updated to operational

status. User equipment and receiver selection will be made

in two increments. Four contractors will be asked to

compete for the second increment of the receiver selection

process by providing upgraded and refined user receivers for

42

extensive testing and evaluation. Two of the four

contractors will then be selected to continue the

competition and will provide additional receiver design,

refinement and modificaticn, leading to prototype production

and extensive field testing.

When Phase I is completed in mid-1983, Phase III,

production, will commence with the selection of one of the

two contractors to produce the user receivers. .anufacture

of these receivers will commence shortly after selection,

and commercial availability of recaivers is expected in

1985. (Ref. 15] The GPS should be fully operational with

world-wide coverage using a constellation containing a total

of eighteen satellites in 1988. (Ref. 16, 173

The GPS is considered to be composed of three

inter-related system segments: the space system segment,

the control segment, and the user system sagment. Of these,

the space system segment may be considered the most

important since the other two require the space segment to

be operational at all times. Operationally, the space

system segment consists of 18 satellites placed in orbits at

altitudes of 10,898 na and having orbital periods of about

12 hours. [Ref. 18]

43

The original plan for the satellite constellation

consisted of 24 satellites in three planes of 8 satellites

each. These three planes were to be placed 45 0 of longitude

apart and at an inclination of 63o to the equator.

[Ref. 19] As a result of 1980 budget cutbacks the space

system segment has been reduced to a total of 18 satellites.

(Ref. 20] Several different types of 18 satellite

constellations have been proposed, but all have drawbacks.

One which appears the most promising is the 'Nonuniform 181

constellation containing 6 planes, each of which contains 3

satellites spaced 120 ° apart (fig.3.1.). The areas of user

outages (fig.3.2.) are a disadvantage of this configuration.

The user outages are indicated by the darker areas shown on

Figure 3.2. These outage areas will appear to rotate about

the globe at the indicated latitudes, and positioning with

the GPS will be intermittent, or of decreased accuracy for

users at these latitudes. The final constellation

configuration is to be selected early in 1982. DEef. 21]

Each NAVSTAR satellite has the following

characteristics: weight of 982 pounds, length of 17.5 feet

from tip to tip of the 5 square solar panels, three atomic

44

* Proposed Satellite

o eleted Satellite

Fig. 3I. The Ionuniform 18 Satellite Constellation

45

0 3 L

7.o0U

Vi

apm. ;s

frequency standards (clocks) , and three nickel-cadmiumL

batteries for operations in darkness and during peak load

periods. Stabilization of each NATSTAR is maintained by

four skewed reaction wheels which produce three-axis

stabilization. NAVSTAR station-keeping and momentum damping

capability is provided by a hydrazine propulsion system

within each satellite. Life expectancy of each unit is five

years, and expendables are expected to last for up to 7

years with normal usage.

Figure 3.3. illustrates the parts of each NAVSTAR. k

twelve element shaped-beam helix antenna transmits the

navigation message on two L band frequencies to user system

segments. Simultaneously it receives satellite status,

clock corrections, ionospheric data, and ephemeris constants

from the control segment on two S band frequencies.

(Ref. 22] The two coherent L band frequencies are centered

on 1227.6 MHz (LI), and 1575.4 MHz (L2). [Ref. 23]

[Ref. 24] Both frequencies are necessary to determine the

total ionospheric delay correction for signal propagation

time (Table I ). Zrrors produced by the ionospheric delay

in signal velocity contribute most of the error in the

pseudo-range. By utilizing the Li and L2 frequencies and

47

VA

Fig. 3. 3. 'IAVSTAR Satellite

48

TABLE I. lonospheric Group Delay as a Function of Frequency

R A+ B+ C +ASC f2 -:3 4 ..... C f2

GD(neglecting third order and higher terms which contribute minimalerror)

where A ionospheric delay constant of conditions

R = true range

F: C = speed of lightf = carrier frequency

B K(averaged jarth magnetic field strength alongthe path)

GD = group delay between points G and D

Term rlealected: Range Error atf = 1.5 GHz

B4!5 1 inch

f3

C3 inches

f 4

49_ _ ... __ __ N

the following algorithm, the error is decreased but not

totally eliminated. [Ref. 25]

T -T = A

GDLI GDL2

Where: A-r GDLI f2

L L

where A = ionospheric constant and T GDLI, GL are the

respective ionospheric group delays at frequencies L i and

L L2. Both frequencies are also modulated to transmit the GPS

time codes and the navigation message.

GPS time is transmitted using two pseudo-random noise

chip codes--the Coarse Acquisition (C/&) and the Precision

(P) codes. l3odulat ion on the Li frequency carries both the

C/A and P codes in phase quadrature, while the L2 frequency

carries only the P code. These codes identify each acquired

satellte by matching the unique pseudo-random noise code

pattern generated by each satellite with similar user

generated codes.

50

•~ ~~~~ ~ ~ 1 .. . . . -- ,. .. -.T - - .. - ..

Both codes are required to measure the navigation signal

propagation time. The phase shift is measured and used to

match the user generated signal with the incoming satellite

signal. (Ref. 26] The C/A code has a repetition period of

one millisecond and has a short code stream, transmitting

1.023 million bits per second (bps). This code is easily

acquired and matched with the user generated pseudo-random

noise code but only provides a coarse time signal for

computing the pseudo-range. However, the P code has a long

code stream, repeating only every seven days, and it is

transmitted at the higher rate of 10.23 million bps. With

this higher data transmission rate and long repetition

period, a more accurate time signal is available for

pseudo-range determination. However, the long period and

high data rate makes the P code difficult to match with the

user-generated pseudo-random noise code pattern unless the

receiver has a highly accurate time standard, and the

approximate receiver location is known. In lieu of this,

the usual method of acquiring the P code is by matching the

C/A code of the desired NAVSTAR, then using the Handover

Word (HOW), to locate the correct P code sequence within the

7-day pseudo-random noise code. GPS time is transmitted

51

"Pow- -71

with the HOW every six seconds and provides very accurate

GPS time to the user. Both of these codes provide the

proper time reference for computation of the pseudo-range.

The C/A code provides a less accurate time reference than

the P code. The selection of the code depends only on the

navigational accuracy desired.

The navigation message is also modulated on frequencies

Li and L2 but at a much slower rate (only 50 bps) than

either the C/A or P code. On both frequencies, the

navigation message is identical and consists of 5 subsets of

6 seconds duration each. Bach navigation message begins

with the telemetry message (TLR) and the HOW message, which

are used to transfer from the C/A to P codes. The remainder

of the entire navigation message allocated to each

subsection is transmitted after the TLM and HOW messages.

Information in the navigation message includes approximate

satellite ephemerides, status of other NAVSTAR satellites,

parameters for clock corrections, atmospheric signal delay

corrections, satellite performance status, momentum dump

status since last upload, time since last upload, almanac

and identification codes for all satellites, and provisions

for inclusion of any special or important messages from the

52

L , , .-. .. .. A

control system segment to the user. ERef. 27, 28J

Navigation is thus achievable by combining the information

transmitted by the space system segment in the form of the

C/A and P codes and the navigation message.

Accuracy and proper function of GPS are governed by the

second segment--the control segment. The oontrol segment

monitors and uploads the individual NAVSTAR satellites,

checks operational characteristics, determines the ephemeris

and almanac data and corrections required, computes the time

delay and clock corrections for each satellite atomic

standard, determines the atmospheric delay corrections for

pseudo-range computations and includes special messages in

the body of the navigational message. These functions are

under the command and operational control of the Master

Control Station, currently located at Vandenberg APB,

California. Four unmanned remote monitor stations, located

in Hawaii, Guam, Elmendorf APB, Alaska, and Vandenberg APB,

collect the pseudo-ranges to each satellite, the change in

pseudo-range of each satellite signal, local meteorological

data, and the remote monitor station atomic standard

parameters. Each set of c,.a is transmitted, on request, to

the Haster Control Station, where all the data is combined

53

and the satellite upload corrections and parameters are

computed. These corrections and parameters are then relayed

to the Upload Station, Vandenberg AFB, for transmittal to

the respective NAVSTAR satellites. The uploads occur when

the satellites are first visible to the Upload Station and

,. when the User Range Error exceeds 4 meters, as computed by

the Master Control Station. [Ref. 29, 30, 313

The last segment, the user system segment, combines the

information provided to the space system segment (by the

control segment) with the information received from the

sp.ce system segment to obtain the pseudo-ranges from four

NAVSTAR satellites. These ranges are computed using the

following algorithms:

a i = C(tR-tti) - CAtAi

where =i = true slant range to satellite i, c = speed of

light in a vacuum, tR = time of received GPS signal by user

receiver (assumed to be simultaneous from all satellites),

tt time of transmission at satellite i, and

LtA- =atmospheric propagation delay time from satellite i

54

(fig.3.4.). Assuming a user clock error, the true range can

be expressed in terms of a pseudo-range:

RP. R + CtA. + c(A. AtS .1 1

where Rp. = pseudo-range including clock error, AtR = time

delay in the user receiver cf the received signal, and

At5 time delay in satellite transmission from satellite

i.

To determine the slant ranges:

i- i Ji-) 2 + (Ys -Y) 2 +(Zs_Z) 2

s i- =_X) 2 (YS iY) 2 (Zs i_ Z) +a Ac+tA C(AtRAtsi

The following quantities are known: x , y , z obtained

from navigation ephemeris data; A tAi computed using the

ionospheric delay correction obtained from observing

frequencies Li and L2 (tropospheric delay neglected); A tSi

obtained from the navigation message as the satellite clock

error. Thus, there are four unknowns: x, y, z (user

position coordinates) and AtR ( user clock error or bias).

To solve this set of equations, four satellites must be

observed simultaneously. [Ref. 32]

55

Should only three satellites be observable, the altitude

may be assumed and the solutions can be computed for the

other two location variables and the user clock error.

Solutions of this type (altitude hold) significantly degrade

position accuracy.

The above description is over-simplified since the

relativistic effects of N&VSTAR satellite velocity, as

observed by the user, are neglected in terms of the

satellite time delay correction and the atmospheric

propagation delay correction. The actual user velocity, in

terms of velocity in the x, y, and z directions, is

calculated using the Doppler shifts of the two frequencies

Li and L2 and the estimated satellite position taken from

the ephemeris data. With the relativistic effects as an

integral part of the data set, solution of the navigation

position problem becomes more complex. Phase I receiver

manufacturers have included them in the solution of the

problem by several different methods. Sethods of solution

depend on the usage of the receiver (whether static, high or

low dynamic) and the expected cost/benefit ratio for the

receiver (in terms of the total cost, weight and size).

Current state-of-the-art design technology and navigational

56

0

I -

u 4~

,, - .4Jt4-

.4.J J- -

U,-l

0 -

0

•- -.

I CC

(00

C-C'

Fi.34 suoPag ecito

I57

algorithms, developed since Phase I, will be included in the

Phase Il receivers and will significantly improve accuracy

of future receivers.

B. IANPACK RECEIVER

The Phase I receiver used for this research was the

Texas Instruments GPS Manpack, also known as the Manpack

Vehicular User Equipment (MVUE). MVUE is a single channel

low dynamic receiver, usable for receiver velocities from 0

to 25 meters per second. [Ref. 33] System configuration is

illustrated by figure 3.5. Composing the system are the

Manpack control data unit (CDU), antenna and vehicle mount,

and vehicle power converter. The Manpack is the receiver

unit (fig. 3.6.) with provisions to connect an external data

output, through the NVUE Instrumentation System ("IS)

interface, figure 3.7. Facilities for an internal

preamplified antenna mount, or an external preamplified

antenna, and an attachment for either an external power

source or the integral battery pack are also shown (figure

3.7). Together, the NVUE and the SSP/9900 microprocessor

data processing unit have storage capacity for 47,360

sixteen bit words. This storage capacity allows

[interrogation of NAVSTAR satellites sequentially, with the

58

~77

-V

1c

LJi

C ILU X-

It tj

59.

20 0 04 4w

Z0 C i z

w ai

LL.

J w CD'

w z oL4c Z W

60

LIMITE

-C-

611

receiver computing a position fix every 24 seconds. The

relativistic velocity effects on the received signals are

determined by matching the incoming frequencies with the

expected frequencies. The user velocity relative to the

satellite is determined from the cbserved frequency shifts.

The nominal frequency shift for each satellite transmission

is computed using the simultaneously transmitted ephemeris

data. Total weight of this receiver unit is 33.5 pounds,

and it has a volume of 17,958 cubic centimeters.

The second element of the NVUE, the CDU (fig. 3.8.), is

the input/output device for the non-automated Hanpack user.

Inputs are entered by keying the pressure pads set in six

rows of four pads per row. To obtain a navigational

position, the approximate user latitude, longitude, and

Greenwich Mean Time (GMT) are keyed. After the warmup

period of 3 to 5 minutes, the satellite signals are

received, and the CDU will display the actual position.

Additional input may include NAVST&R satellite selection,

user altitude, and other position datums. Output options

include position, GMT time, and system error messages.

Other inputs and outputs are listed in the NTUE manual.

CRef. 34]

62

Li

-- I I T-- 4 -+

I Ii 1i II I I

! *I I IL I

S I I I I It( 1) =(12) 1 (13) 14) 1 (15) l(161) (17) 1 (18) 1019) 1 (20) 1

L . .L _.L . .L .J _ I _... J. . . _ _ AJ

ON AUTO

ENT RAO SEL GRDSTSY- MAN

j.. I CLRLA

AS EE

23 TIM

(9) (10) (1') (12)

GJH 10(7L~.L~

4' 5 6' RNG

MIN OP QR

Fig. 3.3. MVIJE CDU Keyboard/Display Liyout

63

Automated users may elect output either to a radio

interface or through the MIS interface. For this research,

a special interface unit, designed and constructed at the

Office of Testing and Evaluation of Systems, NOAA, was used

to link the HIS interface board within the receiver with a

Hewlett Packard 9825T computer. With this interface, user

latitude, longitude and GPS time were logged onto magnetic

tape for future processing. Figure 3.9. and Table II are

the engineering documents used in construction of the MVUE

interface unit.

Other NVUE elements are the antenna and the antenna

external preamplifier vehicle mount. Located within the

antenna are two antenna subassemblies associated with

frequencies Li and L2. Both of the subassemblies are

connected to the external preamplifier with two lengths of

coaxial cable, enclosed in a flexible conduit. The entire

unit, antenna and preamplifier, is connected to the Manpack

receiver with a power cable and a coaxial antenna cable.

Cable length is determined by the location of the external

antenna but is limited by the minimum strength required by

the ,anpack receiver. The combined antenna and preamplifier

unit is 36 inches long, and it weighs 1.9 pounds.

64

i= 4

a. U,+

(26 --- Os 41 +

00

cm a

ccE

qr inc

G6 04~

.1-L.JOz cuc

Ln04C__en

jr4

CM65

- ~ - -~-.-.--

TABLE H. MANPACK Interface Wiring Code

Wire Color OUT IN Wire Color

Code BIT BIT Code

90 1 1 26 1 0

91 2 2 27 2 1

92 3 3 28 3 2

93 4 4 29 4 394 5 5 30 5 4

95 6 6 31 6 596 7 7 32 7 6

97 8 8 33 8 7934 9 9 34 9 912935 10 10 35 10 913

936 11 11 36 11 914937 12 12 37 12 915945 13 13 38 13 923946 14 14 39 14 924

947 15 15 40 15 925948 16 16 41 16 926

98 PCTL 17 42 PFLG a927 CTLO 18 43 PSTS 908

928 CTLI 19 44 STIO 916

901 1 BIT 1/0 20 45 STII 917

902 PRESET 21 46 EIR 91822 4723 4824 49

GNO 25 50 GNO

6o

The power converter must supply 24 volts DC to the

Manpack receiver. During this research, a power supply

which converted 120 volts AC to 24 volts DC was used. This

power supply was considered more stable than the power

converter provided with the MVUE, and the voltage output and

power were controlled and monitored during the testing

period.

C. DIFFERENTIAL GPS

As mentioned earlier in this chapter, atmospheric

propagation delays are the major source of error in the

navigational equations because ionospheric propagation is

poorly understood. The tropospheric propagation delays are

neglected in the navigation equations, and the resultant

error is dependent on local tencerature, pressure, humidity,

and MAVSTAR elevation and azimuth. Tropospheric corrections

are usually neglected if the observed satellite is more than

50 above the local horizon. (Ref. 35]

Additional error sources are receiver and satellite

clock perturbations, lengths of signal paths from different

satellites, multipathing delays, satellite ephemeris errors,

and receiver noise. Combined, these factors lead to errors

of 3.6 to 6.3 meters, as table III illustrates. CRef. 36]

67

TABLE III.RANGE ERROR BUDGET

Uncorrected Error Source User Equivalent Range Error,

feet meters

SV Clock Errors *

Ephemeris Errors 5.0 1.5

Atmospheric Delays 8.0 - 17.0 2.4 - 5.2

Group Delay (SV Equipment) 3.3 1 .0

Multipath 4.0 - 9.0 1.2 - 2.7

Receiver "laise and Resolution 0 0 1.5Vehicle Dynamics

Root Sum Square 11.8 - 20.7 3.6 - 6.3

*Two hours after update

58

One method used to reduce the magnitude of this error is

that cf differential, or translocated, positioning. To use

this method, one receiver is positioned over a known

geographic position. Pseudo-ranges to each satellite are

determined, and a differential correction is obtained for

each pseudo-range. These corrections are then applied to

the satellite pseudo-range observed at a second mobile

receiver. With the differential corrections applied, the

net error of the mobile GPS receiver is significantly

reduced. (Ref. 37]

I

69

iv. gz&1

A. POSITIONING NETWORK

One of the considerations in planning this test was to

examine the various methods of positioning the shore-based

equipment. It was desirable to use closely spaced (1-2 kin)

locations within the immediate surroundings of Monterey

harbor for reasons of ready access by personnel during

installation of equipment and for optimum accuracy of the

positioning systems chosen for control. As mentioned

previously, the positioning systems chosen for the

comparison (the three theodolites and the ARTEMIS) give best

results at relatively short ranges.

Two predetermined station marks were selected for the

test--I) USE LON 1978, a bronze disk set by the Army Corps

of Engineers, located on a sand dune on the Naval

Postgraduate School property 400 meters from the ocean at an

elevation of 40 meters above sea level and 2) MUSSEL 1932, a

bronze disk set by the Coast and Geodetic Survey in a rock

outcrop 7 meters above sea level at the shoreline on the

property of the Hopkins Marine Laboratory of Stanford

University. The third station mark was set in the center of

70

A

the Monterey Coast Guard Pier for purposes of the test.

This was a DMA bronze disk stamped GEOCEIVER ST& 31370.

This location was chosen for access to power facilities

available at the pier and because of central location to the

test area.

The position of the Geoceiver station was established

using a Magnavox Geoceiver (Model AN/PRR 14) obtained from

the Department of Satellite Geophysics, Satellite Tracking

Branch, DNA/Hydrographic Topographic Center. The equipment

was operated in accordance with DMA requirements to obtain a

documented first order station. The data were submitted to,

and processed by, DNA/HTC in Brookmont, Maryland, using

precise ephemeris information to provide the WGS-72

position. Michael Ellett of DMA provided the technical

expertise for acquisition of the data and maintenance of the

geoceiver during the test period.

A third order, Class I, closed traverse was conducted

following the standards of the National Geodetic Survey

(NGS), using a Wild T-2 theodolite to measure angles and a

Motorola MRA V Tellurometer to measure distances. The

TRANSIT position computed by DMA/ETC was used for the

initial position of the survey, and a Polaris observation

71

using the theodolite provided the initial azimuth (fig.

1.1.). A total closure error in distance of 0.061 meters

and angle of 5.4" was achieved. The positions determined

from the traverse and used for the control network are as

follows:

USE RON 36 36'08.424"N 1210 52'140.301"W

MUSSEL 360 37'21.882"N 1210 54' 16.048"V

GEOC-IVER STA 31370 360 36136.246"N 1210 53' 29.693"W

B. POSITIONING EQUIPMENT

As previously described, three positioning systems were

used during the test-the three theodolites, the GPS, and

the ARTEMIS. In addition, the geoceiver was set to acquire

data during the shipboard portion of the test. During the

time of day that the GPS satellites were available (Table

IT) only one or two observations per day of the same

TRINSIT satellite were possible. The geoceiver data were

recorded but were not used for this study.

72

14 s n

.3 4 -

34

.3. .4 I,

3 . a

.3

33 3. a

a P.* a RI'6a.. u 2

.3 I6

a 11 s

Fi X .. rstAe ihTrvreSec

0VrrRK IER.SA 77

W 4-

LUJ 0

cc - ~ r CD 0 1 ON ItJ 4

(v 0 J LoC. C .. Cli 04 - CD 0CS 0 l m% (7N (% %

Cc % uL) 1.- ON Lii

-J ~ 4-) 4W'W U) L0 Ln Ul) Lo

(A 4-

cu 0

cm~ 4-

ui~U CD-i

m Lo M- - ) D C a (

_j La. ) 0 1* P- r* % Z>4- .ea 0 Cl ) m D

Lii cu a

00 J e -Lm 0 i0 0

Lii rn Lo

coo >Lii 0 - u CJ CD

74

1. Thoitll~

This section describes the placement and operation

of the equipment during the test. One theodolite was set on

a tripod at each of the three stations, and each was

occupied by an observer and a recorder. Communication

between stations was maintained via portable radios.

Because tests could be conducted only between 10 PH

and 2 AM local mean time (LET) due to GPS availability

times, it was necessary to plan a systems of lights for the

stations. The three theodolite sites were made visible by

placing battery powered lights directly over the station

marks. In addition, battery operated light packs for the

theodolites were used. All shorebased lights and light

packs for the test were provided by the Pacific Marine

Center, NOAA.

The ship system was designed with an omnidirectional

pair of lights anchored around the support pipe for the GPS

antenna. Also, the equipment shelter on the boat deck of

the ship was lighted by spotlights so that the GPS antenna

was easily seen by the shorebased theodolite operators.

During the test, each theodolite observer initialled

on one of the other stations and recorded the reading on a

log sheet. The theodolite operators then intersected the

75

lights on the ship GPS antenna using the theodolite

telescope and tracked the ship until a 'mark' was called

over the radio from the personnel aboard the ship. The

angle measured at this time was recorded on the log sheet as

a fix. The time interval between fixes was approximatoly

one minute. Each series of approximately twenty fixes

constituted a set. At the end of each set, a final reading

was made of the angle to the reference station. The fixes

were numbered consecutively throughout the test, and the

Greenwich Mean Time (GMT) of each fix was recorded on the

ship log sheets. The length of the sets depended on the

ship position and heading.

To correct the observed theodolite angles, the mean

of the final pointing and the initial pointing was

determined for each set and for each station. This value

was assumed to be a constant throughout the set and was

removed from each angle. A collimation error for each

theodclite was also removed from each theodolite

observation. The same observers and theodolites were used

nightly at all stations except &USSEL, where two different

individuals observed over the test week. This continuity of

personnel and equipment was designed to reduce operator and

instrument error to a minimum.

76

2. ARTEM

To colocate the different pieces of equipment at the

GEOCEIVER STA 31370, a wooden tower was built as shown in

figure 4.2. This tower was set over the station for the

period of the test. An YPS equipment vehicle was parked as

close as possible to the tower. Electrical power was

obtained from the Monterey Coast Guard facility.

Prior to the ship test, both GPS antennas were

placed on the tower and centered over the station mark. For

the first two nights, 9 and 10 Hay, equipment problems

allowed the use of only one system each night and data were

acquired only to determine the stability of each system.

During the next night, 11 May, position outputs for the two

simultaneously operating GPS receivers were recorded. This

static test gave a determination of any difference between

the two receivers under almost identical conditions.

For the ship test, the GPS antenna was centered over

the mark and was anchored at the top of the tower; the

geoceiver antenna was centered over the mark and anchored at

the middle level; and the theodolite on the tripod was set

over the mark at ground level. This arrangement provided

colocation of all three systems. It was assumed that no

77

cc0 a

C3LCL

L~Ln

IV-

ccwLAa-

39 u3

:0. m0

LaJZJ

780

error was introduced by the presence of the wooden

structure. The shorebased ARTEMIS antenna unit was offset

from the theodolite to eliminate interference in the signal

due to the tower. An offset position was computed relative

to the station mark and was used in postprocessing of the

ARTEMIS data.

Except for the antennas for the GPS, the ARTEMIS,

and the geoceiver systems, all other equipment was installed

in the equipment vehicle, where an operator was stationed

throughout the test to monitor the data quality.

The GPS Manpack is a manually operated system. As

mentioned previously, it was decided that manual operation

of GPS was detrimental to the goal of the test, and that an

interface that would provide an automated printout of

positions from the anpack as well as a printout of GPS time

would be desirable. Knute Serstis and Gary Whitsell of the

Office of Testing and Evaluation of Systems (OTES), NOAA,

assembled two interfaces that provided data storage on

cassette tapes as well as the desired printout. Each system

operated in conjunction with a HP-9825 computer and a

HP-7845 printer/plotter. Software was provided by OTES,

NOAh. This arrangement permitted storage of the satellite

79

messages for further processing and provided an automated

printout of the ARTEMIS data. A GPS Manpack had been sent

to OTES for testing during the construction of the

interface. Technical expertise and operation of the

interface during the test were provided by OTES.

The data from the GPS receiver were printed at an

interval of 214 seconds, and consecutive data were

continually recorded on tape and printed during the test

periods. In addition, a manual recorl of the GPS position

from the CDU was logged as the 'marks' were called via the

radios.

The ship positioning systems were installed as shown

in figures 4.3. and 4.4. An equipment shelter was secured

to the boat deck of the R/V ACAIIA. A length of six-inch

diameter polyvinyl chloride (PVC) pipe was attached to the

aft wall of the building on the centerline of the ship. The

GPS antenna and the system of lights previously mentioned

were set on top of the pipe.

The ARTENIS mobile antenna unit was set on a

platform located on the centerline of the ship and at the

top of the building high enough to not interfere with the

signal from the ship radar beam. For the period of the

80

oC i00

03 0

C0 a

0L

a -

____ ___ ___ ____ ___ ___ ____ ___ ___ 9.205 m

4 __ ____ ___ ____ ____ ___ 7.935 m

_________6.6m

I r~y 11 71 r-3.6m

00 0

Mean HigherHigh WaterLevel

Fig. 4.4. Equipment Installationon the R/'1 ACANIA

82

test, the weather conditions allowed ship operations without

use of the radar, thus there was no possibility of damage to

the ARTEMIS unit.

All other equipment--the ARTEMIS receiver unit, the

GPS receiver, control data unit, power filter, and

interface, the HP-9825, the HP-7845 plotter/ printer, and

the real time clock--was placed inside the building. An

operator monitored the equipment during the ship tests and

another individual radioed the 'mark', or fix, to the

theodolite operators on shore. The equipment used at the

shore location was installed in a vehicle and consisted of

the same components as on the ship. Operation of equipment

duri.; the ship tests was monitored by an individual located

in the vehicle.

C. LOG OF EVENTS

The test was divided into two portions--i) a shorebased

test which consisted of two nights of stability testing, and

one night of receiver comparison tests, and 2) the ship test

consisting of three nights of shipboard operation. The R/V

ACANIA from the Naval Postgraduate School was the platform

used for the shipboard tests.

83

1. Stablit

Operations began on 9 May, 1981, GMT. One of the

GPS systems was placed over the station mark, and continuous

data was recorded during the satellite availability periods.

In this manner, system 1 was tested on 9 May, and system 2

was tested on 10 May. A position was defined as a quality

observation if four or more satellites were used by the

receiver to obtain the output.

2. Static Tes

:1 Both GPS antennas were centered over GEOCEIVER STA

31370, and data from simultaneous operation of both

4 receivers were recorded at the equipment vehicle. One

" antenna was placed on top of the wooden structure and the

* other immediately below it on the middle level. The

' vertical displacement was assumed to cause negligible error

in the horizontal positions. Data ware acquired on 11 May

frcm both receivers with a quality observation defined as a

position determined from four or more satellites.

3. Z=

For the entire period of shipboard testing, the wind

speeds were less than 5 knots and the seas were calm with 1-

2 feet or less of swell. Visibility was excellent throughout

84

the test week. The ship track for each set is shown in

figure 4.5. A log of subsequent operations follows:

12 May---The GPS Manpack receiver was installed on the R/V

ACANIA and jPS and theodolite operations were conducted.

The ship was allowed to drift during sets 1-3 (fix 1-69)

with 23 fixes per set, in a east-northeasterly direction.

13 May---The ship cruised at 4 knots on track lines from

south to north for sets 4 and 6, and lines from north to

south for sets 5 and 7. During fixes 139 through 157 the

radio at the Coast Guard pier was disabled, and no data were

acquired by either the theodolite observer or the operator

in the equipment vehicle.

14 May---The shipboard GPS system was inoperable. The

trouble was traced to a faulty antenna cable, requiring

replacement. No data were acquired.

15 May---The ARTESIS, the GPS and the theodolite network

were all used for this test period. Data from sets 9

through 13 were acquired (fixes 165-262)

16 May---The control data unit on the shore based GPS system

failed before the start of ship operation and could not be

repaired without factory maintenance.

85

LO

rn0

5 LLA

-C)b' IL cu

I (.))__ W 0

-j~

LO LC)-.-

4- -J 4w W

8b L

I total of 252 fixes were recorded by the theodolite

network and all GPS positions corresponding to the

theodolite fixes were manually logged. The ARTZMIS azimuth

and distance information and consecutive GPS positions on

the ship and in the vehicle were stored on cassette tapes

and printed by the RP-7845 printer/plotter.

87

V. a&ZRA METHOD

Processing of the thesis data was divided into four

interrelated, but separate, tasks. These tasks were the

basic geodetic survey, theodolite and ARTEMIS ship position

determination, differential GPS ship position determination,

and the statistical analysis of the ship position

comparisons. Each task will be discussed separately.

A. BASIC SURVEY COMPUTATIONS

As described earlier in the Test Procedures, a basic

geodetic survey of the test area was conducted to place the

theodolite control stations on the world Geodetic System of

1972, or VGS-72. To begin the survey, a Magnavox AN/PR 14

Geoceiver was placed over the center survey disk, GEOCEIVER

STA 31370 1981, and the digital Doppler data, from TRANSIT

satellite 68, were recorded onto paper tape. Thirty-three

passes, over a period of two and one half weeks, were

recorded, but analysis of the equipment maintenance

diagnostics indicated a possible error in the records.

These data were therefore discarded.

A second set of forty satellite passes of TRANSIT

satellite 68 were recorded over a ten day period and were

determined to be of acceptable quality according to the

88

standards of the DNA Department of Geodesy. This second set

of data was then submitted to DR&, Department of Geodesy,

Computation Section, for analysis. Also submitted were

meteorological data taken during the passes and additional

information required for documentation of a first order

TRANSIT position. Satellite 68 Doppler data for the survey

was then processed in conjunction with the tracked satellite

position by DNA computer facilities. A final geodetic

position on the WGS-72 ellipsoid was obtained for GEOCEIVER

STA 31370. Coordinates of the station are listed in the

Test Procedures section. The computed height relative to

the ellipsoidal surface is -33.39 meters. Computation of

the geographic position utilized a minimum of 38 individual

TRANSIT satellite passes recorded for GEOCEIVER STA 31370.

This provided for an accuracy of one part in 105 (first

order accuracy standard) . [Ref. 38]

To orient the positioning network, an initial azimuth

from GEOCEIVER STI 31370 to station MUSSEL 1932 was computed

from two sets of observations of POLARIS. All observations

and computations of the star sighting utilized procedures

outlined in the National Ocean Surve? Photogrammetric

Instructions Number 4, Revision 1, 4/24/73, the .ydrographic

89

Manual, and the Manual of Geodetic Triangulation, SP-247.

[Ref. 39, 403

Having the initial azimuth, distances were measured

* between all three control stations using a Tellarometer aRA

5 microwave distance measuring unit. Twenty observations of

distance were made for each of the three lines measured:

GEOCEIVER STA 31370 to MUSSEL, GROCEIVER STA 31370 to USE

MON, and MUSSEL to USE MON. The average of the distance for

each line was then corrected for instrument error and

meteorological conditions following the instructions

contained in the Operators Manual and the Tellurometer

Manual. [Ref. 41] Resulting slope distances were then

reduced to geodetic distances using the respective

corrections for slope to horizontal distance, reduction of

elevation to sea level, and the chord to arc distance.

These corrections were computed using the WGS 1972

ellipsoidal parameters and the methods outlined in the

Hydrographic Manual, the Manual of Geodetic Triangulation,

and Introduction to Surveying. (Ref. 42] The resulting

geodetic distances meet third-order, class I specifications

for traverse accuracy, 1:20,000. In addition, angles to the

other stations were observed at each control station, giving

90

a total of three sets of angles. Each set was comprised of

four direct and four reverse plate observations with a wild

T-2 theodolite. Each pair of direct and reverse

observations was corrected for initial pointing errors.

Observations and computations were made using the methods

detailed in the Hydrographic Manual, the Manual of Geodetic

Triangulation, and the Manual of Second and Third-Order

Triangulation and Traverse. [Ref. 43] The resulting

closure meets the specifications for traverse accuracy of

three seconds per traverse station or a closure of 6

seconds.

With the geodetic position of GEOCEITER STA 31370 known,

as well as the azimuth from GEOCEIVER STA 31370 to MUSSEL

and the distance from GEOCEIVER STA 31370 to MUSSEL, the

geographic position of MUSSEL was computed on the WGS 1972

ellipsoid using the direct geodetic computation. This

computation used the Rainsford's method modified by T.

Vincenty and including the Helzert's elliptical terms.

[Ref. 44]

Next, using the azimuth from MUSSEL to USE 30N, the

geodetic distance, and the results of the direct geodetic

computations, the geographic position of USE MON was

91!A

determined. Also, the ARTENIS geographic position was

calculated as an offset using an azimuth and distance from

GEOCEIVER STA 31370. With these computations, the geodetic

survey was completed.

Positioning of the ship, f/V ACANIA, required one

azimuth, or line- of-position, from each of the control

stations MUSSEL, GEOCEIVER SZA 31370, and USE MON. These

lines of position were used to triangulate the vessel

position. Angles from each of the three theodolite

locations were observed as described in the Test Procedures.

Each of the three angular measurements were then corrected

for mean pointing errors and the individual theodolite

collimation errors. Each corrected angle was then adjusted,

by control station, to obtain an azimuth from south, as

required to triangulate the vessel position. The three

resulting azimuths and the corresponding control station

latitude and longitude were then used in a least squares

three line of position intersection computation. Results of

the computation were a vessel position in latitude and

longitude, error ellipse semi-major radius, error ellipse

semi-minor radius, azimuth from south of semi-major radius,

and circular standard error. (Ref. 45] Figure 5.1

illustrates these error ellipse parameters.

92

THE ERROR ELLIPSE

The mean displacement of a point to be expected in the direction ofthe two coordinate axes, X and Y, due to measuring errors, can be ex-

* pressed by the quantitiesvrx and oy.

When the north-south axes are rotated, two axes (X and Y) perpendicularto one another are obtained, corresponding to the maximum and minimumvalues of u and a- The maximum and minimum values of w, and theclockwise rofation angle (cK) of the major axes of the error ellipse areincluded as part of the output of the Geodetic Package and can be usedas input if desired.

The circular standard error (CSE) is also included in the output with theerror ellipse parameters. The value of CES is the radius of a circlewhose area is equal to the area of the error ellipse. That is:

max min

EXAMPLE: Error Ellipse

Az-Max 144.90

Max 0.235 mMin 0.211 m

CSE 0.223 m

S

Fiq. 5.1.

93

For this study, 262 vessel positions, or fixes, were

observed and recorded. On examination during processing, 51

fixes were found to contain errors and data omissions which

rendered the vessel positioning computation invalid or

indeterminate. The remaining 211 fixes comprised the data

set which was then broken down into smaller sets for

analysis and comparison. These sets were numbered 1 through

13 and categorized according to vessel track. All sets were

then processed using several different methods to determine

the most accurate approach.

B. THEODOLITE AND ARTEMIS SHIP POSITION DETERMINATION

Before processing the GPS Manpack data, the ARTEMIS data

were processed to ensure the vessel positions, as obtained

by theodolite, were correlated with the correct automated

and hand recorded GPS data. ARTEMIS data consisted of an

angle, SUSSEL-ARTEMIS OFFSET-vessel, and distance from the

ARTEHIS OFFSET to the vessel. The angular measurement was

first converted to an azimuth from south usins the azimuth

of ARTEMIS OFFSET to MUSSEL as the correction and adding the

measured angle. Vessel position based on ARTEMIS was then

computed using the direct geodetic program (also used for

the basic survey). Comparisons of the resulting geographic

94

positions and the theodolite derived geographic positions,

were made for corresponding fix numbers using the inverse

geodetic computation on the UGS-72 ellipsoid. The inverse

computation utilizes two geographic positions as inputs for

calculating the forward and back azimuths and the distance

between the positions. [Ref. 46] For comparison, if the

two positions differed by more than 20 meters, the ARTEMIS

positions on either side of the suspect ARTEMIS fix were

then used to compute the inverse computation to the

theodolite ship position. In seven fixes, the fix position

values were changed in the May 15 GPS data to give more

accurate results.I

C. DIFFERENTIAL GPS SHIP POSITION DETERMINATION

With the check and adjustment of the GPS data and the

theodolite determined vessel position complete, the first

GPS data processing was the comparison of the uncorrected,

non-differential ship GPS Manpack receiver data to the

theodolite computed ship position. Using the inverse

*geodetic computation, the forward azimuth and the distance

between the theodolite computed ship position and the

ncn-differential GPS ship position were completed. These

results were then used in the analysis of the differential

495

AD-AIIh 511 NAVAL POSTGADUATE SCHOOL MONTEREY CA F/S 1?7"APPLICATION OF THE GLOBAL POSITIONING SYSTEM TO MEARSHORE MYRO--ETC(U)SEP 81 V E NEWELL, D D WINTER

II~ll lllllll

23Iflllllllllll

LU 111112.5........ 11111=

11111 5 B I~ iiII.

MICROCOPY RESOLUTION TEST CHART

processing to examine the stability of the shipboard Nanpack

receiver compared to the GPS Manpack receiver located over

GEOCEIVER STk 31370.

There were two methods of analysis of the differential

GPS. The first of these was the determination of a static

correction to be applied to all the GPS data. This static

correction in the differential GPS computations was to

account for a suspected bias between the two similar, but

not identical, Manpack receivers. A mean static correction

was computed for the most stable reception period during the

static tests on 11 May. With the two receivers colocated

over the GEOCEIVER STA 31370, a mean correction of -0.954

seconds of latitude and -0.3 07 seconds of longitude was

found between GPS Manpack System 1 and System 2 during times

07:16:00 GMT and 07:25:12 GMT. This static correction was

then added to the latitude and longitude of GPS Manpack

System 1 (located over GEOCEIVER STA 31370) data.

The second method of analysis or processing in the

differential mode used the latitude and longitude obtained

directly from the GPS Man pack receivers. In either

processing method, the computation of the differential

corrector to the shipboard 3PS Manpack receiver data was the

96

same. An explanation of the computations and processing of

the differential correction for both methods follows.

Computation of the differential corrector for each GPS

fix position was accomplished by using the geographic

position obtained with the stationary Hanpack receiver,

located over GEOCEIVER STA 31370, and the geographic

7 position of GEOCEIVER STA 31370. These two positions were

then used in the inverse geodetic computation to obtain a

differential correction vector (forward azimuth, and

distance) from the stationary GPS reqeiver to GEOCEIVER STA

31370. In figure 5.2, the differential correction vector is

labelled "A". Using the direct geodetic computation, the

differential correction vector was applied to the

corresponding shipboard GPS Manpack receiver position, and

the differential GPS position was obtained. These .sulting

positions were computed on i one-to-one basis using the fiz-

numbers as identifiers. Vector "B", the differential

corrector vector which is identical to the vector "A" in

figure 5.2, is applied to the shipboard GPS Manpack receiver

and results in the differential GPS ship position.

Comparison of the differential GPS ship position with

the theodolite-computed ship position was accomplished using

97

T-2 COMPUTEDSHIP POSITION

DIFFERENTIAL GPS SHIP

POSITION:! GPS SHIP

DIFFERENTIAL POSITION

CORRECTOR VECTORAPPLIED.

GEOCEIVER STA 31370

DIFFERENTIAL STATIONARY GPS POSITION

CORRECTOR VECTORCOMPUTED, A

Fig. 5.2.DIFFERENTIAL GPS PROCESSING

98

EIthe inverse geodetic computation for both methods of

differential GPS processing. The resulting forward azimuth

and distances for both processing methods were used in the

statistical analysis.

D. STATISTICAL DATA ANALYSIS

Statistical analysis was done on data separated into

sets according to the method of data processing. The static

data and stability data were graphed as functions of

latitude and longitude versus time. User range error,

furnished by the Space Division, GPS/OL-AQ, Vandenburg, AFB,

was plotted as a function of time for each day of the test

(see Results). Other graphed relationships are the

gecgraphic positions of the stationary GPS Manpack receiver

data versus the GEOCEIVER STA 31370 and the theodolite

semi-major radius as a function of fix number (see Results).

Other statistical data analyses were represented by

histograms. Plotted on these histograms are the relative

frequency on the left margin, variable bar increments scaled

on the bottom, number of variable observations in each bar

increment on the top, and the total sample size on the top

margin. Within the histogram the frequency bars are denoted

by asterisks, the mean by a capital M, the quartiles by

99

periods and the empirical density function by the capital

curve. The empirical density function was computed using:

- U'(Z)14(3I1)) i=1 W 314

CRef. 47] where: ?(z) = the empirical density function, z

the frequency of observation, x = individual observation,

N = the total number of observations, V = weight function, B

bandwidth,

1=17 -z

B(N) = range of all observations / 4W

The statistf.cal values listed below the histogram are

defined in Table V. Computed histograms show the distances

between: 1) the theodolite-computed ship position and the

differential GPS ship position, both with and without the

static correction, by set; 2| the theodolite-computed ship

position and the differential GPS ship position, without

static correction, for five satellite observation periods on

each of the three ship test days; 3) the theodolite-computed

ship position and the non-differential GPS ship position for

the five satellite-observation periods on each of the three

100

TABLE V. STATISTICAL DEFINITIONS

The definitions of the basic statistics computed by the program are

listed below:.

MEAN: Average of the sample

- MEDIAN: Mid-Value of the sample, if there are an odd number ofsample points, or the average of the two middle valuesfor an even number of points

TRIMEAN: 0.25 * (Q1 + 2Q2 + Q3), where Q are the quartiles

MIDMEAN: The average of all sample values lying between the upperand lower quartiles

MIDRANGE: Average of the maximum and minimum

GEOMETRIC MEAN: GM nx * x

HARMONIC MEAN: HM = n /x

VARIANCE: Unbiased estimators for variance and standard deviationare used - the square of the standard deviationIn 2 -2

STANDARD DEVIATION: S = 'x - n x-

COEFFICIENT OF VARIATION: Standard deviation / IMeani when themean is less than 10. ** -30, the coeffic-ient of variation is set to zero.

MEAN DEVIATION: The average of the sum of the absolute differencesbetween the sample values and the median

RANGE: Maximum - minimum

MIDSPREAD: Q3" Ql' also called the interquartile distance

M3: Third central moment. Unbiased estimator is used

144: Fourth central moment. Unbiased estimator is used

COEFFICIENT OF SKEWNESS: M3 / (Standard Deviation) ** 3

COEFFICIENT OF KURTOSIS: (M4 / (Standard Deviation) ** 4) - 3

BETAl: Biased estimate of third central moment; can be used intesting for normality

BETA2: Biased estimate of fourth central moment

MAXIMUM: Largest sample value

MINIMUM: Smallest sample value

QUANTILES: The alpha-quantile, x(alpha), is the solution to theequation:

Probability (x .le. x(alpha) ) alpha

101

ship test days; and 4) the GOCEIVER ST& 31370 position and

the uncorrected stationary GPS position for the three days

of ship ests. In addition, the azimuths between the

theodolite-computed static corrections, for each subset,

were graphed as a series of histograms. To illustrate the

difference between the GEOCzITER STA 31370 position and the

stationary GPS position, the azimuths between the two

positions are graphed on histograms for each of the three

test days.

102

I

The error sources contributed by each positioning system

are described. Estimates of the effect of each on the final

results are determined where possible.

1. ARTEMIS POSITIONING SYSTEM

The sources of error in the ARTEMIS ship position can be

divided into tvo categories--those sources of error inherent

in all microwave positioning systems and those unique to the

AREMIS system. kll are investigated and the contribution

of each to the total error is determined.

1. Mioav Ss erroL=

Sources of system error include range holes,

inaccuracies in calibration, non-ideal geometry of the lines

of position describing the fix, and interference with

reflections of the signal from nearby objects.

Range holes are a cause of unreliability in

microwave positioning systems. The problem exists when the

sea surface is calm enough to reflect part of the

transmitted -signal. The reflected siignal will arrive at the

receiving antenna at a later time than the signal which

travels directly, causing a phase difference between the two

signals. A zone or region of destructive interference of

103

the directly transmitted signal and the reflected signal

will occur within which the vessel will receive an

attenuated signal. When the phase difference of the two

paths is exactly 180 ', no signal will be received provided

the reflected signal is as strong as the direct-path signal.

Prediction of the range holes, therefore, is useful in

survey planning so as to avoid this problem.

The range hole is a function of shore unit and ship

unit antenna heights, the wavelength of the system, and the

distance from the shorebased unit. If the operating area is

known, antenna heights can be determined and equipment can

be installed so that no range holes exist in the survey

area. If fixed antenna heights are known, the region of

uncertainty can be determined from the equation: [Ref. 48]

The following equation is used for this computation:

Dn = 2(hl)(h2)/nX,

where A is the wavelength (constant for AIRTHIS at .032 i),

hl and h2 are antenna heights, and an is the distance from

the shore based unit to the nth range hole. For this study,

the antenna height of the fixed unit above mean sea level at

the GEOC-IVER STI 31370 was 3.8 meters. The mobile antenna

was installed on top of the equipment shelter on the boat

104

deck of the R/V ACANIA at a height of 7.9 meters above the

the waterline. Two range holes within the survey area were

determined to be located at ranges of 1876 meters for n - 1

and 938 meters at n = 2 (fig. 6.1). These two ranges were

computed for the case of mean lower low tide. The change in

width of the range hole due to a tidal fluctuation of 1.5

meters during the test is shown in figure 6.1. (During the

night of 15 May the sea surface was sufficiently disturbed

by a two-foot swell and a surface wind of 2-4 knots such

that surface reflection was minimal and no loss of signal

was observed at the computed ranges).

Calibration of the ARTEMIS system was accomplished

in two stages. As part of the study, the ARTE.MIS was

obtained for a one week test of the system from 13 to 17

April. During this time it was calibrated for azimuth and

range over the baseline between the GEOCEIVER ST& 31370 and

MOSSEL. The azimuth readings required no adjustment but the

range was reduced by two meters. The equipment was returned

to Harinav Corporation to fulfill prior commitments and then

was returned to Monterey for the ship test on 15 May.

Before the evening operation, the ARTEMIS was used to

measure the same baseline as previously. A plus two meter

105

UU,C~i V)

L60

L6..

r % LL.

C..'J.

0 vi

106

bias was observed at this time, but no immediate

compensation was made. This error was assumed constant

throughout the test and was included in. the calculation of

,he ship-determined ARTEMIS position lurin4. postprocessing.

One of the advantages of the ARTEMIS system, unlike

other microwave positioning systems, is that the geometry is

always ideal for the lines of position that determine the

fix. The range-azimuth technique of determining the vessel

position defines an uncertainty ellipse having axes that are

always mutually perpendicular. This allows prediction of

errors as a function of range only, independent of the

angle.

The manufacturer's specified error in range

measurement of the ARTEMIS is + 1.5 meters for all line of

sight ranges. One axis of the uncertainty ellipse is

therefore constant. The other axis, that describes the

error introduced by the system azimuth uncertainty, is 60 cm

per 1000 m range. (This is determined from an angular error

of 0.0330 or 2' of arc). Both system error values represent

an accuracy of twice the standard deviation value, or a 95%

probability of the true value occurring within the margin

allowed by these parameters.

107

The positional uncertainty can be expressed In terms

of the radius of a circle of equivalent probability within

which the true fix is expected to occur. The method of

computation of this radius is described in Appendix A. The

technique allows the conversion of uncertainty ellipses to

radii of circles of equivalent probability. Values from

this computation for the radii of circles of 90% probability

at different ranges used in the survey area for 15 May are

displayed in figure 6.2 along with the circles of 90%

probability, the magnitude of the radii at the ranges used

for the study do not exceed 5.4 meters. One goal of this

study was to measure the difference in positional accuracy

between ARTEMIS and GPS. A maximum error of 1.0 meters, an

order of magnitude smaller that the tested system accuracy,

is desirable. At the present time, however, there are few

systems that can provide the order of accuracy system for

the position of a moving vessel, so the 5.4 meter error

observed at 4000 meters had to suffice.

Error caused by reflection of the signal from nearby

objects is difficult to measure. To minimize this problem.

the offset position for the ARTEMIS unit was planned so as

to be clear of possible interference from the wooden

108

0 CCM

V I-

Cj

0V

109

structure and the eluipsent vehicle. No signal interference

was observed despite the ship's notion. This source of data

unreliability was therefore not considered to be

significant.

2. AREI System Errors

Errors unique to the ARTEMIS system include a

pointing error source and the lack of resolution of the

azimuth measurement in the dynamic mode of operation.

The operation of the ARTEMIS system requires a

manual alignment of the fixed antenna so that it is

perpendicular to a known object. This is necessary to

establish an original reference direction upon which all

angular measurements are based. It is possible that a

pointing error couli be produced by the operator if the

crosshairs of the telescope used to align the antenna are

not exactly sighted. This would introduce a systematic

error into all subsequent angle measurements. To minimize

this source of error, the alignment of the antenna was

checked before and after the period of operation. The

original azimuth displayed on the antenna unit display

agreed exactly with the final azimuth value. Any

contribution to ship position error was considered minimal.

11.0

The ARTEMIS was operated using the dynamic mode

option (see Equipment Description--kRTEMIS). The resolution

of the system in this mode is 1 meter for displayed range

and 0.10 for azimuth values. This resolution limit

introduces an uncertainty of +0.5 meters in range and 0.05.

in angle measurements. The computed uncertainty ellipse as

expressed by radii of circles of equivalent probability, is

shown in figure 6.2. The system error and all other errors

mentioned above are assumed to occur within the scope of

this resolution uncertainty.

Another source of system error has been proposed in

a study conducted in June, 1980, by NOAA. (Ref. 49] The

test compared the ARTEMIS system with tvo ranges measured by

two Del Morte Trisponders (a short range microwave

range/range positioning system) and two lines of direction

measured by two shore-based theodolites. The results

described an underway bias of -0.050 of unknown origin in

the angle measurements. This discrepancy has not been

resolved. The preliminary test conducted in Monterey in

April, 1981 compared the ARTEMIS system with three lines of

direction determined by three theodolites. The purpose of

the test was to resolve the underway bias problem. Results

are enclosed in a separate report. [Ref. 50]

111

B. CONTROL NETWORK

The entire survey was oriented relative to the TRANSIT

position on the Coast Guard Pier (GEOCEIVER ST& 31370).

Therefore, errors inherent ia the position of this station

mark affected all results. Assuming that the geographic

coordinates published by D[ A/RTC represented the best

estimate available of the GEOCEIVER STA 31370, the remaining

significant sources of error affecting all results were

survey errors. The control network for the study was

* idetermined from the traverse and it was assumed that all

positions used in the study were oriented relative to each

other on the UGS-72 ellipsoid. A change in the ellipsoidal

surface within the area encompassed by the traverse was

assumed to be insignificant.

Survey errors stem from the methods used for obtaining

the initial azimuth for orientation of the survey, the type

of traverse chosen, and the geometry of the traversed

triangle.

To obtain a starting azimuth for the traverse, two sets

of observations on Polaris were recorded at the GEOCEIVER

STA 31370. A striding level attachment to the theodolite

was not used. The instrument level bubble was observed

112

repeatedly during the observations to ensure constant

instrument alignment with the vertical. Failure to keep the

instrument level could have introduced a + 30" uncertainty

into the starting azimuth of the survey. (Ref. 51] To

consider the significance of such an errt-', all stations

were recomputed using a deviation of + 15" from the computed

azimuth. The resulting shift in the positions of MUSSEL and

US! SON are listed:

USE SON 0.133 m distance at 230036847.1611 azimuth

IUSSEL 0.132 m distance at 230033, 4.08" azimuth

The change is considered insignificant relative to the other

errors described above.

The survey was a closed loop traverse using GEOCEIVER

STA 31370 as the origin and closing position. The weakness

in this method involves the measured distances. m ethod of

checking the observed angles in the triangle exists via the

angular closure if a significant deviation in one of the

angles can be detected. The distance measurement, however,

can produce an undetected bias in the survey if the distance

measuring instrument contains a systematic error. This

error would shift all the surveyed positions. Correct

113

A

measurement of distance by the dRA V Tellurometer,

therefore, was important. The equipment was calibrated over

a known baseline and the errors were within the centimeter

range. This error was therefore negligible.

- IThe geometry of the triangle was not optimum. Ideally,

the triangle would have been equilateral to minimize the

source of observing error. The choice of traversed points

was based on the location of preexisting station marks and

the desired geometry of the theodolite fixes to the ship.

The closures computed for the traverse met third order

specifications for accuracy, and the survey was not adjusted

to remove the closure error. (Ref. 52, 53J

The survey errors were considered to be within the

uncertainty of the TRANSIT station location and were

therefore not considered significant for the study.

C. THEODOLITE NETWORK

The errors generated by the theodolite network are

divided into equipment errors, operator related errors and

system errors, of which the litter is the most significant.

The equipment used for the theodolite network were three

1-second theodolites which have a measurement resolution of

+0.5". All theodolites were in good condition.

114

Instrumental errors of a theodolite are listed in rable VI.

[Ref. 54]

The major source of equipment error was introduced by

not changing face, the process of observing angles a second

time with the telescope reversed. Due to the nature of the

test coupled with 1-minute fix intervals, it vas not

possible to observe all the fixes in both the direct and

reversed modes. ks an example of the magnitude of this

error, the horizontal collimation errors were determined for

' e each theodolite prior to the study. The results are listed

below:

Zheodoliti CliaonError

# 14405 - 1.0"

# 14482 + 2.7"

# 14452 - 0.7"

All the measured angles were adjusted for this error as

described in the processing section. The largest influence

on the remaining instrumental errors was the non-verticality

of the main axis. If the non-verticality were large (for

example, one division off-level in the lower spirit level

bubbles, the angle measured could be in error by 7".

CRef. 55] To minimize this error, each theodolite was

115

TABLE VI. Theodolite Instrument Errors

ERROR ELIMINATED BY:

Eccentricity, misplaced indices Reading both sides of circleand changing face

Graduation of circles Changing zeros

Micrometer run and vernier error Selection of suitable zerosetting

Non-verticality of main axis Error is not can.r-*4 ,. lAny

observing proced..r.

Bad setting of the upper spirit Correct setti;t): rvi -

level tion of correct ic,

Verticality of cross wires Changing faca . q samepart of wire

Horizontal collimation Changing face

Vertical collimation and index Changing facesetting error

116

relevelled if necessary between sets of observations. The

magnitude of this and the collimation errors were considered

to be insignificant relative to the system errors discussed

later in this section.

The operator-related errors are a function of the

individual performance of the observers. The experience and

personal comfort of the observer, for example, are two major

factors which can influence the observed angles. An

observer's performance and resultant impact on the quality

of the results depend on the judgment and reaction time of

the observer and are influenced by variables such as fatigue

and age. The subjective nature of the problem makes an

estimate of the error difficult, therefore, for this study

the error was considered random.

The major source of error in the theodolite positioning

system dealt with the timing of the fix. A delay existed

between the time the 'mark' was called over the radio and

the actual time the angle was observed. The timing problem

was most critical when the ship was close to one of the

three observer locations because the rate of change of the

bearing of the ship was greatest at this time. Hence a

small delay generated a large amount of error in the angle

117

recorded. To indicate the effect this error source may have

had on the position of the ship, two fixes were chosen

arbitrarily--one at close range to each of MUSSEL and

GEOCEIVER STk 31370. A one minute value was added to the

angle which was most affected by the timing error and the

fixes were recomputed. The same process was repeated for

the three fixes with the one minute value subtracted from

the critical angle in each fix. The results follow:

Number Original Fix + 1 minute - 1 Ninute

210 36 0379 20.6228211 360 37120.61468" 36037120.63081"

0* 0121 53#41.97184" 121 53141.96933" 121 53141.97426"

259 360360I45.88048" 360 36'15.87574" 36036 145. 89082"

121 52'54.0506211 121 52054.04635" 121 52'54.05459"

No significant error was generated from this source.

Another source of system error was associated with the

geometry of the fix. Ideally, the three theodolite angles

would describe two equilateral triangles to the ship;

however, within the operating area, the geometry varied. As

an indication of the size of this error, representative

fixes were chosen and the circular map accuracy standards

(CHAS) of each of the combinations was determined. Figure

6.3 illustrates the approximate error caused by variations

1ills

'7I

S.

Ln m

L61.

C~j CA

000

C~C

L613

V) 0

Ila,

in the geometry. Vithin the area outlined in the figure the

geometry creates a CHAS of not more than 10.4 meters. This

is therefore the source of the largest probable error in the

theodolite method of positioning. Nevertheless, this method

in combination with the ARTEMIS system provided the best

estimate of the ship location.

D. TRANSIT POSITIONING SYSTEM

The method used to determine the TRANSIT position

(GEOCEIVER STA 31370) is the most accurate satellite

positioning technique presently available. Error sources

are reduced both during data collection and during

postprocessing of the recorded values.

inA places requirements on the collection of data that

will minimize error. For this study, only satellite passes

whose maximum altitude ranged between 10° and 80 0 were

recorded. 3eteorological data (temperature, pressure, and

humidity) were collected for each pass, and a determination

of the receiver clock drift was computed after each pass.

The amount of clock drift as a function of time gave a

real-time indication of receiver performance. Equipment

checks were conducted routinely, and the results were

included with the data package submitted to DNA.

120

An error budget of the TRANSIT system for point

positioning is listed in Table VII. [Ref. 56] The precise

ephemeris method of postprocessing accounts for all of these

errors: 1) for each pass the tropospheric correction is

computed from the meteorological data; 2) the ionospheric

correction is determined by observing the shift in frequency

between the two transmitted frequencies; 3) the receiver

clock error is known; 4) the satellite orbits are observed

by the tracking stations so that precise satellite positions

are known; 5) errors due to uncertainties in the

geopotential model and surface drag forces are accounted for

during computation; and 6) the receiver altitude relative to

the ellipsoidal surface is computed approximately from the

given HSL elevation.

The accuracy of the computed TRANSIT position is stated

as 1.5 meters per axis within a 90% confidence interval.

This is assumed to provide the best estimate possible of the

GEOCEIVER STA 31370 relative to the WGS-72 coordinata

system.

E. GPS ERROR SOURCES

Global Positioning System error sources are separated

into two main categories, the space vehicle system errors

121

TABLE VII.TRANSIT System Errors

SOURCE ERROR (m)

Uncorrected propagation effects(ionospheric and tropospheric effects) 1-5

Instrumentation and measurement noise(local and satellite oscillator phasejitter, navigator's clock error) 3-6

Uncertainties in the geopotential modelused in generating the orbit 10-20

Incorrectly modelled surface forces(drag and radiation pressure acting onthe satellites during the extrapolationinterval) 10-25

Uncertainties in navigator's altitude(results in a bias in longitude) 10

Ephemeris rounding error 5

122l

and the GPS user receiver errors. There are seven possible

sources of error in the space vehicle system--errors

associated with satellite equipment, satellite clock,

ephemeris, atmospheric delays, signal multipath effect,

satellite update, and satellite propulsion firing. Table

III lists these error sources and the approximate user

equivalent range error for each. [Ref. 571

The satellite equipment error, also termed the group

lelay, is caused by differences and uncertainties in

processing and transmission of the message through the

individual satellite's circuitry. Each satellite has the

group delay calibrated prior to launch and this known group

delay is included in the satellite time corrector within the

navigation message. However, unknown delays in signal

processing and circuitry not accounted for by the corrector

contributes an estimated one meter to the error budget.

Satellite clock error is caused by variations in each

satellite clock. Satellite clock time may depart by up to

976 microseconds from the correct GPS time. Corrections for

the predicted clock variations are included in the

navigation message. Residual clock variations contribute to

the error budget and are included in ephemeris errors since

123

the clock errors are small and have the same effect as

ephemeris errors.

Ephemeris errors are the result of satellite clock

variations, drift by the monitor station clocks, monitor

*0 station signal processing delays, lack of precise monitor

station geodetic positions, solar wind and pressure

parameter variations, and earth gravity model error. These

sources all contribute to the ephemeris error budget. The

ephemeris error budget is monitored over long time periods

and translated by the master Control Station into an

User-Equivalent Range Error (URE). When the URE exceeds the

prescribed value of 4 meters the satellite ephemeris data is

uploaded to reduce the URE. (Ref. 58] Most of the URE is

included into the user clock corrector obtained in the GPS

position computation. However, not all of the ephemeris

error is removed during the position computation, and the

residual error has been estimated to be approximately 1.5

meters. See Table IiI.

Atmospheric delays in signal propagation are the result

of two effects, ionospheric delay and tropospheric delay.

Of these, the ionospheric delay, which depends on the

frequency of the transmitted signal, is greater. With the

124

two frequencies, Li and L2, transmitted by each satellite to

the user receiver, the ionospheric propagation delay can be

modelled. Using this model, the majority of the propagation

delay is accounted for, but a residual delay results in an

error. The other part of the atmospheric delay error

depends on the tropospheric propagation delay. Propagation

delay in the troposphere is a function of signal path length

through the layer, the receiver air temperature, and air

pressure and humidity. Since the tropospheric delay is not

a function of frequency and is highly localized, the

tropospheric delay corrector has not been incorporated into

the atmospheric correction computation and is therefore an

additional source of error. These two error sources combine

to yield an atmospheric delay error estimated to have a

magnitude of between 2.4 to 5.2 meters.

Ionospheric scintillation of the GPS signal, in addition

to the ionospheric propagation delay, cause errors in

positioning. scintillation has the effect of random

interference with the propagated signal as the signal is

transmitted through the upper ionosphere. [Ref. 591

Interference caused by scintillation combines with the GPS

user receiver processing noise to cause a random ranging

125

error and, if the interference is severe, interruption of

the pseudo-random number binary code causes loss of signal.

Errors and signal losses caused by scintillation cannot be

predicted or estimated.

Signal multipath errors result from the reception of a

satellite's signal transmitted via different propagation

paths. Received signals thus may have been distorted or

reduced in strength to a sufficient degree so as to affect

the range determination from the satellite. Error due to

multipathing is highly localized and cannot be modelled.

Estimates of the error magnitude are between 1.2 and 2.7

meters.

The last two space vehicle system error sources are

attributable to the Raster Control Station (MCS) satellite

accuracy and maintenance functions. Usually the MCS uploads

each satellite at a given time each day and the ephemeris

data is changed(Table VIII). Should the user not be aware of

occurrence of upload and if the receiver therefore does not

obtain the new ephemeris data, ranging to the satellite in

question will contain a possibly very large error. In

addition, if the MCS, through the Remote Stations,

determines that the four skewed reaction wheels cannot

126

kn CDO OL 0 0U) (mC 0 0 ULA Layl -' (" o- 0 - q ~

00c~ k 00 00o 0)%0 0 V)0

to ulA O 00 CDL 00C DU. )C LAO0 LA'O L -C 0 C LA D LCDO qwO aA en 4wmA%0~ Q %0- 1 00- %00% 00 0 U

Ln LA 0 0 LA

OX - LA LA LA LA c

LA 00 00 00 L LLA(n CD 0 00 00( D '

LA L L % LA CO LA co mrf0D 00 00 00 D 0 0

0 ~ - LAO L 0 00 DL OLn Ln O Ln ALL LA LALA

ua- D0 m nm a. -( - - -c

--cc

ccKK

127

maintain satellite antenna alignment with earth, the

hydrazine propulsion system is activated and the satellite

is rotated. Once the propulsion system has used, the

satellite orbit cannot stabilize immediately, and the

ephemeris data will be inaccurate. A period of up to

several hours may be required for orbit stabilization and

for the new ephemeris data to be computed and uploaded.

During that period, the ranging computations to the

satellite will have large errors. The magnitude of these

two error sources cannot be estimated but is assumed to be

several orders of magnitude larger than the previously

stated errors.

The second error category is that related to the GPS

user receivers. These sources of error associated with the

user GPS receiver, the fanpack, are receiver signal noise,

receiver vessel velocity, signal resolution, method of

signal processing, ephemeris update capability, and, for the

differential GPS computations, system accuracy comparison.

Received signal noise errors result from interference

introduced into the receiver GPS signals through processing

and internal receiver circuitry. Additionally, quality of

signal resolution is dependent upon receiver type and can be

128

a source of error. The combined errors from these two

sources are estimated to be 1.5 meters. [Ref. 60]

Estimates of these error sources for the TUE have not been

made but are estimated to be of the sane magnitude.

Velocity of the receiver vessel results in GPS errors of

varying magnitude. The higher the velocity, the larger the

error. This error is due to the time lag required to obtain

and process each H&VSTAR satellite range with the VTUE

single channel receiver, to compute the average vessel

velocity and to use the average velocity for positional

computations. The Manpack requires a six second time

interval to obtain and process the pseudo-range from each

satellite used. Thus the fix position is an extrapolation

of the four satellite ranges to obtain a convergent fix

position. at one knot, the vessel travels 0.51 meters in

one second. During the full twenty-four seconds required to

obtain four satellite ranges, the vessel would have

travelled 12.24 meters. With the vessel travelling at four

kncts, the distance travelled during the total time interval

required for a GPS fix position would be 48.96 meters.

Processing the pseudo-ranges from four satellites requires

the solution including the average vessel velocity during

129

the 2:-second ranging period. Use of this average neglects

the relative velocity of the receiver antenna which is

mounted on the mast. During this test, the GPS antenna was

mounted 9.06 meters above the waterline of the R/V ACANIA.

Assuming five degrees of pitch and roll and a vessel roll

period of 15 seconds, relative average antenna velocity was

0.16 meters per second. The antenna velocity error

translates into a random error of 0.16 meters. The other

velocity errors have not been estimated but are assumed to

be less than one meter for the low vessel velocities used in

the test.

One error source mentioned previously requires further

clarification: error due to the relative position at the

time of the fix. The Manpack requires up to twenty-four

seconds to obtain and process the satellite ranges. The

position of the fix and the associated time of fix are then

recorded. However, the time of the fix may not have been

the same time the computational results were printed. Thus

the output position may not have been correlated with the

correct theodolite computed ship position since the

theodolite lines of position were observed when the GPS

position was printed. Errors from this bias source may

130

range from 12.24 meters at one knot of vessel velocity to

48.96 meters at four knots. This error varied with the

processing time required to obtain a convergent position.

An estimate of this error magnitude was not made because of

its wide variability.

The type of receiver and the method of signal processing

also contributed to the error. Multichannel receivers

designed for Phase I testing have a lower magnitude error

for satellite signal reception and processing than do the

single channel receivers, such as the Manpack. In the

multichannel receivers, four or more satellite pseudo-range

data are taken simultaneously. The single channel receiver

sequentially receives each satellite and computes the

pseudo-range data. Simultaneous data reception reduces the

user velocity error and therefore has less position aliasing

than that found in the single channel receivers.

The method of signal processing may induce positional

errors through the use of non-optimal weighting functions in

the position computation. Should the ephemeris or ranging

data change significantly, as happens during uploaoing,

signal processing using Kalman filters as used in the

Manpack do not change the computed position rapidly enough

131

-, I, .. .... ... .. ... .. .. . . .. .. ....... . . .. .. .

to reflect the large changes in ranges and result in a slow

change drift of the GPS position. Estimated time

requirements for user receivers to stabilize after an upload

are up to two hours. [Ref. 1J]

The capability of the user GPS receiver to receive

ephemeris data immediately after upload reduces the

ephemeris errors caused by using an outdated ephemeris. The

ephemeris data is repeated every thirty seconds and its

length requires sections of two message subframes.

[Ref. 62, 633. The use of current ephemeris data in

the computations reduces the error for the ephemeris to the

value shown in table III. The Manpack has the capability to

accept new ephemeris data immediately, but during the test

large errors were introduced into the computation on

satellite ephemeris upload because this selected optional

function either did not operate properly or was neglected by

the operators.

For the differential GPS user, an additional error

source is introduced into the error budget: the assumption

that both GPS receivers will observe, obtain the

pseudo-ranges, and compute the position solution in exactly

the same manner. If this were true then both receivers

132

should be able to compute the identical position if placed

at the same position at the same time. With the Manpacks,

the assumption that the units were identical and the

position solutions were identical was not made because prior

testing established that the receivers differed in position

*i computations by as much as 3 08 meters. [Ref. 64]

For the static testing on 11 May, one error source not

previously discussed was the vertical separation of the two

Manpack antennas, colcated over the GEOCEIVER STA 31370

* control station, and the possible interference of the

uppermost antenna, System 1, with the GPS signals received

by the lower System 2 antenna. Separation of the two

antennas was 2.1 meters. Calculations using the speed of

propagation in a vacuum and the two GPS frequencies produce

wave lengths of 0.19 meters and 0.24 meters. Since

separation of the antennas was much greater than either

wavelength theoretically there should have been no

positional errors due to antenna interaction or

interference. Differing elevations of the antennas above

the ellipsoidal surface also may have induced errors in the

position computation. Estimated errors for the elevation

differences are less than 0.1 meters. Total errors for the

133

antenna vertical elevation separation and antenna

interference or interaction have not been estimated.

Thus, the error budget for the Global Positioning System

suggests a combined error of up to 57.35 meters. Using the

differential technique, the bias errors can be reduced in

magnitude or eliminated. These bias errors include

ionospheric propagation delays, tropospheric propagation

delays, variations from predicted ephemerides, and satellite

clock perturbations. Teasley, Hoover, and Johnson (1980),

evaluated the error budget for a more idvanced GPS receiver

and found the predicted filtered solution error was reduced

from approximately 11 meters to approximately 3 meters with

differential GPS processing. (Ref. 65] Error budgets for

the Nlanpack and the differential correction vector

processing method used in this test did not attain the four

meter accuracy but were not expected to because the

receivers used were less accurate, and of different types.

k134

Results of the GPS tests have been separated into four

sections: stationary GPS evaluation, evaluation of non-

differential GPS, differential GPS evaluation with static

correction, and evaluation of differential GPS without

static correction. Tables TV and VIII list the optimum

satellite observation periods and the, NAVSTAR upload times

for each day of the testing period.

A. STATIONARY GPS EVALUATION

Stationary GPS receiver evaluation was separated into

the stability tests, the static tests and the tests of the

stationary GPS receiver used during differential testing.

Figures 7.1, 7.2, and 7.3 illustrate the results for the

stability test conducted on 9 and 10 May. The receiver was

less stable than expected and depended to a great extent on.

the NAVSTAR uploading times and the URE of NAVSTAR 1 (fig.

7.4 and 7.5). Correlation between times for the NAVST&R

upload and changes in the stability can be observed at 0620

on 9 May (fig. 7.1), 0655 on 9 May (fig. 7.2), 0815 on 10

May (fig. 7.3), and 0845 *n 10 May (fig. 7.3). The

relationship between stability and URE is illustrated by

comparison of the UBE for NAVSTAR 1 at 0640 on 9 May (fig.

135

61

STABILITY TEST RESULTS36 37 15 9/10 MAY 1981

00

Lu~ EA x Axc x

a (10 MAY tIAVSTAR 1)

36 36 30 (10 MAY A

4AVSTAR 3,5,6) (9 MAYNAVSTAR 5,6)

121 53 45

i&ILl IL Xr

30 - , , le x L ji

121 53 15 x 9 MAY - SYSTEM I

lO MAY - SYSTEM 2

( 3 SATELLITE JPOATESSI I

0545 0600 0615 0630 06A15 0700

GREENWICH MEAN TIME

Fig. 7.1.

136

36 37 15STABILITY TEST RESULTS9/10 MAY 1981

*007

-j45

x xE

36 36 30- A 3-

121 53 45-

11-I

121 53 1SF9 MAY - SYSTEM 1

.10 MAY - SYSTEM 2

SATELLITE UPDATES

0645 0700 0715 0730 0745 0800

GREVMIICH MEAN TIME

Fig. 7.2.

1137

36 37 15 STABILITY TEST RESULTS

9/10 MAY 1981

x 9 MAY - SYSTEM 1

00 olO MAY - SYSTEM 2

( ) SATELLITE UPDATES

i--

45 (10 MAY NAVSTAR 1)(9 MAY NAVSTAR 510 MAY NAVSTAR 6)

36 36 30

121 53 45

30 x l AX x xA

0w

121 53 15

l I I I I

0800 0815 0830 0845 0900 0930

GREENWICH MEAN TIME

Fig. 7.3.

138

~u1

LA- OA00

LLai

Ld.O

>. >. >.I-

If: -<Q CC

2t Zz

I% LJ.)

139 C.

o

L6.

onto

'Ir-

140'

7.1 ) with the stability plot on figure 7.1 for the same time

period. On figure 7.2 the stability plot variation at 0745

on 10 Say remains unexplained but may have been due to a

receiver malf unction.

Results for tests on 11 May (fig. 7.6, 7.7, and 7.8),

exhibited less stability than expected. Also, system 2 was

less stable than system 1. The largest difference in

reading between the two receivers occurred at 0750 (fig.

7.7). This difference may have been caused by the location

of the antenna of system 2 under the antenna of system 1.

At approximately 0800, the satellite constellation elevation

angles were the largest, between 56 0 to 800 above the

horizon (fig. 7.9), and antenna interference may have led to

the large differences. Neither the satellite constellation

azimuth angles (fig. 7.10) nor the user range errors (fig.

7.11) appeared to have any effect on the stability.

Comparison of the stationary GPS receiver positions with

the position obtained with the geoceiver is depicted in

figure 7.12 for the stability and static tests. The

comparison indicates, that there is a nearly constant offset

of the two positions with the GPS position almost due south

of the geoceiver position. Further comparisons between the

141

mf

STATIC TEST RESULTS

11 MAY 1981

SSYSTEM 1

a SYSTEM 2

( ) SATELLITE UPDATES

36 36 35-

uJ

I - -4

36 36 30

121 53 35-

_ 30

121 53 251 I0705 0710 0715 0720 0725 0730

GREENJWICH MEAN TIME

Fig. 7.6..

142

STATIC TEST RESULTS

*SYSTEM 1 11 MAY 1981

*SYSTEM 2

)SATELLITE UPDATES

3636 35

3636 30

121 53 30

lpl

143

STATIC TEST RESULTS

V £SYSTEM 1 11 MAY 1981

aSYSTEM 2()SATELLITE UPDATES

36 36 35

36 36 30

121 53 40

uLj

35

121 53 30

0755 0800 0805 0810 0815 0820

GREENWICH4 MEAN TIME

Fig. 7.8.

144

SATELLITE

100 ELEVATION ANGLES

VISIBLE GPS SATELLITES

90 MONTEREY, CALIFORNIA90 MAY 1, 1981

80 - .5.NAVSTAR I (SV 4).4NAVSTAR 3 (Sv 6)-'NAVSTAR A SV 8)

t70 - -NAVSTAR 5 RSV 5)-qNAVSTAR 6 (SV 9)

20

10

2 4 6 8 10 12 14 i6

GREENWICH MEAN TIME (HOURS)Fig. 7.9.

145

e - SATELLITE

200 AZIMUTH ANGLESVISIBLE GPS SATELLITES

160 MONTEREY, CALIFORNIAMAY 1, 1981

120

80

S40

k NAVSTAR 1 (SV 4)uj -I- NAVSTAR 3 (sV 6)

-i-NAVSTAR 4 (SV 8)4AVSTAR 3' (SV 5)NIAVSTAR 6 (wY 9)

-40

-80

K -120[ -160

-200 I

2 4 6 8 10 12 14 16GREENWICH MEAN 71ME (HOURS)

Fig. 7.10.

Ia

.0 co

U- -j

-~ Z3I

-' n c IL

ul VI V)V V

IC -c cc -

147

GEOCEIVER VS. GPS POSITIO.1 cOrlPUTATION

GEOCEIVER POSITION VS. GPS SHORE-BASED POSITIONFROM DATA COLLECTED

i 9, 10, 1I MAY STATIC ANDSTABILITY TESTING

d. ]36 37 1

00

, - 50

GEOCEIVER 40

GPS 30

20

_[ io

I I 36 36 0

T21 54 1Zl 5310 30 50 40 30 20 10 00

Fig. 7.12.

148

stationary GPS receiver position and the geoceiver position

reinforce this conclusion. In figures 7.13, 7.14, and 7.15,

the mean distances between the two positions are 127, 173

and 143 meters, respectively. The higher means for the data

shown in figures 7.14 and 7.15 are due to the URE (fig. 7.16

and 7.17) for those dates and the NAYSTAR upload frequency.

If the data from 13 May is deleted because of the high URE

for that date (fig. 7.16), and the median distance used for

15 May is taken to be a representative value (130 meters on

fig. 7.15), The difference is approximately 128 meters

Between the two positions. The azimuth was also nearly

constant with the observed median values of 181 , 2070 and

1780 from south as shown in figures 7.18, 7.19, and 7.20

respectively. Neglecting the data on 13 May because of URE

values (fig. 7.16), the remaining two values are nearly

identical and have a mean of 179.5 from south. Thus the

data indicates a nearly constant offset between the position

obtained with the geoceiver and the position obtained with

GPS. This difference is about 130 meters and has an almost

north-south orientation.

149

Figure 7.13.

- U.* I-

4 IL *WIUIWW W

ILlI 1I -Q

4 IL *W(NW6 0, 4m

4L ILI6

IL~

IL 1

-I Q6

4 oea WI

I u . -4 -4L 44 M -0IO'

40 U...... OX .

4 LL.

0m Z,.3

'Cc.

* IL 150

Figure 7.14.

4d..

I IL o

U.

L - . .0 . .4 .. ... t- -

go *41** I0 * *0 a0

I IL I I~40CII4N

*~ e eMQ e ~ o o e)e !4,1 4o¢ IL 4-.d4I tOOO0.

.............. ............ .. 4............

..........- .............. ..

|I *....... *......... ... . .. .... ..... . ...... .....* 4e4

* I . <'- IaIL.-

I *4 44 44 44 44 44 44 *4 44 44 4* 4 -II,

4 IL

@6ILI, I1

, IL. ... ,4 . . ?*? ?

oi 'A. I' .. a

IL 4444*4 44.44 N Z~3OOZ

.......... 0, a .........I IL- .- t-C-40 ---

@60 a

6 IL I ~ 44151

Figure 7.15.

U. 0L N

* U.I

o4 IU.4 44 4

* IL4ILUJUADWWcm a'. a

I IL.In

04r.

-, IL4

IL IL amj

4 4L

*a %F-ab-L6 I ow -Wpp

I0

in IL 0 L 0P .

6 cea 4A0n %L.044*

NI IL ... N nOOO

4. 0**00 UAU

L 4 IL -a IL ::'a aR xL * x

at IL I4M-

mu ca L aIn

mu a L a . IJ152

q~ao

LI-A

at -C -c-

1533

= yn- ru

I.R-

1 1.

(SHUN)

1.5

Figure 7.18.

nI U* 4 4 4 4.4.4 6 00 0EL 44 4044 *444 44e44 44 44444

* LL

*6 00 4-a

U. OIL .. .

*Z 4-L 4 E **WA

WEE L .4 4*4. 4444 44.4 4.&64p EL X444*444.* 444-M

ELU CL .

wE 44%44444

IEL U. 20 EOEn

.9L44

4 144-CIA

GL '

Os~~u ftI' 4O A O

*. EL 4A C92.E

EL 444 r ~49-' I V. 4444 0 90 -

*~~. EL 44464. *=

EL~~~~ 4.444 LO o

* EL 44F44 'm OOOOC=6*E . * * ~ *IllE

-~% I EL 4444- - oO0&**44 1E EL 444***! J

,A 4 U ,.

of b444 44 I W~~ %EU. I44 4 4 U6Ehl

EL 44444444

4 L C W-g EL *44* UA J-g

-. ~~O w L 444 4 ~

EL155 E KW

Figure 7.19.

* IL 4W&(AMAIWIL-' U.*4*4* ~ f~W(.

at IL 1 X*4*p *

1 IL I I

- I ILno- M4

IL NN tf.00 000** z

I W.N I U.

.4X44I W.*lab-

*0 44*0 *

* IL 12I2

* ~~4 IL L

* IL L*, It

t1 IL I 4 "* IL I

AL 1

LA. 6*4*914

*..................*. *.. . . . . . #A. *e UAMW-QWWW* IL"A

I IL AICci I.

* I I I I.41156

Figure 7.20.

* 4 W.W* 4W1

6~~~~~~z ........... 4 .. *

4. 81. 44,~~~~U U. ."*44.44444444444444

* 8. 4 Z -3-

w. U. M44C

IL 8 U. 81. l9)M

I~~~U 81 44 *44.*

U. I* w44 - U

no 81.. VIP

4-0 444 I WD .Z0 0Q1. 4 I6

4U 440 ZAC 404

4 U.1

68. ~ I wp-V9~C

I I - U. 0.2 l £ .

*w 84 36 6 I* 816 4 CwLW

N~o Z4 81.C4U .

157N4 11.

B. ERULUATION O NONDIPYENETIAL GPS

Evaluation of the non-differential GPS positions was

done on a set by set basis. Figures 7.21 through 7.31 depict

the distance in meters between the theodolite-computed ship

position and the shipboard GPS receiver position. Sets 1, 2

and 3 have the smallest distance difference, as shown in

figures 7.21, 7.22, and 7.23, with mean values between 118

and 133 meters and standard deviations ranging from 7.45 to

16 meters. These smaller mean distances are a result of the

low user range errors for 12 Say as shown on figure 7.32.

The highest mean distances occurred during sets 4, 9, 10,

12, and 13 when the ORE values were high and the NAVSTkR

uploads occurred (fig. 7.16 and 7.17). The remaining

distances vary between 144 and 278 meters.

The results during the best satellite observation

periods are depicted in figures 7.33, 7.34, and 7.35, where

the non-differential GPS ship position was compared to the

theodolite-computed ship position, the distance between the

two being plotted as a histogram. Mean distances on figures

7.33 and 7.35 are representative of the five satellites

obseVed and the improved accuracy of the system, with mean

distances of 134 and 149 meters and standard deviations of

158

L-

1

Figure 7.21.

6 *4,4U,., *44,

IL 4 ,, r I& M ,,W W U JW U .

IL Z; 6

* L I ,

00 .

A L vC

*ox . .. .. .-

. . . . .e . . . . .• . . .e i eee@• . .. . . . .. . .e . e . . . . . . .+ O

JLIL

4L L

* 6.lop I L L lb I IL IL IL ILlbmL I

, I _

lU ?. ,I ?,..

11 I 220300.

, 444, ...444*4 .4.4944

1+ o" al- 20o'

01 U 'A 44 *.4U444 44401 .

46 16 UP W II

IL I. 41

. . . . .. . . . . . . . . .,0.--.,

4L L 4 A3uz ,

U. I A44444444 *L 14 -. K ,= c

41. I C WPd 46 IL U6 ~V W%4I

IL 4- - O' 2 #.

41~Ut 6 L Z 164

4L ax

MA lb M64944 0 0 0

4 IL 4441 t4A 6 lb I L IWI~1

AN6 * - MO00

IL~xqz

* IL 6 w59

Figure 7.22.

.

LL.ol U. I *ico

.I cc t

0l

II U. 1L

I U.I L

UA U..4 IaU. usIL - Ow

at 4L Z2Zft

uL I ot.

C6ILI

IL~

4 NO '.Me@m

IL -CU Z44***4U . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .4 IL 4 F5P.It5%4w

*L 4-4 Uj OI#u" w

* *44444IL44*

*l IL 4. "

Od I I ZXML .0 515

*04 Mftwx5* II' IL I m..0

IL C LU WI -W-IL' 4 Itt= 0W

Zag IL160

Figure 7.23.

44-i *.4.44a94ea.444.4 NN-N

* MIXS4#.4e*4*49 re'e

4 a, .u.nJWWW-

.. . . .. .. . . . . ..° .. .- ' " .. .. ...Ia

* '" 4o e cet--

41

as I

a a020

* a 4 - b4 u-

I ............... .. ... ..... e.4............. .

I a WI waWassI

,,.a ,.- 0

a30 .40 ...

* a,", 4

a I * £00002

I Ua1 a0 d-.* .a. * . 4 . . .

.........................................................................................

Ni .ee* ,eetek eaeoeeeee e* oe4*.4*4 o..eo. *.. 0l€. .cw[

*a *- f-ic-flq

Sa. ..- -00000

a, a_ g~~oo o °o e eo e e *= o0n ooem

, a. *.4__.... ..... 4.e:e: .e..::..*e g o a *44 4.o...4...e..... * ne= n-.* a t- o ~ 4 e 4oe o oo o o e 2' r J•

CI I'as * a I 2 2s y..4-C "

.a Xtztt~tKtXt~trt~tz~t~~rrrrrrz~zztz 6we

: - -- " & 4I a,,4,-- .4w4w

Figure 7.24.

U. La

4 0.******

01 L I 0JIC

U6 -- a 4 7740A16 z L

I L

000000IC

I IL A J,

I ou.134-121 O &

!a 4 Ill LUC

IL 10 cc- 4

-1 U. M- Z4.4E

* ILII

4 .. .. .. . .. .. . .. .. . .. .. .

44644 0 0* 4 L

V.1 QVV1. %*

-t I *44*444 .7 00w

us ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ I .4 4 4 * . .*4 .............. 4 ......... 0W

19 10 * 4 4 UJI(WU

Figure 7.25.

U.N

*.*4**44. *6 444 444~ NN~AI3f.

U. 3 ~ c - o~-0-40

CI IL II*

tL

03 .* U. L63U IL 0

u . Tol c

LL IL 40

03 I6

of _,a UABeal.

-0 44ILw444 w6of '"Mzz**4~*****4 04444

IL4 0

*A In.

-- 4 A = 44c4,444

@. II.

4~~~- 'a4 MP.O

03 IL 3 119 440.t.~.616 IL

w w IL 3=,A IL 4 2 tL~

W03 IL 163

Figure 7.26.

w.%L.rI~~~~~~~~~~~~~~ *e*zeszeU.*4~ e4*Ot***t*tC sz N0J~t

I ** 416 4 4 4 ~ 4 4 4~~. . 4 4. . * o e

* U6*~

NI .U%#.............. . . 4 .. .... 4*....*.** * t000. 0

* *1* . . . .. . . . . . . . . . . . . .

U.)' *.SU.~r *4 . * ** * NwCwM

U. IIC

I I U INu 0 JJ.M.'5

U. u Im U

ofb4 ~ 1 1

*LU U.- -,a w

I U.0u. * 4*****4, 04

I U.. I0* ~ * s 4 * 1

U.- 4 ~ cN = U. ** 40 @*04*4 .4

0 * U6 *0 4 40 0 00 * 4 N ~ c -4CU 4w. U. 4.fl a lrc0Ne

ow IL 4 -f I

-d~~~~t I ... 0 0 0rw* 4Id)~~u Ac.ae S

4~~~~ . *. .>... * rrul

*~u c.g - c c oo*

I..U. * o Z-oo.Z.,nS

o .I *1U.Iz

4 U. fcat

4*CU0 4

0 ** 0 K J g4*.* U.0 *044 044 SM~ -14w USz-I~ ~ ~~~~4 400U0044*4 a0W0

*A xI. 4 25 aT

CI U. - urU1Z64

Figure 7.27.

1 W6 -oQ9 L. I )+' "'

S1lb I

0I N

I IL I NNU-V%0

I IL I

I ..... .-.-.. - .o - *--°' ".........''...... ........ . ... . " -,. IL 4I 4

iL I00 0 : .....

w 44 4

Ike

* -I l llll I IL lit9 20it2llll l l ll l l

1114.11111+11o111111

* IL -1IL

; .4..................e*044................44...

*...44.4* .IL 4..4.4 44..4..4. . *1*4,* .44. 41,4.44 z1OO • • • • •

Il IlLk *I4. * 4 4* 4 4 4 . 4 N 0

* l ..... 4 * . * .. '* 4,,..,-.4.

* I *44* 4** 4444 *........

-. 4 IL

9 • b .... . ........4 .. 4 .

*, lb. .. IL..''.'-

OI I 4 4 * * .. 4 @44.cc

*1 + ll o eI *4*4.o4.4*4..44. , . 4. . N iA

w" XmewZ

-. - : : : :: 1 : :::::::::: :::::::::: ::::::*::: g. . ...-.

- I... . .. ..lb... . . .... .... e.. , . .. . . . ,

I~~~~~~m lb* 4 44 4 4 t. 4 4 44 4 4I IL*444 444 *44O 444 4O*~o*

4obf1 ,0III~l b *4*40 *O*4 .@4. 4.*40.e N NO COO

- I *O.* ..........4 4444 44 44 * .4. . 4 4 . Q4 IL ..........

-, I *44 444 0440 444 4444 *.4 N qNN10

*~~~~~~~ lb 'It444 0 4 4 4* 4 4

* I I *04 *40@ 404.4 .444165.

Figure 7.28.

.. ...... ... .o. ... ..I U.* ~~~q * 4

011 -ZZ

U6 I

11 1~~

I I"Lu **...*o **-

41 ILU 4 4P.*'V+m

-g II IIL.. .. . . .. . . .. . . .. . . . . .*4... ,-

* IL.

*I L

OO-

NI~~~~~Z IL I I4*44*4*44*

o, IL * 44.,44 4**..4. 44. 4*

... ... ... .. ... ... .. ... ... ..O : t~ ~ a +l +d" WP''O

*, IL4tI,

(L ** _, *"%|

* IL+ .•. . .. . ** 4. . . . .. * . ... .

I IL* U.* 4 * * O,#". -CI U6

IL .

LL Z*A a

U. I 10000000

4 IL

-j I I I - Z A

I IL

* *444 4IL4* qU44I

It L 4 4zxz

%20*t IL I%2

4 6 IL 166

Figure 7.29.

C,*4 UU.l U

4L4

* IL 44 .44 4.041044 4. 000 Il U' N ,'4MAI

* tAL O0,40 O.0 04. 0i 0 41*0 00 0 4 * o ce

+;'°I A.44.444,44**4.4 - 4444

%16 zA W

0 Ii lNI lteelteee ILti i4*4444 **t4e44 *4ei 444ft44 e

o IA o4.o.4o4.4 .o o . 44 4,

IA.4 .. °o4o44*4 o444 ,o4 o 'U* IL * * ° * ° ° * -,

"* ILO l t l°t °l° 4, I Z iZ

°I : z

*I I A. I 4 = .I-,Iq

I6

. . . . . . . . . . . . . . . . . .. . .. .. . . I. .. .. ... ..

U. [email protected] 4

* - IL.. .11 % I9 0... .

IL ................... . WIII

I IL I *

N _. IL . : I w , QJNCO--

I IL *44444RO* * IL.. -

w C I I 0 z

* L6 I '0

16

-43 I IA.1 C1

IAt Ic IL* * . * 4 * 4 M016 IL 0 ; 4~

* IL I * - 00000

I IL I F 4 l -,It N0

01A

I 6 .z

0 Ni

I~~~a IL I 0.41A

I IL I lAI167

Figure 7.30.

4L~

NI IL4.........*..,e..ow-*. NJN nNwoL* * ** * * * ** * * *

6 M6*te*e...*.e**

1i W4MIIJ6A

. . . . . . . . . . . . . . . . . . . . . . . . ..t *..e .. e

* - 4

* ILIL,~~U at.O.*.0...O..***.4.*4.*ee UF4AI

%L wL * ... 1

I IL

01~~A =OI- W .A~ILI

* ILII

*t . . . . . IL .. . .. . . 0 . . r - -

C I I8 ILI..49L

CI I16I

Figure 7.31.

A., *0 4 0 0 0 0 L 0 400 00 00 00 0. 0 - .. 04 0 .4N4fNS ~ ~ ~ ~ ~ ~ ~ " *0 0 0 0 0 b 4 0 4

0 0 0 0 4 0 . 4 0 0 440041 .

* 00 0 04 40 0 00 00 0 00 0 04 00 0 000M640 0* S.. bWWS&SMSIU

*S&0 00*. 0004 0404 *00 44* 4.00 .000 o 0 0

0 0 : : : . . . . 0 0 0 0 4 0 0 4 0 0 .. 0 * 0 0 . 0 0 4

S 0 00 Sb0 000*0004000*000 0000*0& 00 000 0.00 4 InW6b* ~ S S

W6S . S%~,S-

I ~ ~ O4GON* Sba.* Sb 0 ~ " %.m .~4

-4 Sb .00

* 0 Sb .4

* of 0.-6 V 2 !; m-v =UA

.6 10. 0 0 . 0 0 0 0 . O 4 $5

I4 Sb.. 0 ZZZ

UA Sbt00 ,* 1 0, -A -'iG0 Q

Pa 9044 oQ -

04 Sb -I^

A.9 4000 U.0 0000 004 0000 0004 0000400 000 400.0A0I 0 40lO

40

40000000000000400#*

%1 1*04 il uIgZ g~gg t~zrxtgzggrzrzz

10 0 Sb47 5

SM~~~~~~~I 5 b0 0 4 .0 0 . 0 0 0... 0

A. I Sb 000 000 000 000 000 000 * 169A

.71

LLACM'

Ch0

-S 4

e..a I.J 1 - U

- I-,..

* Si,

qr um- tj L

170

.1*mom

Figure 7.33.

-: ****4*U.444

U.. ,

6~~~~~ It44L 44 @0@

o a- I • * ,

: io a- a .

~~~~',IIi. . ,4-l

too

of a.

I a a n, O

4 a - ......

a a-

< : -z

W 0 1

IL . W. "-a -a-,-a-

a- aoo

LA 436

*00 a-00"Oo- 0400

4, .. .... .,. .. ..0I a- TQ Cowls-

+- . **********a-*** a** *

oi M- - .G-

-e * * * * * * * ** * * * . *+ 40. 44a- 4****4 -.

Isa-e4**4 0 4.2

. . . . .. ... ... . a- -m . W ..-.- -a W Za.0..* a- a -1.-, ,u .r]I- r r ,

- . 0

US 4 -.- 4- 0

0I a I - a-Mi 20.6w a a .a1

* ~ aa- a a .4 .~ w .bIL- 4 -4 - -- *w.a

z ea a a a- Z1.Am

a a- -~ 00004 a- Ca- 4a

4n

- I - 4444444444 a- U*WU171I

..............

Figure 7.34.

U.. w

", L** 0N N N

0 N

f' I

4, 6 4 I.QI.I

o, ... .--o .,P

OI . I !

I

06 NI• e•I 0 ,d

16 To6

06 . o -. 4.Nu I

6 1 Up &

0* ~~U U. I1 2 Z Z

~ 9.

*L I' w

06 vNd0

N C 4

Uj fN - W9

In aInI N 2w ; - v rM 0 . JL9. (I

6 N 4 z .172

Fiqure 7.35.

-, ** o 444 *. 4 .4 4 4* OCn,,p ., C. ,O 4

O ******************* -

as I f LI94.

I0B *.k ,-,,r -

1. -

B. % .

* B. I Ns . . I

.. , 3 .... 3oi . I"N C$ ,

L a E@ Q

04 -Z 49

4L 0~ Q6

W.B NL

V6"B

, , tL 1 o e I L

l . .. 'Wo

B.1 I -

5 e B. •o r, e m ~ l* l e e o+e e o * *

B. I J Uk

W6 2 B. IK

iK i

.J % . 1

4 .. B . . . . . . . .. ..,

0, B. I -UU 1 lO+~k

* B.1 ~l 1 1 o ooe eooe I i I LUG ~ Id* i •e0i oo@i ~i eee@eee0 w x z g zz g t gg z z t z t p g rt e 2t w rg e , e r Ilj -ooo

173

--

10 and 52 meters. The large standard deviation for 15 May

is apparent in the bimodal empirical frequency curve (fig.

7.35) and is caused by the YAVSTAR upload (fig. 7.17). In

this case, the GPS receiver did not obtain the new ephemeris

until halfway through the time period of the observations.

The data for 13 May (fig. 7.34) reflects frequent 3AYSTAR 1

uploads and the resulting high VRE values (fig. 7.16) for

that date.

C. DIFFERENTIAL GPS EVALU&TION WITH STATIC CORRECTION

The results of the evaluation of the data of 12 May

(sets 1, 2, and 3), computed using the differential GPS

positions and the static corrections, are shown in figures

7.36, 7.37, and 7.38. Mean difference between positions for

the theodolite ship position and the GPS position are 29.1,

27.4, and 64.8 meters, with standard deviations of 7.7, 7.5,

and 22.7 meters. These distances are larger than those

determined for the theodolite ship position and the

differential GPS position (fig. 7.39, 7.40 and 7.41). This

higher mean distance indicates that the static correction,

as applied, results in a larger error in the differential

GPS position. Therefore, further evaluation of differential

GPS with static corrections were not conducted.

174

Figure 7.3f.N

" -.. o.-.*a ;;:. .. *. I . . .o

, xx "

I -, I

m * zzzzle

an ,. *- *-

IL .. a - . . -

I I lb"lb lb

OLILZ

I 6

** ,* _ poE -

-I

*A 4401~ --.°t 4 -* 4 *-.

- 1 .

- . . . .. ..... ... .... ..... . I

N I *i **i444*.* 44*.4I4i***.*.C* tII•D! N" U I Z

w1..

IL ILe) e eo o e ,1

,IC ,r M.,J - "-

a I

• *I*lb **.**.*..*.**.*** r ZI

I %l

IL lb I L N I"I

Ib I L0- -

of 16

* lb Cj *1 4* L

175

Figure 7.37.

II I

U. z UtfJW4C0II

CIL

IL I -ZP

Of ILL .6 11 W4IOa

1.I t* IIf

*L ::o toA w 2ZZZ-

a. N6i --2 It%edo L *e q**e *C~C ***e C***~q** ***"M eI IL ***.**CC CCCC C*CC C*CC CC** CSC

14L*C. * C C C C C * C * C C C C C0 .

* 4 * 0 * 0 ILt l g - -

* . edO IL. . . . . ... * . . . * . C C * C * * . * C *.' - 0 0 0 0

I1 LC b C C C C 4 C C C CC * * C * C .Z C C

P4I i* IL * ..I***N.Ooo

IL L 1 ~

41 I IL * *C** C*C* ***C1764

Fioure 7.28.

-, L9** -**9 . • ......eIL4 . 9" -* - ** 4

*I * 44. ,. k*--4.I--I.

U. wb i *ZZZ*

-- d

U*[AL

41. 1#,

I oft

,. . ....... . .................. .

61 Z.I .44I i

,," ....... :. .... .. i 6.

* -, L***4**9*4*** - jer

ILM.

@6 - sus arl.'.

n~ .- C.

.y f ..... .... ..,

, . ... £ . . ...±. ... . . . .. .. ... I

17 7 |

Figure 7.39.

-, *44*u.*4*.4. .. 44*44..O.444-0 00 -

*44~44444 04444k *4*4 o0~0C0

£ *e* t4*4* *4*** **s4*-, ~ ~ ~ ~ ~ ~ ~ ~ L *4Sk44C44*444* bW.M

44 Ob

a: IL

0g~Q wbI

'A IL* I .. 'I

=-I

10 ---

0N6l

. .4 .b .-

N d I lb 4 O

.1 . .b . .* .4 4 * .** .~ .. . . WVI ..... .

lb ~ ~ ~ ~ .*: 0 ...*4~4 * 4* S S44 ..

*~~~~~~. 17 -- )ft.S* 4 54 5 44 4 A

lb * S.4. 4.4* ... * .. *S .44. 0.- V-4C* ~ ~~ lb 8. 4.4.

aS ~~ ~ ~ ~ ~ ~ ~ ~ : lb *4 4 * * * 4 * * * ** 4 * t UJUAJ6 ~ ~~~~~~~~~~ IL* 0 4 * 0 * 4 5 54 4 . . 8 p0

*. . .. . . . . . . . . . . . . . . . . . .-

* lb *~Z~ E178

Figure 7.40.

US1. I0C.q.'. .

U. I 4 00f4I* . ,

0 1 lbl 1 du

* 16.1

I V1. I q3 .1

U. I 1.:Mm

I L6. .1lo2020 46 cc

IL'

Ib I W.O~

9" f L 4 4P 4 0 *00#4 00 -000 m- 0

4 . *. . . . . . . . 2040T% **r4 ** * 4 r r C 1, xr r- rrr r77 7-! ,

IA . InW~~c~"M Cue,44 *4 * * 6

w W. ICU. 2-4

611 I 116* *4 4

*~~k lb* a414

417

Figure ?.41.

$ 4 4 . . . . . . . . . . . **oe o*o*e **o 4 ..O .e .

°lb I 000* NP

I I

~*°~--. +,,-,,- Z ,OE

*~W xA.*zeI4 Mz

IL

IL

I IIIi iiii

: -- 1 o,..D ,,,o . .. ,, .°°... ..... .. ........ . -

C O IL I * U J4

* £i i ~ l l ll l l l . l ~ l ll l l l + l u l

4I IA. II::1 :: M. ;- *oo_ . .

0I0

-, _ . I.. .. . . _ ,

4Aoo Q. I 01 000

I w .. ... . @..*e ..e~*... .. ....... IZjm.**cee~4*.~*o** ****4444**,O*O444* 4~I91

-,~~~~~0- **rl*9 4* @ * ~ wv0~,e

44 zM 9 Pmor -mo I..

lil. .b . . .A . .. ...

4- - - IjI

-s lb **4*********** V

0-1 . .A . . ..

* IL *~#

- K KZXKLK K1rKK K~KK W2g~t~TT~TTTZ~T180t l

D. EVALUATION OF DIFFERENTIAL GPS WITHOUT STATIC

CORRECTIONS

Evaluation of differential GPS without the static

corrections were computed by sets for the five satellite

constellations for each day of testing. Set 1 (fig. 7.39)

shows a mean distance of 10.1 meters between the two

positions and a standard deviation of 5.0 meters. The

median azimuth between the two was 2570 from south (fig.

7.42), which also was the approximate course of the vessel

during the test. These values indicate that the error is

caused by the bias, or alias, in GPS position due to the

Sanpack 24-second fixing interval and positioning algorithm.

For this set, the GPS position was obtained after the

theodolite position.

During set 2 the mean distance difference between the

two positions was 11.7 meters, and the standard deviation

was 6.8 meters (fig. 7.40) ; the median azimuth was 710

(fig. 7.43). These values were slightly greater than those

of set I due to the increased URE for NAVSTkR I during the

set (see 0700 on fig. 7.32) Additionally, the azimuth was0

nearly 180 from that computed in set 1, with the vessel on

the same track as set 1. Unlike the error obtained from the

181

Figure 7.42.

. . .-,1 ...... ...... .. .. .. ... .,. . . .1.o . .. .

4u.4.4.44444.4 .4i , eo eo

05 IL IO,,gOOtd

qI IL I .

. .. . ..... .d--.

I L.

. IL-4-0

,'0 1 I rON I

44 4 44 .4 4 .4 4 44* 44- 4o .4

W ML

oIc ... . . .

4e Lb 4 ZZ

I w4 Z44 .. ... .... 4 .ZQO

4 +

It.~

Of IL II I. I '

4

..........Z .................. 0-0 ...... a .......... J n OZ '4.4

• * * IL I , L r [I.. I % rP 4

LU 1IL4 44 4 4484442£ U

'A~-S3 4 L.14W~

-, IL 44444444444444-L LU ZOLU

.4 5 Lb4 44444444 444444 '

I LbI

0 1 ILI I

40~~ Z: : 95 U

4 IL 1 44182

Figure 7.43.

U..

- I 4~~.4.4440 44 I4

*ot U. 4 *o 0 U 4F

-II *0

4IL4.*00 * *'4

U, 2 I--OMzg

ZOO 10 444IL44444444.04444

. . ... ... ... ... .. ... ... ... .. .

4IL44 04*** 44*A4$. -004

4IL 444 44 44* 44 "Moor-4"

II. I

U. ~ I ol UILILft

4 IL 4 LD ZZ

16 1 xxvw -s

U.44 I ** ,4 * 02 .. *. .

ILL. I00 44 0 N 'j, O"000

fI 44.

* IL . IL

ILL 4A 49Z II'll 14 Owl

U.' 0L I In

U. I4 4U* * 0

L 4 * * . 3 4

. 'A 3". Czo

....I IL I

0** IL I L 00000a

M4L~ 04:4 ww44 ~ .0I

4I4444 0*4444 12 IIL~0.4 *0 -6 0

at IL 4. Ot

49 l 00* # . , 4 0 0 * *X ta 'NI ~ ~ ~ ~ ~ ~ ~ ~ ~ -444L4444044*4400044 - 2IC* ~I ~~~~~~ ~f ac.* 0 * 40 4 4 4 4 44

0 4 4 4 4 C.ALO-C

I ~ ~ ~ ~ ~ ~ ~ ~ ~ w -MA44 4 44 4 4 4 4 40 4 4 4 4 4 w -04~~~t ILt 3 N.z4

NI L 44 44 444 44 44 *4 440 44 44 44 4 1 183JI

Sanpack bias on set 1, the error in set 2 was due to a

similar time delay bias because the theodolite position was

obtained after the GPS position.

Set 3 values on figures 7.41 and 7.44 are similar to

those of set 2, except the mean distance values were

greater--4.7 meters for set 3 versus 11.7 meters for set 2.

Mean distance increases for set 3 were due to the upload of

SAYSTAR 1 (fig. 7.32) and the availability of only a four

satellite constellation. Set 2 had a five satellite

constellation available.

For set 4, figures 7.45 and 7.46 show the distances and

azimuths for the set. These histograms fail to illustrate

the constantly increasing distances between the two

positions. These increasing distances result from the

uploads of NAVSTARls 5 and 6 and the large URE for NAVSTAR 1

(fig. 7.16). Onboard the vessel the new ephemeris data

uploaded for NAVSTAR's 5 and 6 was not obtained until late

in the set. Also near the end of the set, NAVSTAR 1 was

uploaded with the result that the last several distances

were very large, as shown in figure 7.16.

In set 5, a very high URE value for NAVSTAR 1

contributed to the extremely large position difference

184

Figure 7.44.

-I~ "I*.*4 UA

00

IL

%& . 0W6 & ** * * Z*d

IA.I

wL ~as£

W64 09 L"

Icoo 0

Il

IL: IL Q. I4

*9 we 491

I I- . ; *OL * 4 -. -000000'

11 *I ** 4

IL I

IL I C.~I1~O0O

185 Kv..~

Figjre 7.45.

o, -~* 4 ... C

'@4 4'o~~o0 IO

I AL

@6 - I:::::: .0~

C-f

40

to.oI'~

* 4 4

IL I" I 0

e.44o *ee e~ e e ee e eo e~ * 4'e **44**** e

18 I

* 15. ~ ZLZ

@1 4'%Ar w z zwi z Tc rrxxv z rT u.uzw TT t c xz r rr TT. I K 4

I' .4 rO~. dc

r-j

01 4' 44' 186

Figure 7.46.

06 -.&*W. z4 6z

of M.. 2x

* 1U.

@16 .0I* 11 611

j 4 4

@6W

U. 060 g~c0V6 CL

U, 02.p CI %4U . -

IL4*O0e*,e4..e. . * .

U6 I

W. I* *.*..~~~~.. . . * O. O . . 0

In 420. . . . .O O . . . . . es~.********~U.* * * t * * ** * I fdJ

* 64.64 *-0 ,. ao 3oOt

I w14 4.4 f M

6 64 I 0 0. . .S 6 64. 6 0

W64 4 0tUGS~~ X.6 U

114 664. 6 * .j 1874

between that determined by the theodolites and that for the

differential GPS position (fig. 7.47). The azimuth remained

nearly constant between the two positions with a median

azimuth of 43 (fig. 7.48). These values for set 5 show

that the large URE for NATSTAR 1 was the most important

contributing error source (fig. 7.16). The large tTRE may

have been caused by activation of the hydrazine

stabilization jets to dampen momentum on NAVSTAR 1.

For set 6, the mean distance difference between the two

0positions was 60.0 meters and a median azimuth of 93 (fig.

7.49 and 7.50). These values indicate that the upload of

NAVSTAR 1 near the beginning of the set (0715 on fig. 7.16)

was received and utilized in the Manpack position algorithm,

but the higher distance between the positions also indicates

a large (RE (also shown on fig. 7.16).

Data from set 7 were not evaluated because the

stationary 1fanpack receiver did not record the GPS position

for each fix due to radio failure. The mean distance

between the two positions for set 8 was 121 meters with a

mean azimuth of 68 0 (fig. 7.51 and 7.52). Both the

distances and azimuths show that the URE for MAVSTAR 1 at

0530 (fig. 7.17) was increasing in magnitude and thus

188

Figure 7.47.

!:wo

f 444* *4~~* *4 4 44 4 4 4 44 -044

* U. d~ U*WUJ~lUU.

'U U. 4 t

U. **.0*U.I

W . wow

* U.

4 40U.U f*4 4 4 4 4 4 *4 4 * 4 4 ~ U4

* U. 4 **

of U.g..

U. .7

*L * U.0 4 0 01 ZZ 0-z

-,0 . 0-0 C -.

U.L MA C . .

*j. U.-w I.:A~~IL*.40**4444*44

W.m4 0 6 4 "W

I w a s U. I 0 O

I .......................... -t %J14

ts 4 0*0L0 * *

4 0 0 0 0 04 4 V% 99Q C

-em 22222122Z1Z2222222221 ZTZ 2122212222~ .**I4.4444.S4.4444 ~U et-9*

I * 4'. 4.4 .4. *.4 ... .. 4 0 O 0080

Figure 7.48.

*44*4.L4*** 4.*4 . 1 NOO0

NC U.*** 4..4 4.* .

*~~~U U.*4*.**.4 Ow O~f..19

*6 OX .-l eI..l.

IL IIa

U*. at am U

I .

*1 U.TOO -

U. I v _4 ,4

91.e

r.. .0 WGQ3

I. U. W ZQ O.A

* U.0. %L

N. N IlOGGO

U. 0 V% W%ft .A ZO~e0'e

Figure 7.49.

LL

-~ ****~4U*4***i

4 I . 4 UMIAJIIW

0 !

N I N

0:9 000424. *

• S IL

* I,. !

* IL. --

C6 *+I **

•o UL jwe ewet l

IA.l

* 1w-IL . . .. '1-,

0 IL - 1-1-dl[ =. I

I IIg

*~~O L ILI .I 20 0 0

osx 3.haac.

ILL

* IL4~I flV100VI

I Naa. ... . -S : 04w~444

IU Ricwa~~ua

a Is UA.;* I a. IL 14iC N4NI

4 IL 4 02

Iu ~u ~. a 19!

SA-AI P 511 NAVAL POSTGRADUATE SCHOOL MONTEREY CA F/S 17/7APPLICATION OF THE 'GLOBAL POSITIONING SYSTEM TO NEARSHORE HYORO-(TC(U)SEP 81 V E NEWELL. 0 0 WINTER

UNCLASSIFIED

4 T 2~

11111111.5

I.. 13611111 2.01.25 1.fl4 11.6

MICROCOPY RESOLUTION TESTCHR

Figure 7.50.

IL 4. vUJWIII01 IL6

IL I * ~~of10Cdf1IL I 4o

N AZC40*9%-: ** ****,4**4**cc.~

f I -ILj

4L -o

it I Z44~

04.I

NO I - 4

4.I IL

IL I .46 g4

ILa 4 TNt c*44*IL*44*4t.*4*

*~c IL4 11,1NP40.q

t* ml- 4 Z 'W'a

Was IL

=0 IL * - 11!144

to I 11z

-c IL-c I -Cac IL I 0

* IL I I 000000-~. . . . . .. . . . . .*. . . . . . . .

IA444*44*444 * O *****192z. W~aU

Figure 7.51.

Not

44- , *ee.4e.444.444 *lb0404

014***4*4444l4*4 ooc4444*444404*4*4b*44* 4*4Q6

at lb6 1 700e000

lb S cco.a

Ul .b . . t I" Z ca.a lbc-

I 8 lb -A

* lbm

lb a.

lb ILI

... ... ... ... .. .*CZ.*lA44 *44 .*I **o4**44**44lb4**4***4 @--cc m-~

M4 ft 4 rzzxlb p e 44c4

* lb S *

*~AA lb lb flO OQ

U , lb

*0 lb SUAUUhlhMA

4 lb 4 l~t193

Figure 7.52.

Ia - M

of U. 1044 4 - li000004

* IL 0044 444 I~Om

u s I

.. . . . .. . . . . .. . . . .. . . . . . ... z

0I&

I..I a:@t IL ftO .

09 W

*~I ZVI;I £

IAU 7S LW

IIL go IL S

.04 QL 0 0-.qe~m- ....... .... t...................~

44 IL.

do4 445 S4

IL9 Im a z *4

IL 4b-vILs

OS IL I - P4194

resulting in a greater positional difference. Another

factor was the bias introduced by the Manpack 24-second

reception period. This is evident in the azimuth for the

theodolite ship position obtained from the GPS shipboard

receiver.

The data for sets 9 and 10 were combined for evaluation

and the results were a mean distance between positions of0

241 meters and an azimuth of 267 (fig. 7.53 and 7.54).

Values of that magnitude show that the URE for NAYSTAH I was

the error source. Data were sampled from 0600 to 0640 (fig.

7.17). Thus the correlation between the high NATSTAR 1 URE

and the high mean distance difference between positions.

Again, receiver bias also introduced an error into the

evaluation.

Analysis of set 11 shows a bimodal distribution in both

the two position differences and azimuths (fig. 7.55 and

7.56). The peaks on the distance histogram are

approximately 133 and 160 meters; the peaks on the azimuth

histogram, approximately 220 0 and 2360. Both peaks are

caused by the upload of NAVSTIR 3 halfway through the set

and the increasing magnitude of the NAVSTIR I ORE just after

the upload as shown by figure 7.17. The total positional

difference for set 11 was approximately 151 meoters.

195

Figure 7.53.

Os . ,4t p+~++ ,b

16.

' l 6 * .1-. IEII,

. . . ...

V6

I~

@61

61 lb

I "

U. 41 19,41

N6p4

e y . - .

o *AL 0+.,, so o :s o ,. o u. o o o 4: 0 o e e+D ,m, d,44, "I ,V6 mmd x~o @

Ot # z a.t ..,4

0V6 3tba 6 N o00 0

6 . U. 6 46644

4, . lb ,-4 . 141 . 1@6 I lbll l l l l l lI II I I III I II II I I I. I lIA 41

4 ob o* oeo :o o 0 o ',,'O

40.b. . .. . . . .

- , , • .- ;+ .-.

.4so 4 lbqo.I flvs

1A .l b . .

Ii ua u wzl' ,m

4 4 lb

........... v ............. ............. - ujwaa

* lb 4 ' * a

.46 lb 44444044444.4 4444044 4 R N6 %I4

6 lb4444*44@ 444444444*44 .

I~~ *of 044 4 4 4 * 4 4 4 4.4 cc -644

@6 lb I -w 64 h 2.I lb 6 * jlb g c-wlb 6 l~d U.4-aW a

4. lb 4196

Figure 7.54.

01

416 N4****41, lb44 *4 * 4Q * * * 4 * 4ZO4 * 4444*****e 9a

01 lb 6llldlll

l6 * Is06*..6 N+ "

C,,,i lbza6

of

W6 C . .... . .

@°6 ,,.. .. §

4L IL 'A - -* lb 4-0, )UM

: 4 IT .

I 4L III

W. 001

"* ob l -o -- "

I Mb ,7

6*.~U I44@ Odogs.O* *4

I Q* i cooa*

45. lbv G"

g .b6 ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ a wl..O .*.* ** 4 O 4444* O * O ** O **

* *.lb4*.4**o*@e*4*O~e~e*4**O9***4** ******* lb *~' A0@ %

@6 lb 6 9-00000

*C lb I ,'*44

lb 6 ~ -Z W~IJ U0

t lb197

Figure 7.55.

MAU 4AMf

*~A IL4W.MvWS

I~M IL C: £ *4*O **, *O

SIL S-W

IL IL

At IL6OIW4

*~c IL 4~SWSS

L I aSA6CCxI ILI .22 3

& IL

CI I I -0000@

-, I *..4**4e4*4.e* Wvn.

@6j ILa I. c~

I£ S. GL Ufr

*-*4444*O4 V6

U. .. .... .. . Att

IL 4. ILA

*0 IL 4% I-

-g IL *44*4q4*4*4 -* ~~ aL *..**.* *. .- A04 &

S SKKZKKZIZZ~rIZKKKKKKKSKSZZrKKSZSZ198000

Figure 7.56.

1 l

@6 Olb .

. .b . .... R .. t4.. w

* to. 4 NN NNme6 *4 u # 4 * 4 4 E *4 4 * 4 4 4 4w*44 4 4 4 4 4 *

No W6a 4054. .4

*6 *-V

C lb I ONtI'

lb *4b 0 f -- co* lbC a *

06 lb0C0N00-00000

4p I *--W W i lt

*11. w 21 mCI

- 0 66 lb--wsC

I *Ia

06 lb

Cy lb C93 4 O r

09 lbCWIJMUUlb s 2 ~ Ndat

010 lb N tn 0IS~

C Ilb VIX . 6-SItitflt4m 9b 1 4,

-s a*.....4*....a. a wN c.*~~~~~Uf lb* 44444 * *44 . V

ab *O *4 * *4to44 11*6 lb 4-4 C0.

lb*I4 4 44 * 4 4 a gt.Uw

C lb4***4444***444 mlWf UW Io a .. 4* 44** 4.*. * a 2lCJ.-

*.. ... ... ... ... ... ... O 10 a

Evaluation of the combined sets 12 and 13 (fig. 7.57 and

7.58) result in a mean distance between the two positions of

172 meters and a mean azimuth of 2620. The mean distance

was large because of the large magnitude of the NAVSTAR I

OURE (fig. 7.17). & contributing source of error was the

bias introduced by the Manpack receiver.

Overall, the data using a five-satellite constellation

were as good or better than the data using only four

satellites. The results for 12 May using five satellites

(fig. 7.59) show a mean distance of 11.4 meters and a median

of 10.6 meters. This compares favorably with the results

from sets 1 and 2 (fig. 7.39 and 7.40), which had mean

distances of 10.1 and 11.7 meters, respectively. The five

satellite data from 13 May showed a very large variance due

to the large NAYSTAR 1 ORE. Therefore, data were not

compared to remaining 13 day data. Data on 15 May for the

five satellite constellation had a bimodal distribution

because NAVSTAR 1 was uploaded. Once the uploading

occurred, the peak distance was 137 meters (fig. 7.60) which

compares favorably with 121, 241, 154, and 172 meters for

sets 8, 9 and 10, 11, 12 and 13. Thus with a five-satellite

constellation, positioning results were as good or better

than the four-satellite constellation.

200

kWL

Figure 7.57.

I IL =CC a-I U P*WeQN.SfI

41 4-4

0I04 4 S

IL 1 6I

Ww L II 14-~Nt 4 444.IL*44*4 444*444*44*44*444*4 @444404*4 NI O34

I 4 I . 4 4 4. 4 * 4 4 4*04 4 * 4 4 *4 4 * 4 4 4

du *0*444*4 4*t4*4**4*444 4 4 *4 4 4

0n IL .IN .00IttI ttt ...... . 'I 91* w #P 'r O f q

OS.~~z I I %0 I

j IL 4.t UJCIL a.m leIu= 400000lcvldd

q9 U..aS IL S N

*~4 IL4 C 4W50.0$ ILS - NCOO.

4* a .IL %A -&G.~~I

It IL Im IJ 3 .

WOL M. Ico -zNA co ILh IL " 4i

45 4 I 4N II~l L *444*44444*44444444 ONIL QL

IL .4 * * * 4 ** 4 4 * 4 4 .~ . ....... ......IL *4*464****44444**44* N' 1~ UjII

40 1~ W6 0,0 m~2LUL4.464 4 4 44 4 4 * ** ,. : z~~

4 I L **.***44*45*444*0442 M,45 IIL **~*44***4~444444444NU6

4 IL@1 IL -U.

* IL S o fl iN

42 IL: '"UM~o~9 IL 4 41- 44444

00 *4 'o JUL.4444 4 @* 4 444 zoo::::W ,I I *4*444*4*4*44444W4 N us o14 4

* IL444e44444*4444444404 . 10 NO7

Figure 7.58.

* IL 4 MWWWIL

. . ........ . . .

IIL . . ..... IAO'vO

I *& IL

I I

IL xIU

01 1L 4 r m w aI LI N 0 0-0

W6 IL 4y 40-UWI

IL 4**440 N 04..-.JM6*~~~V U&.I .. . . .. . . . . . . .

I IL ***,*4 I-NNN.4 IL 4 ~ ZZZZ

W.I IL- aas WL*.. 44 4 ** .** * ZL*

0iL IL L

U4 7u; 4-04014

I~~~ ILW, 4 4 4IL~W- 4 I - 00. .*

M6 -I- N 6O4~~o * I 4c*.~jm l0I a

IL I N ~~wz4UNUIww I IL V K C.

of AL ~C 1 O4 IL 4 OZE I

U IIL6 + 1 3-Am

of I. W"* -Vr

-L q 0000Pf

wU. 4, U. U14 .9

-V UI0Ctl

I20.2A1A0.

Figure 7.59.

0 *46

a ***444444* L*44 *-.444

S Sb S sa . a Sn -b

* Sb *L

G a SI a a~ .r~ p ,O.AN

*** a.a

W. ! at UAs as.z z".4

I uW.

O. 0400N50

a 1s. N16a~4~v u.a * r0O@

4 O.f a05 u. 'l6S~A

wtu as a

OS Sb S N 0000000

* S6 4 mI 4SSPOOS Sb a

A0 I Cb5aS 0 ud-

*L -14 i..4 4 3 -A 0 Z "SM~~I 41 1S. ** 4 * 4* 4 * 4 4 .JN ~ ~ ~ ~ ~ ~ ~ ~ 0 5 b **44044*. 0 . f

* S 1. * 4044.04.4*4. * a2035

Figure 7.60.

6..

. ..... * .. . . . .. 4. .. . . .cR. . . .. e . N

o f 44 4 4 . * O 4 4 * * ~ * * 4 4 4 - f

0 Im Ijw

ILV IL or

IL11 IL~~ @ 4*4.4 N

Of U...

*. I IWWP

IL .

04 3

IL ow S lC*I LbS

*I IL % .

IL S

us U. S-a IZ

W. I

01 ~IIL

r~~~~ z rnrz a aaw.

* IL ~ 1.'~I.,%96%

-W W I I E23 1 *

05 5 -. MAU. -4p ..d W&

t* ILL w 01A. .

4IM I ; 4Z .- @*I-32 -.1

fA6 S L I A wUj -

~~.9S"4 ILl a,

05 IL% S -c--S IL S UJIL m

* IL I14 l~204

Figures 7.61 to 7.74 show the difference by fix between

the latitude and longitude plots of the theodolite ship

position, the differential GPS position, and the uncorrected

GPS position. During most of the testing, the longitude of

both methods of GPS positioning was in closer agreement with

the theodolite ship position than was the latitude. Shown

on figure 7.9 and 7.10 are the satellite elevation and

azimuth data for the testing period. This data indicates

that the satellite constellation had a larger effect on

latitude than longitude because the predominant ranging was

computed from one quadrant.

The large differences between the GPS position relative

to the theodolite ship position were easily identified

during the shipboard testing period. Latitude and longitude

plots (fig. 7.65, 7.66, 7.69 through 7.74) for sets 4, 5, 9,

10, and 12 show the deviations of the GPS receiver positions

from the predicted positions that were obvious to the GPS

operators during the shipboard testing. For sets 4 and 5,

the stationary GPS receiver positions began to deviate

greatly from the geoceiver position, resulting in a large

differential correction to the GPS ship position. These

large differential corrections caused the differential GPS

205

LATITUDE COM1PARISON SETS 1 ,2

SHIP TEST RESULTS

I 12 MAY 1981I ~ T-2 POSITION

36 3800 LUNCORRECTED GPS POSITION

I * TRANSLOCATED GPS POSITION

50

SET 1

40

~30

2OF SET 2

10

3637 0010 20 30 40 50

FIX NUMBER

Figure 7.61.

206

L0flGITUDE COMPARISONi SETS 1, 2

SHIP TEST RESULTS

121 53 20 12 M4AY 1981

10

53 00-

SESE \1

50

40k

30[1 T-2 POSITION

aUNCORRECTED GPS POSITION

I eTRANSLOCATED GPS POSITION

121 52 2 1 -II III10 20 30 40 50

FIX NUMBER

Figure 7.62.

207

LATITUDE COMPARISON SET 3

SHIP TEST RESULTS

12 MAY 1981

AT-2 POSITION

aUNCORRECTED GPS POSITION36 37 50

TRANSLOCATED GPS POSITION

40 - ET 3

30

20

10

36 37001so5 60 70 80 90

FIX "JUMBER

Figure 7.63.

208

LONIGITUDE COMPARISON SET 3

SHIP TEST RESULTS

12 MAY 1981

121 52 40 T-2 POSITION

SUNCORRECTED GPS POSITION


30-

20-

x

- SET 3o 10-

5001

121 51 50~50 60 70 80 90

FIX 4IUMBER

Figure 7.64.

209

LATITUDE C0MPARISO."l SETS 1,5V

36 38 40 SHIP TEST RESULTS13 MAY 1981

20

SET 4

38 00

SETS5

40

20x

4x

37 00

ST-2 POSITION

40 a UNCORRECTED GPS POSITION


36 36 2070 80 90 100 110 120

FIX I48MER

Figure 7.65.

210

LON!GITUDE COM1PARISON' SETS 4, 5

121 53 50 -SHIP TEST RESULTS

13 MAY 1981

30

10

5250oX.

xc~x

3 0 X J R x x V * bS E

SET 4

10

ST-2 POSITION51 50 - UNCORRECTED GPS

POSITIONrRANSLOCATED "UPS

POSITION

72 5 0 80 90 100 110 120

FIX 4UMBER

Figure 7.66.

211

LATITUDE 'COMPARISOM SETS 6, 7

SHIP TEST RESULTS

13 MAY 1981

3638 20~T-2 POSITION


*TRANSLOCATED GPS PrX ,1.38 00

40

20 SET 6cc SET 7

37 00

40

36 36 20110 120 130 140 150 160

FIX NUMBER

Figure 7.67.

21 2

LOr'GITUDE COMIPARISONi SETS 6, 7

SHIP TEST RESULTS

K ~ ~~121 53 50K13AY98

30

53 10

xx SET 750-

30

'T-2 POSITION

a UNCORRECTED GPS POSITION

52 10 o TRANSLOCATED GPS POSITION

121 51 50o110 120 130 140 150 160

FIX NUMBER

Figure 7.58.

213

36 38 20LATITUDE COMPARISON SETS 8, 9

363820SHIP TEST RESULTS

15 MAY 1981

38 00

40 SET 8

20

SET 9

S37 00

40

I.20 m T-2 POSITIONUNCORRECTED GPS POSITIONj


36 36 00 I160 170 180 190 200 210

FIX NUMBERFigure 7.69.

214

LONGITUDE C0MPARISOM SETS 8, 9

121 3 40SHIP TEST RESULTS

15 MAY 1981

400

20

52300SE 9

* - 40 TRNLCAE *

POITO

-0- OSTO

2 01718192020

52006

]$

LATITUDE COMPARISON SETS 10, 11

1 SHIP TEST RESULTS15 MAY 1981

36 38 30

10' SET 1

37 50

LiJ

SET 10* 30

10

x

36 50 - T-2 POSITION

UNCORRECTED GPS POSITION


3636 30I I2O220 230 240 250

FIX NUMBER

Figure 7.71.

216&

LONGITUDE COMPARISON: SETS 10, 11

SHIP TEST RESULTS

15 MAY 1981

121 54 10

SET 7053 50K

30

SET 71lo-

52 50

30~ T-2 POSITION



121 52 10 20 21 220 I3 4

FIX NUMBER

4 Figure 7.72.

217

LATITUDE COM1PARISON SETS 12, 13

SHIP TEST RESULTS

36 38 30 -15 MAY 1981

38 10

50

30 SET 12

6 T-2 POSITION

37 10 *UNCORRECTED GPS POSITION

*TRANSLOCATED GPS POSITION

aSET 13

3636 30240 250 260 270 280

FIX lUMBER

Figure 7.73.

218

LOM!GITUDE COMPARISON SETS 12, 13

SHIP TEST RESULTS

15 MAY 1981121 53 10

A

52 5SET 13•52 50

30

I-,

10 SET 11

51 50

A v T-2 POSITION

a UNCORRECTED GPS POSITION30 -SET 12 " TRANSLOCATED GPS POSITION

121 51 10 1240 2O 260 270 280

FIX IUMBERFigure 7.74.

219

position shown on figures 7.65 and 7.66 to deviate a large

amcunt from the theodolite position and the uncorrected GPS

position.

220F

The objective of the study was to determine whether the

application of GPS in a translocated or differential mode

provides sufficient position accuracy for near-shore

hydrographic surveys. Several assumptions were made prior

to the test: (1) two GPS receivers required for the

differential mode performed the computations similarly, so

that in the worst case a constant bias would be displayed if

both receivers were colocated; (2) the process of updating

the ephemeris information would not contribute a significant

positional error to GPS in the differential mode; (3)

reasonable accuracy could be obtained by using the output

from the receiver control display unit, that is,

sophisticated processing methods were not required; and (4)

the Manpack receiver results would not differ when used in

stationary or low-dynamic environments (shipboard at 4

knots).

None of these assumptions proved to be valid. It was

originally expected that the two receivers would produce

similar positions if both sets were operating at the same

tine over the same location. One limitation of these

receivers, however, is that current ephemeris information is

221

not received unless the receiver has been activated prior to

the update. (Ref. 66] For a period of time after the

receiver is activated, the portion of the satellite message

that contains the updated ephemeris is not acknowledged.

Therefore, position degrades rapidly when a VAVSTkR is

updated and the current ephemeris information is not

utilized by the receiver. For this study, unfortunately,

the initial times of activation of the two receivers were

not recorded. Hence, the effect on position caused by this

receiver parameter was not defined.

The URE plots illustrate that the ephemeris updates did

contribute significantly to the range error and,

consequently, the resulting position error. The position

determined by the Manpack receivers was not affected as

seriously by the clock error in NAVSTAR 1 when a

five-satellite constellation was used. The inaccuracy of

pseudo-ranges received from this satellite required frequent

updates from Vandenburg, especially on 13 May. A receiver

that had the option of totally eliminating the erroneous

pseudo-range would have given the best results for the

study. Since the receivers operated independently, and the

ephemeris information was not concurrent, the differential

mode was effective for only a few cases.

222

k simple correction was applied to the values output on

the control display unit of the Manpack receiver. This did

not permit application of tropospheric or refraction error

corrections, other than that applied by the receiver during

computation of the position. & more sophisticated

prccessing method, resulting in greater accuracy, would

require acquisition of the pseudo-ranges and subsequent

postprocessing to determine the receiver positions.

The best results during the onboard portion of the study

were obtained when the ship was allowed to drift at

approximately I knot on 12 Bay.

In conclusion, use of two Phase I Hanpack receivers in

the differential node meets the required accuracy of 10

meters only if the ship speed is less than 1 knot and other

parameters such as ephemeris information and the number of

satellites used for the position computation introduce a

minimum amount of error. Further testing using differential

GPS should incorporate more sophisticated processing methods

and more advanced receivers to achieve the survey accuracy

required for nearshore hydrographic surveys.

223

UDESCRIPTION QZ ERRO TBEZ23

From application of error theory in the evaluation of

positional information, it is possible to establish a

meaningful accuracy statement subject to uniform

interpretation. To provide a logical and acceptable basis

FOR COMPUTATION AND COMPARISON, POSITIONAL ERRORS ARE

assumed to follow a normal distribution. The assumption is

valid because positional error components generally follow a

normal distribution pattern when sufficient data is

available.

The statistical treatment of errors utilizes measurable

errors obtained in the sources of positioning information.

Analysis of the linear components provides a two-dimensional

expression of the accuracy of the positioning system in the

form of an error ellipse. The use of an error ellipse is

complicated by the problem of axes orientation and

propagation of elliptical errors. Therefore, the ellipse is

commonly replaced by circular form which is easier to use

and understand. The linear standard errors are combined and

converted to a circular distribution and the final accuracy

statement is expressed in terms of circular errors.

224

The circular form assumes the circular probability

distribution function. This function expresses the

probability that the radial error will be equal to or less

than the radius of the probability circle. The dispersion

of errors within the distribution is measured in terms of

precision indexes, each of which represents the error which

is unlikely to be exceeded for a given probability. The

preferred circular precision indexes are the circular

standard error (a , the circular probable error (CPE or

CEP), the circular map accuracy standard (CHAS), and the

circular near-certainty error, or 3.5 sigma (3.5 a ). For

each precision index, the given probability is 39.35%, 50%,

90%, and 99.78%, respectively.

Each of the precision indexes can be converted to one of

the others by using the following table:

Circular Error Conversion Factors

39.35% 50% 90% 99.78%

39.35% 1.0000 1.1774 2.1460 3.5000

50% 0.8493 1.0000 1.8227 2.9726

90% 0.4660 0.5186 1.0000 1.6309

99.78%. 0.2857 0.3364 0.6131 1.0000

225

The rapid approximation for determining the circular

standard error is

CSE - 0.5000 (ax + ay)

wherea x and ay are linear standard errors obtained from the

sources of positioning information, and therefore are the

axes of the error ellipse. For a truly circular

distribution, where ac is identical to the circular standard

error, a x and a y are equal and the angle of intersection

between the two is 900. In all other cases, a normal

circular error distribution is substituted for the

elliptical distribution. The substitution is satisfactory

for an analysis within specifiedasin/amax ratios, wherea sin

is the minor or smaller linear standard error of the two.

Distortion in the error distribution between elliptical and

circular forms at lower Gain/ amax ratios limits the use of

the circular concept to ratios greater than 0.2 .

The CHAS is the precision index used as the U.S.

National Map Accuracy Standard. It is interpreted as

limiting the size of error which 90% of the positions will

not exceed. CHAS is computed from circular standard error

as follows:

CRAS 2.1460ac

226

The two-dimensional error distributions for positions

determined by the ARTEMIS system and the theodolite network

are described by CHAS. The accuracy of the specific

positions is expressed by a statement of probability and

error magnitude. The accuracy statement does not mean thatthe error in position is exactly the value shown in figure

5.1, rather it expresses the probability that the error in

position will not be larger than the error given.

2I

227

LIST OF REFERENCES

1. Stansell, T.&. , Ths TRkYSTilHavigAtion Satellit,Magnavox, 1 8~.

2. Denaro, R,get~al; "GPS Phase I User Equipment FieldTests": N t v.25, No.2, Summer 1978.

3. Personal communication, Lt. T. Rialon, NOAA, 2February, 1981.

4. Munson, RC. RADN NOAA ,2ggjonggy5.gim--Reoy±on theeWork of WG t~4&, fiOS Putication, June 1977.

5. DUnaP. and Rees, J. yraao-cA

7. Van der Ree, I., The ARTEMIS Position Pixing SystmA R~geRe~in sy--U. aper presenea Before!"e

Canadi.an Hydrographic Conference, Halifax, N.S., 1980.

8. Q2 i±.Stansell, 1978.

9. Dew, . E.0 LCOL, USAF., "The NAVSTAR GlobalPositionin System", rceg o f tZSg

1. Henderson D. W, and Strada J.A. "INAVSTAR Field TestResults", Naigtin Y . 2s, No. 2, Summer, 1978.

11. Easton RL. and Buissoni JA. "The contribution ofYaviqi Technology niateliltes to the GlobalPositioning Sgstem". roceedinas of the Second

12. Bartholomew, C A , "Satellite Pre ue Sadad"

Alaa.1.n, V. 5;No. 1, Summer, 198.~ tnad"

13.

14l. Dew, 1979.

1. Personal communication W.Hovr TeaIntuns1. Corporation, 3 March 1§81.Hoer TxaInrmnt

228

16. Jorfelsen, P.S., "Combined Pseudo-Range and DopplerPos t oning for.the Stationary NAVS AR User", IE!!Pjans 80. Position Location and Nay catlo SvmgOlreod1980.

17. Book S.A., et al, "The Nonuniform GPS Constellation".IEEE Plans 80 osition Location and NaviaationiYIR.osi3I.Record,.& 1 80.

18. Milliken R.J. and Zoller, C.J., "Principle ofOperation of NAVSTAR and System Characteristics",Naigatig, V. 25, No 1, Summer, 1978.

19. O.ac i%# Dew, 1979.

20. &viation Week and Space Technologyr 10 March, 1980, p.16.

21. Q tit, Book, 1980.

22. GP§Dijagam. Rockwell International--Space Division,Pun. 3541-F-1, 1976.

23. Pasahow E. "A Constellation of North Stars. NAVSTARGlobal fos!i oning System", OCEANS , day 1979.

24. Spilker, JJ.4 "S.gnal Structure and PerformanceCharacteristCs Navation V. 25, No. 1, Summer,1978.

25. Ibi.

26. 2R ci±, Milliken and Zoller, 1978.

27. Qg 9,U, Spilker, 1980.

28. van Dierendonck, A.J., et al, "The GPS NavigationMessage", AA.L ld, V. 25, No. 1, Summer, 1978.

29. Russell, S.S. and Shatbly, J.H., "Control Segment andUser Performance", N 25, No. 1, Summer,1978.

30. Q2 ci, Dew, 1979.

31. Personal communication, CAPT E. Gonzales, USAF, 5 May1981.

32. O i, Van Dierendonck. 1980

229

33. Global pgsiioninu ystem ma npack/ahicular "SarBRU3-Ient LAYU nual, Texas Instruments Inc., 1979.

3. ibid-

35. O c Spilker, 1980.

36. Z cit Milliken and Zoller, 1978.37. Teasley S.P. et al, "Differential GPS Navigation",

IEEE Plans 8o 0 poition TLocation and Navigatigon- VDOsiUa Record, 1980.

38. OM ci%, Munson, 1977.

39. Unbach, N.J., CAPT, NOAA, Rdroaraohic Mannal, FourthEdition, NOS, 1976.

40. Gossett, F.R., Manual of Geodetic Triangulation, NOS,1950.

01. Poling, A.C., Tellurometer Manual, NOS, 1961.

42. Muelle, 1.1. and Ramsayen, K., rntrndneEIn toSurvevin, Frederick anger Publishil4 Company, ]979.

43. Hodgson C.!., Manual of Second and Third-OrderTrianaulation. and Traversee NOS, 1Y:7.

00. Pfejf~ee CL CDc, &OIA !nd Gergen. J.G., DirectGeodetic Conputation klqorithm, NGS, 1975. '

45. Pfeifer L CDR BOAk, Intersection CouputationAlor , NOS, 1981.N

46. Pfeife 1c L., CDR4 NO. and Gergen, 3 G., InverseGeodetic computation Algorithm., NGS, 1915.

47. Librajy Program, YONIjSL, IBM 3033, Naval Postgraduate

School, Monterey, California, 1981.

48. U 1A Van der Ree, 1981.

49. ?feifer L. CDR, NOA Report of the FiOid Test oLahe MS Ranae-A uth2 Electronic PositioningT~s me nos, 30 June, 1930.

230

50. Nevell V. LT, NOAk and Winter, D., LCDR, NOAA,ARTEMIS Test Results, NPS, 1981.

51. Personal communication, R. Helby, PMC, NOAh, 5 April,1981.

52. U S. De rtment 9f Commerce, NOAA NOSF', P E s 00 l~ong, Stan Ars off Acuracy. ana general

Speci ictions Or Geodetic control Survey February,

53. U.S. Departmen ofComperce !0 NOSo Specificationsto UDDr% l ..ifiatin Stn dso ccuracyo -a dGeneral Sncfctj2As OfGedei Control SurveyStJuly 1975, Revised June 180.

54. Admiralty Manual of Uvdrographic Survelin, Vol. 1,Hydrograpner or tae Navy, 1 bb.

55. 92 Sit, Admiralty Manual, 1965.

56. O2 c1, Stansell, 1978.

57. Q , Milliken and Zoller, 1980.

58. .. plotted data received from SASSO, May, 1981.

59. Goodman J.H., "The Environment and Earth-SpacePropagahion--Challenges for the uture, Nval Aear-Rav I.-q Naval Research Laboratory: Falyswier

60. Q q2t, Milliken and Zoller, 1980.

61. IJlj.

62. QU gjt, Spilker, 1980.

63. 2R W, Van Dierendonck, 1980.

64. Qagci, Henderson and Strada, 1980.

65. =, Teasley, et al, 1980.

66. Persoioal ponmunication, W. Porter, Yuma ProvingGround, Arizona, August 1981.

231

BIBLIOGRAPHY

Air Force ACIC Technical ?eport Numoer A966,ricjs--o

Air Force Systems Command Test an4 gva1.uation Master PlanNumber J-190, 31IVsTkR Globa1 2 9 sl.0f tniv g j ygtew fnr M~Anne.x, by Headquarters, space Division, July 1 980.

Arm vice Technical Report Number 474 nerto

Arm1y 11L~p ervice Technical Report Number 61, Pre-orbiterInvestigation, by R. N. Harpe and others, 1967.

Bossier J3 D. ta~o e

Hoa -L ci eVy ox SCIIeNes ommittee on the Panel forGlobal Positioning System an members of the InteragencyGroupt Mokile aryland, 31 October, 1979.

* .- Bovditch N., Ainerj.c4n Practical N ~viczator 2 vol, Defense* agLng I cy Hydrograp1Lc Center, OR Stock"Number NV

Christiaan Hua genslaboratorium ARTEMIS Mntial, 9oer lign andinstAllgtioa anual for the- ARTEMIIS Position Fi7n vtmltk IL-5 ]12.

Cnossen, R Lo otGS eeVr o ega v4jMe reserued at te .nd annual Keico ter association of.ae~rica mfeeting and Industry Showcase, as Vegas, Nevada,

11-13 February, 1980.

rste atte farine Meetlu of re Inst iate of~avigation, Yorktown, Virginia, 19h0.

DeesqRgigAlny Th:e ge2&tmsnt of Degons: World

Dunn, P. D. and Rlees J. W. A gicatin 0 fG10&

Nerogra c Confertee urlngtn Ontario, Canada,S pteuber, 1980.

232

Euler V. C. and Jacobson, L J.APrpcieg ;i pof GP , pa jer presented at The Ingtitt or N~avigation 36thnnn Meet n g Monterey, California, 23-26 June, 1980

Francavillo 3., 'NAYSTAR Nev Stars in the Sky," S&Lfx.=~Warfare, Fetruary, 1979.

Heinzen M. Roo.,qrghcSrej:Ggei5Lagati 0I Ounjublined paper, ui e a Ots aNceanorp Off ce, NSTL Station, Mississippi, 91i.

Henderson, 0. V. and Strada J.A. NiAVSTAS Zield TgstResults paper presented, at fhe Nititute or avgationINational 97erospace Symposium, Springfield, Virginia, 8March, 1979.

Ingham A. Sea surveying, 2 vol., John Vlcy and Sons,i975.

Jac bqon, L. gser Eonipment fo; the NAI;TAR gIlobajlitini - stem paper presented a h

WeIecomunIcations Conference, Dallas, Texas, 1 December,1976.

Jacobson L. and Huffman L., "Satellite Navigation withGPS," Saiellite Communicati.ons, July, 1979.

Jor ensen, P. S. BAYSTAR/Global poSionj. fl Systemi8-a.Rtefl &.3JLZ.l2as, paper p resented am Tneinstitute orav gatIon Conference, Monterey, California,25-29 June, 1980.

Joseph, K. H,, *tand others, NAVSTLAR GP eceiver forSAtelli.te 1D lctlojg,, paper pree ER aXAG T3-Guidance anq Control Pa9 symposium, London, UnitedKingdom, 141-17 October, 1980.

the A piition o eas aa es asters Tesis, naa

Logistics Review Group Report, NlkYSTA&, Global PositioninoSystem IGPSI Audit Summary, 2b April, 197.

Mainavox Technical, Report SX-TH-33241-80

Rd .Q&r..QQ 2 rnU UL y %. 1. aenpa, Agust, VU

Magnavox Report 140413, Global Positioning S9stea Phase IUser Satz Descrigtions, October, 1915 u.

R~asters es6 N aval Postgraduate Scco, 3oncey,

233

Juh nnual 5yposiurn on Frequency control, Atlantic City,New Jersey, 2-4 June, 1976.

e #' eme erng o erence, H att House Hotel, Los Angeles,Califrna. 18-20 October, 1 77.

Nambeional3 Aeronautics and Space Administration ContractItudy and Analysis, by . Goad, 12 j y, 94

Naval Ocean Systems Center Field 'Test *Operations PlanFTOP-PF-1052. ATSI& Goa Positionind system on.rigate/FFI102 , July, 1978.lb

Naval Ocean Systems Center Field Test 0operationis PlanContract Number N00123-77-C-OO&6, NAVSTkR global politioziingqgl~aoL.kU, by SAI C01ISYSTENS corporation, I N ovestker,

Naval Ocean Systems Center TechnicaL Regort Number 417,

RocVSl neraina eor,1 ASA Global positioningSse hRarAtnaTss

Ranae-zi t Colln Gernicet Avonica stison papebrrese ,e

Space and Mlissile S stems Organization CS/ E DefinitionContract Nlumber P04170 -73- -029

y General Dynamics, g ectronicsl i son, 2 e ruary,1974

Space and Missile Sistems Organization CS/TIE ContractDefjtn til ? 04701-73-,,-0298, nl Rqr :o aoh

1ynamics, iec ronics iv sion.e ra-ry, j97 4 Y eer

Space and Missile Systems Or aization NAVSTAR GlobalPositionn.pe C~tatNme I-DE10 ra mpaodjj t tan r te -UhL22tinn Sr:@

234

Space and missile Systems Or~ani; ation Finai~ Field TestReport for Major Fie]4 Test Ob ective Number 1 01SiporOgrtos 1 June, 1979.

Sgace and Missile systems Organization Major Field Test

S ace 0 and Missile S S s Orianization Re r

Sequent 0YTI 05~~a Ps ong 5stem nase i

University of Texas at Atstin Bid Invitation Yumber 0088gloal ~giignng ecev n Slttmlg= Field Portable,

235

INITIAL DISTRIBUTION LIST

NO. COPIES

1. Defense Technical Information Center 2Cameron ftatignAlexandria, Virginia 22314

2. Library, Code 0142 2Naval Postqraduate SchoolMonterey, California 93940

3. Department Chairman Code 68Mr 1Department of Oceanogr.aphyNaval Post raduate SchoolMonterey, California 93940

4. DR. G. Sacksan 1Department of Electrical Engineering, Code 62SaNaval Post qraduate schoolMonterey, California 93940

5. MR. T. Stanton 1Department of Oceanography, Code:68SmNaval Post raduate schoolMonterey, California 93940

6. Professor G. Schacher 1Department of Physics and Chemistry, Code:61 SqNaval Post radua e SchoolMonterey, galifornia 93940

7. CDR D. Puccini 1Department of Air Ocean Sciences, Code:35Naval Postgraduate SchoolMonterey, California 93940

8. DR. K. Davidson 1Department of Neteorology, Code:63DsNaval Post qraduate SchoolMonterey, California 93940

9. CAPT. V. Reynolds 1Deartment of oceanoqgapgy, Code 68Na al Postgraduate SCOMonterey, California 93940

10. Professor J. J. Von Schind 3Department of.Oceanography, Code 68VSNaval Post qraua.e SchOO1Monterey, California 93940

2 36

11. Professor V. GarwoodDepartment of Oceanography, Code: 68G 9Naval Postgraduate SchoolMonterey, California 93940

12. Associate Director, Marine Operations, C7National Ocean SurveyNational Oceanic anl Atmospheric AdministrationRockville, Maryland 20852

13. Djr9ctorSce of Ocean Technology and Eqieerig Services.National oceanic and Atmospheric dministration

Rockville, Maryland 20852

14. Direct9r IAtlantic Marine Center, NOAA439 West York StreetNorfolk, Virginia 23452

15. DirectorPacific Marine Center NO&A1801 Fairvlev Avenue EastSeattle, Washington 98102

16. Director 1_nqIneering Support Office, TE3.ffice of ocean Technology and Eniingefing ServicesNatiotal Oceanic and Ataospheric dadin strationRockvlle, Maryland 20852

17. Chief, operations Division 1National Ocean Surve y No A ).OA/C71: Attn: CDR Fidel T. SmithRockville, Maryland 20852

18. NOS Liaison for Radio Frequency ManagementNational Ocean Survey (NOA )OA/C7 33: Attn: Mr. Thomas A. MooreRockville, Maryland 20852

19. Chief, Sensor Test DivisionEnqineerinq Supoort Office, OTES/VOAAOA/TE31: Attn: Rr. Euqene A. RussinRockville, Maryland 20852

20. Office of Marine Surveys and MapsNational Ocean Survey, NO&AOA/C35x3: Attn:. Mr. Jack L. WallaceRocky ille, Maryland 20852

21. Engineering Support OfficeOTES/IOAAOA/TE31: Attn: Mr. Knute A. BerstisRockville, Maryland 20852

237

f

22. Engineering Support office 102 ES/NOAAOA/TE3 1: Atta: Mr. Gary R. WhitsilRockville, Maryland 2852

23. LCDR. G. B. Hills* IEO&A 10Department of Oceanography, Code 684.iNata Postgraduate SchoolMonterey, California 939140

24. LT. T. Pulon, NOAA 1* I !OkAANOS

Code: C7346001 Executive Blvd.Rockville, Maryland 20852

25. Hr. T. B. Smith 3kassant Be in al Director Hydrography* Bedfrd Intitute of oceanography

P.O. Box 10~6Dartmouth, N. S., Canada32! 4&2

*26. LI, He~tor Sol i 1Direcci 9n De Hid rografia yNavegacion De La MarinaEgg. Roca Y GamarraChucuit o-CallaoPeru

27. LTCDR Walter Garcia 1Direccion De Hidrografia yNavegacion De La MarinaEsq. Roca Y GamarraChucuito-CallaoPeru

28. LT. D. Minkel, VOkk 1SSC ;068Naval Post t rfla*Sh~

29. Hr.D*3 Peters.Esq. I

Naval Postgrafuate schoolMonterey, California 939140

30. DirectorDefense Sapping kgencyCrog Roph ic/Topograp ic Center6500 Brookes LanesWashington, D.C. 20315

31. Directo1DeJ~s~rajpng qenl~cCenter

00 Booke LanesWashington, D. C. 20315

* t 238

32. Director

Hyarogapn ic/Topogra piic centerCode: PRH6500 Brookes LanesWashington, D. C. 20315

33. Di~:ctorIHDerogs Malpin g AgencyCode:gra~h -,Topograpiic Center6500 Brookes LanesWashington, D. C. 20315

314. Di ector 10Hyd Mga c/ orpi CenterCode: G S8S

6500 Brookes LanesWashington, 0. C. 20315

35. Director 2Defense lapping Agency* Hydrograph ic/Topograplic CenterCode: SD* 6500 Brookes LanesWashington, D.C. 20315

36. DirectorI* Defense mapping agency* Code: PPM

Bldg 56t 0. S. Naval observatoryWasing o D: C. 20305

37. DirectorIDefense Mapping AgencyCode: SSTBid g U6 . S. Naval observatoryWasing4on, D. C. 20305

38. DirectorIDefense Mapping SchoolCode: SD-IBCFt. Be] voire Virginia 22060

*39. LI. V. V. Newell, NOA 7IlOAfl i Divids onPacifc Nai; ne Center1801 Fair view Avenue, lastSeattle, Washington 98102

40. LCDR. D D. Winter NOAA 10UOAA Sb ps NRide & fieckAtlant c Barine Center439 West !Qrk StreetNorfolk, V rginia 23510

239

41. Naval Research Laboratory 1Code: 5110 (John BrozenaJ4555 Overlook Ave. S. .Washington, D. C. 20375

42. Mr. Mohan1Jet Propulsion Laboratory

4800 Oax Grove DrivePasadena, California 91109

43. Ms. P D. Dunn 2Naval Oceanographic OfficeCode: 8412NSTL StationBay St. Louis, Mississippi 39529

44. CcmmanderNaval Ocefnography CommandMSTL StationBay St. Louis, Mississippi 39529

45. Commanding OfficerNaval oceanographic officeNSTL StationBay St. Louis, Mississippi 39529

46. Commanding Officer INaval Ocean Research and Development ActivityAttn: Larry Gronmeyer, Code: 302VSTL StationBay St. Louis, Mississippi 39529

47. Comanding qfficerOceano grap c Unit OneUSYS BODITCH LT-AGS21)Fleet Post OfficeNev York, New York 09501

48. Commanding OfficerOceanog~rap hic Unit TwoUSNS DUTTN (T-&GS22)Fleet Post OfficeSan Francisco, California 96601

49. Commanding OfficerOceanographic unit ThreeUSYS H. H. HESS (T-kGS38)Fleet Post OfficeSan Francisco, California 96601

50. Commanding OfficerOceanographic unit FomraNsS CIAOV ENT (T-AGS29)Fleet Post OfficeSan Francisco, California 96601

240

51. Commanding OfficeNaval Ocepnograpaic officeCode: 8400 ;ttu: D. OuelletteNSTL StationBay St. Louis. Mississippi 39529

52. Commandinq Office;Naval Oceanoqra phic OfficeCode 8100- kttn: 9. HartNSTL StationBay St. Louis, Mississippi 39529

53. Com~apding Office; 1L aval Oceanographic officeCodoe 800; Attn: R. BiggsNSTL StatloBay St. Louis, Mississippi 39529

54. Commanding OfficerNaval Oceanolraphic officeCode: 6000; Attn: C. GetmenNSTL StationBay St. Louise Mississippi 39529

55. CommanderNaval Oceanoqraphy CommandCode 153; Atn: J. ReshewNSTL StationBay St. Louis. Mississippi 39529

56. Commanding OfficerNaval Oceanographic OfficeCode 02; &ttn: H. BeckNSTL StatiopBay St. Louis, Mississippi 39529

57. DirectorNaval Oceanography DivisionOP-952: &ttn* CDR 3. ChubbDepajtaent of the Nav_Washington, D. C. 20350

58. Defense Mapping AgeiC, HQBld 56, U. S. Naval ons PIatoryAtt. PPI (LCDR D. ,Pqcfni)Washington, . C. D30

59. Directo Iav.al Air Development Copland

Codgen4031; Attn: N. Be '!n7Warmln*ter, Pennsylvania 187

60, Sr. Mike SimsCode- K-13Naval Surfafe eea ons CenterDahlgren, Virgini& 22048

241

61. Mr. Bruce HermannCodee K-13Naval Surface Weapons CenterDahlgren, virginia 22408

62. Texas InstrumentsAttn: V. Hoover2501 South Highway 121P.O Box 405Lewlsville, Texas 75067Nail Station 3418

63. Mr. T. CalhoonDepartment of Oceanography, Code 68CuNaval Post raduate SchoolMonterey, California 93940

64. LT. R. Parsons, NOAASHC 2064Naval Post raduate SchoolMonterey, alifornia 93940

65. CDR. A. DurkeeDeputy Project Manager, DOTA, C, Space DivisionCode YE-DOTP. 0. Box 92960Los Angeles, California 90009

66. Mrf Stilianos HertikasUniversity of Nev BrunswickSurveying and Ingineeri1q DepartmentP. 0. Box 4400 / Frederigton, N. B.Canada 33B 513

67. Mr. J. HuseEarinav Incorporated1811 West Belt Drive NorthHouston, Texas 77043

68. r.B. Grayiarlnav Incorporated

1814 Vest Belt Drive NorthHouston, Texas 77043

69. Commanding Officer1. fhi D iv dsonPaciic ia=ine Center 10141801 Pairviev Avenue hastSeattle, Washington 98102

70. DirectorMava Oceanography DivisionMava Observ tot y34th and Massachusetts avenue NVWashington, D. C. 20390

242

71. Commanding Officer 1Fleet Numerical Oceanoqraphy Center3onterey, California 939 0

72. Commanding Officer 1Naval Ocean Research and Develcpment ActivityNSTL StationBay St. Louis, MS 39522

73. Commanding OfficerNaval Environmental Prediction Research Facility3onterey, California 93940

74. Chairman, Oceanography DepartmentU. S. Naval Acade.Annapolis, RD 21492

75. Chief of Naval Research800 1. Quincy StreetArlington, 1I 22217

76. Office of Naval Research(Code 4801Naval Ocean Research and Development ActivityNSTL StationBay St. Louis, [S 39522

77. DirectorNational Ocean Survey "(Ot/C) _National Oceanic and altospneric AdministrationRockville, MD 20852

78. Chief, Program*Planninq and Liaison JNC2)Nationa Oceanic and Atmospheric AdministrationRockvil 9, RD 20852

79. Associate Director3ar ne Surveys and fafs (OA/C31Naional Oceanic and a ospheric AdministrationRockville, ND 20852

80. Chief, Hydrographic Surveys Division (0/C351Natlonal Oceanic and Atmospheric adsinistrationRockville, ND 20852

81. Commanding OfficerOceanolraphic Unit FiveUSS HA KN SSFleet Post OffcNov York, MY 073

82. ISO/FIG Xnternational Advisory BoardInternationfl ,ydrographic BureauAvenue _rps d en J. F. KennedyMonte-Carl1, 3onaco e

243

83. Hy~ aher of the Navy1__iryo Defense

Hydrographic DepartmentTaunton, SomersetTA 129 NEngland

84. Hydrographer, Roial Australian Navy1Hgdorpi Off ce

North Sydney2060New South Wales - Australia

85. Doninjon Hydrographer 1- Canadian Ny dro graphic Service

615 Booth StreetOttawa, Ontario

KiAOE6

86. Kaiala Javatan Hidro oseanografi1Jal an Gunung Sahori-87Jakarta, Indonesia

87. Direccion General De Oceanografia 1Departuentc De HidrograficiaMexico 1,D. F.

88. Director 1Instituto Hidrograf~coRua Das Trinas, 439L is boaPortugal

89. Df fector De Hilirografia y Navegacion de la Marin~a 1D jecio de H ro gafi Naegaionde la Mlari.na

Callao-5 Peru

90. Civil Enqinierin q Depart:!ft rtoPorts and L ghth 6usesAdnsrtoRas Elti -Gate IAlexandr a, Egy7pt

91. Hydrographic Service 1Athens BST 902Greece

244

Documents

-EEMIhhhEEMMhE -EMEhEEEhE -EEomohEEMNOEE · 2 1. SECURITV CL.ASI2PICATI@OO*@ NO PAGE (wm Do. ... accurate method of position determination for onshore and airborne navigation