Calhoun: The NPS Institutional Archive

Theses and Dissertations Thesis Collection

1986

Traverse adjustment.

Klangvichit, Supote

http://hdl.handle.net/10945/22159
 ""

JDLEY KFOX LIBRARY

kVAL P rE 3CH00L
rtrtQ
ONTEREY, CALIFORNIA 9S943-B00S
NAVAL POSTGRADUATE SCHOOL
Monterey, California

THESIS
TRAVERSE ADJUSTMENT

by

Supote Klan gvlchit

September 1986

Thesis Co- -Advi sors Muneend ra Kumar

Glen R. S ;haefer

Approved for public release; distribution is unlimited,

T231249
StCUHl I V CLASSIFICATION OF T"Hl5 PAGE

REPORT DOCUMENTATION PAGE

la REPORT SECURITY CLASSIFICATION lb RESTRICTIVE MARKINGS
Unclassified -
2a SECURITY CLASSIFICATION AUTHORITY 3 DISTRIBUTION/AVAILABILITY OF REPORT

2b DECLASSIFICATION /DOWNGRADING SCHEDULE Approved for public release;

distr ibution i s unlimited.
4 PERFORMING ORGANIZATION REPORT NUMBER(S) 5 MONITORING ORGANIZATION REPORT NUMBER(S)

6a NAME OF PERFORMING ORGANIZATION 6b OFFICE SYMBOL 7a NAME OF MONITORING ORGANIZATION

(If applicable)
Naval Postgraduate School
Code 68 Naval Postgraduate School
6< ADDRESS (Ofy. Sfafe. and ZlPCode) 7b ADDRESS (Ofy, Sfafe, and ZlPCode)

Monterey, California 93943-5000 Monterey, California 93943-5000

3a NAME OF FUNDING SPONSORING / 8b OFFICE SYMBOL 9 PROCUREMENT INSTRUMENT IDENTIFICATION NUMBER

ORGANIZATION (If applicable)

8c ADDRESS (Ofy, Sfafe. and ZlPCode) 10 SOURCE OF FUNDING NUMBERS

PROGRAM PROJECT TASK WORK UNIT
ELEMENT NO NO NO ACCESSION NO

title (include Security Claudication)

TRAVERSE ADJUSTMENT

PERSONAL AUTHOR(S)
Klangvichit, Supote
?j TYPE OF REPORT 3b TIME COVERED 14 DATE OF REPORT (Year, Month, Day) 15 PAGE COUNT
Master's Thesis FROM TO 1986 September 87
6 SUPPLEMENTARY NOTATION

COSATl CODES 18 SUBJECT TERMS (Confmue on reverie if neceisary and identify by block number)
F ELD GROUP SUBGROUP
Surveying adjustment, least squares, observation equation
method, UTM grid coordinates adjustment.
9 ABSTRACT (Confmue on reverie if neceisary and identify by block number)

A traverse is a series of consecutive lines whose lengths and directions have been
determined from field measurements. It is chiefly used to determine the mutual location
of survey lines and station positions.
Data reduction procedures have been applied to reduce slope distances to ellipsoidal
distances to grid distances. Traverse computations were then performed in Universal
Transverse Mercator grid coordinates. The computations included adjustment by the method
of least squares observation equations. Three resection points adjacent to the traverse
line were used to analyse the quality of the results. Adjusted traverse coordinates
obtained by various methods were compared. The best results were obtained by the least
squares method with selected weights incorporated for each observation.

r
">
"'3UTiON/ AVAILABILITY OF ABSTRACT 21 ABSTRACT SECURITY CLASSIFICATION
^UNCLASSIFIED'-UNL'MITED D SAME AS RPT D DTIC USERS Unclassified
22a tjAMF OF RESPONSIBLE INDIVIDUAL 22b TELEPHONE (Include Area Code) 22c OFFICE SYMBOL
Glen R. Schaefer (408) 646-3131 Code 68Sc
DDFORM 1473.84MAR 83 APR edition may be used until exhausted SECURITY CLASSIFICATION OF tmiS PAGE
Alt other editions are obsolete
 Approved for public release; distribution is unlimited.

Traverse Adjustment

by

Supote Klangvichit
Lieutenant, Royal Thai Navy
B.S., Royal Thai Naval Academy, 1980

Submitted in partial fulfillment of the

requirements for the degree of

MASTER OF SCIENCE IN HYDROGRAPHIC SCIENCES

from the

NAVAL POSTGRADUATE SCHOOL

September 1986
 ABSTRACT

A traverse is a series of consecutive lines whose lengths and directions have been

determined from field measurements. It is chiefly used to determine the mutual

location of survey lines and station positions.

Data reduction procedures have been applied to reduce slope distances to

ellipsoidal distances to grid distances. Traverse computations were then performed in

Universal Transverse Mercator grid coordinates. The computations included

adjustment by the method of approximation and by the method of least squares

observation equations. Three resection points adjacent to the traverse line were used

to analyse the quality of the results. Adjusted traverse coordinates obtained by various
methods were compared. The best results were obtained by the least squares method
with selected weights incorporated for each observation.
 12$ l*

TABLE OF CONTENTS

I. INTRODUCTION 8

A. BACKGROUND 8

B. OBJECTIVES 9

II. TRAVERSES 10

A. GENERAL 10

1 Open Traverse 10

2. Closed Traverse 11

B. ANGULAR AND DIRECTIONAL MEASUREMENTS 11

1. Interior Angle 12

2. Deflection Angle 12

3. Angle To The Right 12

C. LINEAR MEASUREMENT 13*

D. ACCURACY 14

E. ADJUSTMENTS 15

1. Approximation Methods 15

2. Adjustment by Least Squares Method 17

III. DATA ACQUISITION AND REDUCTION 22

A. DATA ACQUISITION 22

B. DATA REDUCTION 22

1. Computation of Ellipsoidal Distances 22

2. Computation of Geodetic Distances 24

3. Reduction of Horizontal Distances to Geodetic Distances 29

C. GRID DISTANCES 29

IV. TRAVERSE COMPUTATION AND ADJUSTMENT 31

A. DATA PROCESSING 31

1. Set up of data base 31

2. Modification of an Existing Program 31

 3. New Program
Writing a 31

B. COMPUTATION OF STARTING AND CLOSING

AZIMUTHS 31

C. COMPUTATION OF TRAVERSE STATION

COORDINATES 34

D. ADJUSTMENT BY APPROXIMATION METHOD 36

1. Angular Errors of Closure 36

2. Linear Errors of Closure 36

E. LEAST SQUARES ADJUSTMENT BY OBSERVATION

EQUATIONS 37

V. ANALYSIS OF RESULTS 46
A. COMPARISON OF ADJUSTED COORDINATES 46
B. ANALYSIS OF THE REFERENCE VARIANCE OF
UNIT WEIGHT 47

VI. CONCLUSIONS AND RECOMMENDATIONS 50

A. CONCLUSIONS 50

B. RECOMMENDATION 50

APPENDIX A: LINEARIZATIONS 51

APPENDIX B: TRAVADJ FORTRAN PROGRAM 52

APPENDIX C: INDTRA FORTRAN PROGRAM 64

LIST OF REFERENCES 84

INITIAL DISTRIBUTION LIST 85

 LIST OF TABLES

I. TRAVERSE CLASSIFICATION 16

II. DATA OF KNOWN STATIONS 24

III. OBSERVED HORIZONTAL ANGLES 25

IV. MEASURED DISTANCES AND GRID DISTANCES 26

V. SPECIFICATION OF PARAMETERS 30

VI. QUADRANT OF AZIMUTH 33

VII. UNADJUSTED TRAVERSE STATION POSITIONS 35

VIII. TRAVERSE CLOSURE 38

IX. LATITUDE AND DEPARTURE CORRECTIONS 38

X. ADJUSTED COORDINATES BY COMPASS RULE 39

XI. THE COEFFICIENTS OF ANGLE AND DISTANCE

CONDITIONS 43

XII. ADJUSTED COORDINATES BY OBSERVATION

EQUATION 44
XIII. ADJUSTED STANDARD DEVIATION OF ANGLES AND
DISTANCES 45

XIV. COMPARISON OF ADJUSTED

COORDINATES'DIS TANCES OBTAINED BY
APPROXIMATION AND LEAST SQUARES METHODS 47
XV. COMPARISON OF COORDINATES AT INTERSECTION
POINTS 49
XVI. COMPARISION OF VARIANCES OF UNIT WEIGHT 49
 LIST OF FIGURES

2. Open Traverse 10

2.2 Closed-loop Traverse 11

2.3 Closed-connecting Traverse 12

2.4 Measuring of Interior Angles 13

2.5 Measuring of Deflection Angles 13

2.6 Measuring of Angles to the Right 14

3.1 Traverse Layout 23

3.2 Ellipsoidal Distance 26

3.3 Slope Reduction for Typical Triangle 27

4.1 Azimuth Computation 33

4.2 Determination of Angle and Distance from Coordinates 41-

5.1 Comparison of Adjusted Distances Obtained by Approximation and

Least Squares Methods 4$-
 I. INTRODUCTION

A. BACKGROUND
Surveying is the science and art of measurements which are necessary to

determine the relative position of points above, on, or beneath the surface of the earth,

or to establish the points in a specified position. Surveying operations are conducted

not only on land, but also in the oceans and in space. The measurements of surveying
consist of distances, horizontal and vertical, and directions. In order to provide a

framework of survey points whose horizontal and vertical positions are accurately

known, basic horizontal and vertical control surveys must be performed. A primary

use of control surveys is for construction of control for a map or chart. The
fundamental network of points whose horizontal positions have been accurately
determined is called horizontal control [Schmidt, 1978, p. 122].

Horizontal control generally is established either by traverse, triangulation, or

trilateration. Which one is to be used depends on the accuracy required and the factor
of economy in the selection of survey method. Obviously, there are many degrees of

precision possible in any measurement because no surveying measurement is exact.

Each of these methods may be the best one to use for a given purpose. Ordinarily, it is

a waste of time and money to attain unnecessarily high accuracy. On the other hand,

if the measurements are not sufficiently precise, faulty survey results are produced.

Therefore, the best surveyor is not the one who makes the most precise measurements,

but the one who is able to choose and apply the appropriate measurement with
precision requisite to the purpose.

Before 1950 the main framework of a first-order geodetic survey almost always

consisted of triangulation, which could be replaced by traverse in cases where the

topography made triangulation impracticable. Today, due to the development of

electronic distance measuring (EDM) equipment, the first-order control points can be
established by means of high accuracy traverse [Allan, 1968, p. 370]. Therefore, the

horizontal control is frequently provided by traverse, especially for surveys in area of

limited extent, mostly flat and jungle covered. Traverse in such cases is much more
economic, convenient, and rapid than other methods and the results are equally

accurate.
 In order to achieve high precision of horizontal control points when distributed

over a large area, first or second-order geodetic surveys are required. These types of
survey treat the shape of the earth as ellipsoidal and require using the most accurate

distance and angle measurements. Computation of such a survey is relatively

complicated, based on long geodetic formulas for computing (with necessary precision)

the exact horizontal and vertical position of widely distributed points on the earth's

surface.

Disregarding ellipsoid shape, a third-order survey is used over earth areas of

limited extent. In this type of a survey, the earth can be considered as fiat and all

angles are considered to be plane angles. Surveys of this type are used in the

dcnsification of geodetic control.

B. OBJECTIVES
As mentioned above, the traverse method has been used worldwide mostly for

densification of control stations. However, there are many methods of traverse

computations. The main objectives of this thesis are to (1) compute a

closed-connecting traverse and adjust station positions by the Approximation Method *

with the compass (Bowditch) rule and Least Squares Method (adjustment by
observation equations only), and (2) compare the results of the two methods.

All computations have been accomplished in the Universal Transverse Mercator

(UTM) grid coordinates rather than geodetic coordinates.

 II. TRAVERSES

A. GENERAL
The word traverse generally means to pass across. But in surveying, this word
means the measurement in a specified sequence of the lengths and directions of lines

between points on the earth whose position may be know or unknown. Traverse is the

most widely employed method for dcnsification of local horizontal control. Linear

measurements are made either by direct observation using a tape or IS DM equipment,

or by indirect observation using tachcomctric methods. The angular measurements arc

made with theodolite or transit. In this thesis, the only traverse operations considered

arc for angles measured by theodolite and distances measured by precise I:DM
equipment or tapes.

Two kinds of traverse exist in surveying, an open and a closed traverse.

1. Open Traverse
An open traverse normally originates at a point of known position and docs

not return to the starting point nor docs it terminate on another point of known
position (figure 2.1). Open traverses should generally not be used because they can

not be checked for errors.

A = known station
o = unknown station
= measured distance
= observed angle

A'
o

Figure 2.1 Open Traverse.

10
 2. Closed Traverse

Closed traverses can further be sub-divided as, a closed-loop and a

closed-connecting traverse.

A closed-loop traverse is one that originates and terminates on a single point

of known position, thus forming a closed polygon (figure 2.2). This type of traverse
provides an internal check on angles but no check on systematic errors in distance.

Also, if the starting azimuth (between stations and 1 in figure 2.2 ) has an error, it

causes error in orientation of the entire traverse. A closed-loop traverse generally

should not be used.

A = known station
o =

=
unknown station
measured distance ^>
pt ^
3 i
J
A = observed angle

Xv L 3

/ V 6
*4 A \\
6<
i
&J Ky
Figure 2.2 Closed-loop Traverse.

A closed-connecting traverse is one that begins and ends on two different

points whose horizontal positions have been previously determined by a survey of

higher or equal accuracy (Figure 2.3). This type of traverse is preferable to all others,

since computational checks are possible to detect systematic errors in both distance

and direction.

B. ANGULAR AND DIRECTIONAL MEASUREMENTS

The position of traverse points is determined by the direction and distance from

the starting point. To obtain the direction by means of azimuth, the horizontal angle,

or plane angle, must be measured in the field. Also, the determination of vertical

angles, or zenith distances, may be required to reduce slope distances to horizontal

distances.
 A = known station
O = unknown station
mi = measured distance A
Pi - observed angle

Figure 2.3 Closed-connecting Traverse.

Commonly, both horizontal and vertical angle measurements are accomplished

with a transit or theodolite. The theodolite is employed especially for surveys of high

precision. Two types of theodolite are a repeating theodolite and direction theodolite. -

A repeating theodolite reads directly to 20" or V and may be estimated to one-tenth of

the corresponding direct reading. A direction theodolite usually reads directly to 1"

and may be estimated to 0.1" [Davis et al., 1981, p. 215]. In general, a direction

theodolite is more precise than a repeating theodolite and with it, plane angles are

computed by subtracting one direction from another.

In all types of traverses, the horizontal angles can be measured by one or more
of the following listed angle measurement methods.

1. Interior Angle

Interior angle is the angle measured within a closed figure at the intersection

of two lines (Figure 2.4).

2. Deflection Angle

Deflection angle is the angle measured from the extension of the preceding line

to the succeeding line (Figure 2.5). Such angles must be identified as right or left to

express whether the angle is turned to the right or to the left from the precceding line.

3. Angle To The Right

Angle to the right is the clockwise angle measured from the preceding line to

the succeeding line (Figure 2.6)

12
 4
2
observed interior
H ~
f
^\ /<^

6
A\'"5
1

Figure 2.4 Measuring of Interior Angles.

Figure 2.5 Measuring of Deflection Angles.

C. LINEAR MEASUREMENT
Direct linear measurements may be obtained in traversing by pacing, odometer
reading, tachcomctry (stadia), subtense bar, taping, and EDM. Of these methods,

taping and EDM are most commonly used by surveyors. However, FDM equipment

has a clear superiority over traditional taping for lines in excess of about 250 meters.

Distances measured using FDM equipment are subject to personal, instrumental

and natural errors. Personal errors include misreading, improper centering of the

I3
 Figure 2.6 Measuring of Angles to the Right.

instrument over the stations, failing to exactly center the null meter, and incorrectly

measuring meteorological factors and instrument heights. Instrumental errors,

expressed in terms of the accuracy of the instrument specified by the manufacturer,

contain two parts. For example, if the accuracy of an instrument is designated as ±

(K) ppm + 5 mm), the constant error part is + 5 mm, which is independent of the

distance, and the value of the proportional part is 10 ppm (parts per million) which is a

function of the distance measured. Constant error is most significant for short

distances. For very long distances the constant error becomes negligible, but the

proportional part is important. Natural errors such as refraction are cause by

changing of atmospheric conditions along the measured line between the end stations.

D. ACCURACY
In survey adjustment, a deviation from the 'true' value is considered as an

observational error and the standard error designates the measure of accuracy of the

observation. The meaning of an accuracy is then the degree of conformity or closeness

of a measurement to true value.

The quality of traverse operations is dependent upon the accuracy of angular and
linear measurements; thus, in checking the accuracy of traverse two quantities are
considered, the angular misclosure and the linear misclosure. Although the positional
closure (relative accuracy) is an indication of the overall quality of the traverse and is

used for traverse classification, it docs not yield information on the precision of point
location determined in a traverse [Davis ct al., p. 332].

14
 The inherent weakness in a traverse is that the deviation of each measured line is

determined by a siftgle series of angular observations, further, any error in any angle

will affect not only the adjoining line but all subsequent lines to a greater or lesser

extent according to their lengths [Allan, 1968, p. 371].

Angular misclosure is expressed by standard error of the measured angle times

the square root of the number of measured angles.

Linear misclosure is commonly expressed as a ratio of total misclosure to total

length of traverse.

Finally, some of the most significant features of traverse classification by the U.S.
Federal Geodetic Control Committee (1984) are shown in Table I.

E. ADJUSTMENTS
Adjustment of a traverse is carried out to ensure consistency within the known
positions of the originating and terminating stations and to remove inconsistencies in

observed angles and distances to compensate for random errors. For a more precise
extended traverse, adjustments made on the basis of least squares are preferred. But a
traverse of limited extent can be adjusted by simple approximation methods.

1. Approximation Methods
There are four methods for traverse adjustment by approximation.
a. Arbitary Method
This method does not conform to a fixed rule. Rather, the linear error of

closure is distributed arbitarily according to arbitary preference of the surveyor.

b. Transit Rule

Transit rule is better for adjustment of the traverse where the angles are
measured with greater accuracy than distances, and is valid only when the traverse lines

are parallel with the grid system used for the traverse computations. Corrections are

made by the following rules: the correction in latitude for any station is equal to the

multiple of latitude in that section and total closure in latitude divided by the sum of

all latitudes in traverse, and the correction in departure is equal to the multiple of

departure in that section and total closure in departure divided by the sum of all
departures in the traverse [Davis et al., 1981, p. 323].

c. Compass or Bowditch Rule

This method is suitable for adjustment of the traverse where the angles and
distances are measured with equal precision and uses the following rules: the correction

15
 TABLE I

TRAVERSE CLASSIFICATION

Order First Second Second Third Third

Class I II I II

Station spacing not

less than (km) 10 4 2 0. 5 0. 5

Minimum number of
network control
points 4 3 2 2 2

Theodolite least
count 0.2" 1.0" 1.0" 1.0" 1. 0"

Direction number of
positions 16 8 or 12 6 or 8 4 2

Standard deviation of
mean not to exceed 0.4" 0.5" 0.8" 1.2" 2. 0"

Rejection limit from

the mean 4" 5" 5" 5" 5"

Azimuth closure at
azimuth check point _» , , . .

(second of arc) 7VN 30VN

1. 4. 5VN 10. 0^N 12VN
Position closure 0. 0.04^ 0. 08Vk
0. 2^K 4^K 0.8^
after azimuth or or or or or
adjustment* 1:100000 1:50000 1:20000 1:10000 1:5000
N = route distance in km
K = number of segments
*The expression containing the square root is designed
for longer lines where higher proportional accuracy is
required and the results are in meter.
The closure (e.g., 1:50,000) is relative error of
closure.

in latitude for any station is equal to the multiple of the length in that section and total

closure in latitude divided by the total length in the traverse, and the correction in

departure is equal to the multiple of the length in that section and total closure in

departure divided by the the total length in the traverse [Schmidt, 1978, p. 150].

d. Crandall Method
Crandall method is a rather complicated procedure which is more rigorous
than either the compass or transit rules but suitable for adjusting traverses where the

linear measurements contain larger random errors than the angular measurement. In

16
this method, the angular error of closure is first distributed in equal portions to all of
the measured angles, then linear measurements are adjusted by using a weighted least

squares procedure [Brinker, 1977, p. 228].

2. Adjustment by Least Squares Method

The method of least squares adjustment is based upon the theory of
probability; it simultaneously adjusts the angular and linear measurements to make the

sum of the square of the residuals (error) a minimum [Brinker, 1977, p. 228]. This

method can be used for any type of traverse. Because of the availibility of fast

computing devices at the present time, the least squares method is being widely used.

Further, the least squares solution has the advantage that it determines, quite

objectively, a unique solution for a given adjustment problem [Clark, 1973, p. 121].

In general, adjustment is needed whenever there are redundant observations

(more observations than are necessary to solve the required unknowns). As an
example, to determine the angles of a plane triangle, only two observed angles are

required because the third angle can be obtained by subtraction from 180°. When
three angles are observed, the sum of them will not be equal 180° due to error in

measurements. Therefore, these three angles should be adjusted to fit the functional

model.

The redundancy may be interpreted to mean that among n observations there

exist r conditions or functions (n > r) that must satisfy the model.

Let n be a number of observations and t\q the minimum number of

observations to find the uniquely solution in the model, then redundancy or degree of

freedom in the statistic, r, is

r = n - n (2.1)

Consequently, there are r redundant observations, which can also give a

solution. To detect the error in each observation, the best estimated or the most
probable value must be defined because the true value is not known exactly.

Statistically, the best estimated value of a group of repeated observation is the average

(arithmetic mean).

Once the difference between observed value (X_) and estimated value (X e ) is

determined, the adjusted value (X ) is obtained through a least squares solution, then

the residual (v) can be expressed as

17
 v = X -X (2.2)
a

and

Xa = Xe + dx (2.3)

where dx is the correction to estimated value to obtain the adjusted value.

The least squares adjustment method is based upon the criterion of the sum of

the squares of the observational residuals must be minimum.

When observations are considered as uncorrelated and of equal precision (with

identity weight matrix), the least squares condition can be expressed as

<P = v
i
+v 2 + -- + v n = Xv i
= minimum (2.4)

or in matrix form

(p = y T V = minimum (2.5)

where V is the vector of residuals.

In uncorrelated observations with unequal precision, such as distances and

angles [Mikhail, 1981, p. 68], the Equations 2.4 and 2.5 become

+ wn v n = minimum
= WjVj +W2V2 + ---
Xw vi i
(2.6)

or in matrix form

d) = v T WV = minimum (2.7)

where w- is the i
th
element of the diagonal weight matrix W and v- is the residual

associated with the corresponding i observation.

Generally, the relative weights are inversely proportional to variance, thus the

weight matrix is the inverse of cofactor matrix, Q (when it is square and nonsingular)
and defined as

W = Q' 1
(2.8)
where the elements of cofactor matrix Q are

= °?l°02
% (2-9)

and

2
% = a ij
/<T (2-10)

2
where <r- the variance of the i th observation, a-- is the covariance between the i
th
and
j
m observations, and <7q
Z
is variance of unit weight [Mikhail, 1981, p. 67].

For the case of uncorrected weight observations, the cofactor matrice will be

diagonal with all off-diagonal elements being equal to zero, thus the diagonal elements

of weight matrix in this case are

2 2
w =
ii
l '%= ff
/*ii
(2.11)

Generally, there are two types of equation in least squares adjustment:

condition and observation equations. Condition equations include one or more

observations but observation equations include parameters and only one observation.

The condition as well as the observation equations involved in an adjustment

problem can be linear or nonlinear. However, least squares treatments are generally

performed with linear functions, since it is rather difficult and often impractical to solve

nonlinear models [Mikhail, 1978, p. 108]. Consequently, whenever the equations in the

model are originally nonlinear, they have to be linearized. A method of series

expansion, especially Taylor's series, is often used to obtain linear equations. Only the
zero and first-order terms are used and all other higher-order terms are neglected.

Thus, the linearized form for the general case of m functions of n variables becomes

F = F ° + J mn Ax <
2 12 )
-

where F is the zero-order terms, when x = x , and J is a Jacobian matrix of

coefficients of first order of n variable (Appendix A).

The choice between the observation equations (indirect observation) and

condition equations (observation only) techniques will depend mainly on the

19
mathematical model of the problem to be solved. However, the final answers are

always the same.

a. Adjustment by Observation Equations

The method of adjustment by observation equations is performed with both

observations and parameters. Therefore, the number of equations is equal to the

number of observations. Using the example at the beginning of this section, if three

measured angles and their residuals in a plane triangle are assummed to be a, P, y, Vj,

V2, and v 3 , respectively, and the adjusted value of those angles is x

a j,
x
a2 , and x a3
then the three condition equations in the normal form (zero at the right-hand side) are

(a + v )-x al =
1

(P + v 2 )-x a2 =

(Y + v3 ) =
^a3
by using Equation 2.3 for adjusted values

=
1
dxj x
eF a
2" dX') = x
e2" P
3" dx 3 = x _
Y
e3
letting

-1 dx-i - a
v = B = -11 x = dx* f = el -
, ,
e2 P
-1 dx§ " y
e3
then, expressed in matrix notation as

V 31 + B 33 X 31 = F31 an d the general form of the adjustment of observation

^31
equations become

V nl + B nu X ul ~ F nl (2.13)

where V is an n by 1 vector of residuals; B is an n by u matrix of coefficients; X is an u

by 1 unknown vector in which u is the number of parameters; and f is an n by 1 vector

and equals P - XQ , in which P is evaluated using the estimated value X e>

b. Adjustment by Condition Equations
The method of adjustment of condition equations only has no parameters
included in the condition equations. Thus, the number of condition equations is equal

to the number of redundancies. From the example above, n= 3, nQ= 2, and r= 1,

which is the number of condition equations being set. In this case, because the sum of
interior angles must equal 180°, the single condition equation can be expressed as
a + p + Y + vj + V2 + v 3 = 180°

20
 r

l
+ v 2 +v 3 = 180° -a -
P -
y

A = | 1,1,1 |, V = Zi F = |
180-a-P-y
7r2
v3
Then, the general form of this technique is

A rn V nr = F rl (2.14)

When the conditions are originally linear, the vector F is usually written in terms of the

given observations as

F rl " P rl " A rn X o,nl (2.15)

where A is the coefficient matrix V., P is a constant term (see Section II. E. a), XQ is

observed values, r is redundancy, and n is a number of observation [Mikhail, 1976, p.

173].

21
 III. DATA ACQUISITION AND REDUCTION

A. DATA ACQUISITION
Taverse data used in this thesis were obtained from field work accomplished from
25 September thru 9 October 1972 by CAPT Glen R. Schaefer, NOAA Corps, and Mr.

Jim D. Shea, National Ocean Service (NOS), utilizing traverse methods in Pinellas

County, Florida. Only the first 15 of 40 occupied stations and three intersection points
will be used for analysis (Figure 3.1). The two pairs of known stations for this

closed-connecting traverse are shown in Table II.

The known stations were observed by the US Coast and Geodetic Survey (now
NOS) and adjusted by the National Geodetic Survey (NGS). Station Turtle 2 is of

first-order and the other three stations are of third-order.

The horizontal angles measured by the method of angles to the right (Table III)

were turned with a Wild T-2 theodolite, according to the specifications of third-order

class I traverse, by starting at station Tomlinson and using Egmont Key Lt. House for t

a backsight. The traverse was closed on Turtle 2 with a check azimuth to Madeira
Beach Tank.
The slope distances were measured in the field with a Model 76 Geodimeter in

feet and corrected for temperature and pressure. Distance measurement by the Model
76 Geodimeter are reported to have an accuracy in the temperature range of -20°C to
+ 50°C of ±(1 ppm + 1 cm) with a resolution of 1 mm. [Schmidt, 1978, p. 116]. all

observed distances were converted to meters and reduced to horizontal by the

procedure given later in this chapter. Finally, geodetic distances are reduced to grid

distances by applying the scale factor correction (Table IV).

B. DATA REDUCTION
1. Computation of Ellipsoidal Distances

For the requirement of high precision in the first-order traverse, the measured
distances obtained by EDM equipment are first corrected for atmospheric conditions

and then reduced to the ellipsoid (Figure 3.2) by the equations

S = 2R a Sin- (S
1
/2R a ) (3.1)

22
 re IP 5

A = known station
O = traverse station
IP = intersection point

T'^-o 11

Nor TO Scali

Figure 3.1 Traverse Layout.

23
 TABLE II

DATA OF KNOWN STATIONS

Station Station Grid coordinate m. (
)
No. Name X/Easting Y/Northing
Egmont Key Lt. House 326214.833 3054014. 779
1 Tomlinson 329400.420 3063485. 483
15 Turtle 2 325348.292 3076179. 297
16 Madeira Beach Tank 322698.706 3076362. 405

= {[m 2 -(h -h + h /R a )(l + h 2 /R a )} 1

/2
S )]/(l (3.2)
2 1 1

1/R a = {(Cos 2 a)/M} + 2

{(Sin a)/N} (3.3)

M = 2 2 2 3 2
a(l-e )/(l-e Sin (p) / (3.4)

2 2 2
N = a/(l - e Sin <p)
1 /
(3.5)

where m is the measured distance, S is the ellipsoidal distance, Sq is chord distance, hj

and h 2 are the heights above the ellipsoid at each end of the line, Ra is the radius of

curvature of the chord distance m at an azimuth a, a is geodetic azimuth clockwise

from north, M is radius of curvature in the plane of the meridian, N is radius of

curvature in the prime vertical, a is a semimajor axis of the reference ellipsoid, e is its

eccentricity, and (p is the mean latitude of that line. [Torge, 1980, pp. 48,49,50,51,180]

2. Computation of Geodetic Distances

For a lower-order traverse, the measured distance can be reduced to mean sea
level (MSL) or geoid only. Because the error of only 1 ppm in length results from an
error of 6 m in separation between MSL and the spheroid [Bomford, 1980, pp. 42, 342,

345].

24
 —

TABLE III

OBSERVED HORIZONTAL ANGLES

A) For Traverse

At Stn. Stn. Ancrles (T

No. Name D H S ( ±)
Egmont Key Lt. House
1 Tomlinson 105 21 25 5"
2 TN-01 243 39 18 5"
3 TN-02 168 10 15 5"
4 TN-03 59 55 56 5"
5 Ruscue 291 28 29 5"
6 RE-01 160 43 42 5"
7 RE-02 269 55 50 5"
8 RE-03 92 43 29 5"
9 RE-04 178 22 51 5"
10 RE-05 182 31 19 5"
11 RE-06 196 54 42 5"
12 RE-08 168 51 46 5"
13 RE-09 161 06 42 5"
14 RE-10 236 58 14 5"
15 Turtle 2 78 37 24 5"
16 Madeira Beach Tank

B) For Intersection Points ( IP)

IP #1 *
At Traverse Ancrles (7
Back Stn. Stn. D K S ( ±)
RE-01 RE-02 196 23 46 5"
RE-04 RE-05 184 34 22 5"
RE-05 RE-06 186 29 15 5"
RE-06 RE-08 34 43 31 5"
RE-08 RE-09 2 41 22 5"

IP #2 **
RE-03 RE-04 175 35 01 5"
RE-04 RE-05 97 12 06 5"
RE-05 RE-06 2 32 22 5"

IP #3 ***
RE-08 RE-09 194 01 22 5"
RE-09 RE-10 232 11 34 5"
* St.Petersburg BCH CO Tank
xk St.Petersburg BCH St. John s CH Towe r
*** Bay Pines Veterans Administ ration Ho sp.

25
 TABLE IV

MEASURED DISTANCES AND GRID DISTANCES

At Stn. Measured Distances Grid D istances (T

To Stn. ft m ft m (
tm)
1 2300. 98 701. 340 2300. 91 701. 318 0. 017
2 2605. 41 794. 131 2605. 28 794. 090 0. 018
3 4452. 48 1357. 119 4452. 31 1357. 068 0. 024
4 709. 56 216. 274 709. 54 216. 267 0. 012
5 5050. 37 1539. 356 5050. 21 1539. 307 0. 025
6 6620. 54 2017. 945 6620. 33 2017. 880 0. 030
7 1655. 17 504. 497 1655. 12 504. 481 0. 015
8 1605. 85 489. 464 1605. 80 489. 448 0. 015
9 2114. 68 644. 556 2114. 61 644. 535 0. 016
10 2362. 13 719. 979 2362. 05 719. 956 0. 017
11 1360. 16 414. 578 1360. 10 414. 561 0. 014
12 5060. 98 1542. 590 5060. 41 1542. 415 0. 025
13 6172. 26 1881. 309 6171. 67 1881. 128 0. 029
14 6664. 72 2031. 411 6664. 51 2031. 346 0. 030
15

m -
measured distance V^~— ^n>
So chord distance V\A S
s = arc distance on
ellipsoid
V=
\ s<,
^=S/k
»1 elevation above ^
ellipsoid at
station 1
=
/r«
ha elevation above
ellipsoid at station 2
R~ radius earth curvature
along measured line

Figure 3.2 Ellipsoidal Distance.

However, the process of reduction requires three steps: (1) correct slope

distances to horizontal distances, (2) reduce horizontal distance to geodetic distances,

and (3) change the geodetic distances to grid distances.

'I he slope distance data used included correction for atmospheric conditions.

26
 In a plane survey, such as a traverse, there are two considerations for the

reduction of slopojdistances to horizontal distances, a short slope distance and a long

slope distance.

a. Slope Reduction for Short Distances

Short slope distances (

< 2 mi or 3.3 km) measured by using EDM
instruments separate from the theodolite, can be reduced to horizontal with a simple

trigonometry process as

D = m cos 9 (3.6)

where D is the horizontal distance, m is the slope distance, and is the vertical angle

(Figure 3.3).

[slope correction
U ._ "^ «-,

r V

Figure 3.3 Slope Reduction for Typical Triangle.

The horizontal distance also may be determined by using the difference in

elevation between the two ends of the line. The horizontal distance is

D = (m 2 - d
2
)
1 '' 2
(3.7)

in which d is the difference in elevation between the two end points. The heights of the

EDM instrument and reflector above the survey mark must be observed, and d

becomes

27
 d = (E +
1
I
1
)-(E 2 +0 2 ) (3.8)

where Ej and E 2 are the elevation at each end of the line respectively, Ij is the height

of the instrument, and is the height of the reflector. Then, expanding the right side
2
of equation 3.7 with the binomial theorem yields

D = m-(d 2 /2S + d 4 /8S 3 + ...) (3.9)

The quantity in the parenthesis is designated slope correction. For

moderate slope the first term is usually adequate. When the slope distances and

vertical angles are obtained by separated EDM equipment from theodolite, the
correction of the vertical angle must be determined. The corrected vertical angle Qj is

G
T = 9 0+ AG (3.10)

where 9q is an observed vertical angle by theodolite, and AG is

AG = H CosG / S Sin 1" (3.11)

and

H = (H r - H t) - (^ - H e ) (3.12)

where H r is the height of the reflector, H is the height of the target, H-

t
is the height of

the EDM, and H e is the height of the theodolite [Davis et al., 1981, pp. 103-104].

Equations 3.11 and 3.12 are not needed when the slope distances and vertical angles
are obtained simultaneously by using an EDM transmitter built into a theodolite,

b. Slope Reduction for Long Distances

Slope reduction for long distances (> 2 mi or 3.3 km) involves using

vertical angles affected by curvature and refraction. By assuming a mean radius for the

earth of 3959 mi or 6371 km, then the curvature correction (C), expressed as an angle

in seconds, is 4.935" per 1000 ft or 16.19" per km and the horizontal distance, D, is

D = m Sin(90°-G-C) / Sin(90° + C) (3.13)

28
for

6 = (Y + P)/2 (3.14)

where y and are the vertical angles at each end of the measured line [Davis et al, pp.

106-107].

When a single vertical angle (y) is observed, is the corrected vertical angle

for combined results of curvature and refraction (C&R), then, G is y + (C&R)". The
C&R correction is 4.244" per 1000 ft or 13.925" per 1000 m. The correction of C&R
will be positive when the vertical angle is an elevation angle and negative in the case of

a depression angle [Davis et al, 1981, p, 108].

3. Reduction of Horizontal Distances to Geodetic Distances

The horizontal distance at same elevation above the geoid, must be reduced to

a geodetic distance. This can be done by the equation

D' = (R)(D)/(R+E) (3.15)

where D' is the geodetic distance, R is the mean radius of the earth's surface at that

section, D is the horizontal distance at elevation E above the geoid [Davis et al., 1981,

p. 107].

C. GRID DISTANCES
The traverse computation, based on the UTM grid coordinate system, requires

the reduction of geodetic distances to the plane of the projection by applying the

projection scale factor and grid scale constant. Scale factor can be obtained from a

graph or from a rigorous formula [Department of the Army Technical Manual, 1958,

pp. 3,4,9,17]; i.e.,

= 2 4
k k
Q [ 1 + (XVIII) q + 0.00003 q ] (3.16)

where k is the scale factor at scale working on the projection, k^ is the central scale

factor which is an arbitary reduction applied to all geodetic lengths to reduce the

maximum scale distortion of the projection (for UTM, kQ = 0.9996), and values for q

and (XVIII) are obtained by the formulas which shown in Table V.

29
 TABLE V
SPECIFICATION OF PARAMETERS

1 + e' Cos 2 (p . 1 . 10 12
XVIII =
2
2 v k^"
a
V =
z
( 1 - e Sin 2 ^) 1 / 2
e2 = (ecentricity) 2 = ( a2 - b2 ) / a2
e2 = e2 e2
'

/ (1 - )

q = 0. 000001 E'

E = grid easting = E' + 500,000 when point is

east of central meridian, 500,000 - E
when point is west of central meridian
V = radius of curvature in the prime vertical
a = semi-major axis of the ellipsoid
b = semi-minor axis of the ellipsoid
<P
= latitude

30
 IV. TRAVERSE COMPUTATION AND ADJUSTMENT

Linear measurements and angles must be checked by computation to determine

the position of traverse stations and whether the traverse meets required precision.

Traverse station coordinates are usually expressed in terms of geographical coordinates

or rectangular coordinates such as those based on an UTM projection. Traverse

computations are usually done in rectangular coordinates because of the ease of

computation. In this thesis, only closed-connecting traverse computations in UTM

coordinates were used. When specifications were satisfied, the traverse was adjusted
for perfect geometric consistency among angles and lengths.

A. DATA PROCESSING
1. Set up of data base

Two files on IBM 370/3033AP main frame computer at NPS were established.
2. Modification of an Existing Program

The Fortran program TRAVADJ, originally written by LCDR Saman

Aumchantr, RTN (1984) in Watfiv language for computing and adjusting the traverse

station coordinates, was modified to be able to handle the distances reduction

processes.

3. Writing a New Program

A Fortran program INDTRA was written for computing and simultaneously
adjusting traverse station and intersection point coordinates by least squares

observation equations method.

B. COMPUTATION OF STARTING AND CLOSING AZIMUTHS

The directions of lines by means of azimuth are used for traverse computation,

because sines and cosines of azimuth angles automatically provide correct algebraic

signs for latitudes and departures.

The terms latitude and departure are widely used in rectangular coordinate

calculations of surveying. The latitude of a line is its projection on the reference

meridian, which differs from geographic latitude. The departure of a line is its

projection on the east-west line perpendicular to the reference meridian. In traverse

calculations, north latitudes and east departures are considered plus; south latitudes

and west departure, minus.

31
 Latitudes are also sometimes termed 'northings' and 'Y differences' (AY);

departures are similarly called 'eastings' and 'X differences' (AX).

Because the closure angle of traverse can be checked by azimuth of each

consecutive line, starting and closing azimuths have to be determined in the first step

of computation and azimuths can be computed from a pair of known coordinate

station positions at the two ends of the traverse (Figure 3.1)

The azimuth of the line from A to B, « A g, measured clockwise from north, is

determined by the equation

a AB = Tan'^AX/AY) (4.1)

for

AX = X B X A - (4.2)

and

AY = Y B" Y A (
4 3)
-

where XA and Xg are the grid easting coordinates, and Y^ and Yg are the grid

northing coordinates of stations A and B, respectively (Figure 4.1). The quadrant of

the azimuth of line AB, a^g, is dependent upon the sign of AX and AY (Table VI).

The back azimuth ag A (the azimuth from B to A) is obtained by adding 180° to the

forward azimuth «
A g.
The length of the line AB (denoted as S^g or S) can be determined by the

Pythagorean theorem or by one of the trigonometric relationships

S = AX / Sin a (4.4)

or

S = AY/ Cos a (4.5)

32
 Figure 4.1 Azimuth Computation.

TAB LB VI
QUADRANT OF AZIMUTH

Quadrant of azimuth measured Sign of Sign of

clockwise from north AX AY
to 90
90 to 180
180 to 270
270 to 360

Substituting data from Tabic II into liquations 4.2 and 4.3, the AX and AY
between Fgmont Key Lt. House and Tomlinson are computed as AX = 3185.585 m
and AY = 9470.704 m. The azimuth of the line from Fgmont Key Lt. House to

Tomlinson is 18° 35' 27.6" (Equation 4.1). Similarly, the azimuth from 'Turtle 2 to

Madeira Beach 'Tank is computed to be 273" 57' 12.0". These starting and closing

azimuths will be used Tor computing the coordinates of each traverse station and Tor

checking the angular error.

33
C. COMPUTATION OF TRAVERSE STATION COORDINATES
Computation- of traverse station coordinates is the reverse process of finding

azimuth and distance from coordinates. Therefore, the rectangular coordinates for
each closed-connecting traverse station can not be computed unless forward azimuth
and distance from the previous station are known.
The azimuth is reckoned clockwise from north and obtained by

a jk = a + 180° +
X]
Pj (4.6)

where <*: is the forward azimuth from station j to station k, a- is the forward azimuth
k
from station i to station j, and Pj is the horizontal angle at station j for j values of 1 to

n, where the previous station i = j

- 1 and the next station k = j + 1. The number j

will increase from 1 (which designates the starting known station of the traverse) to

number n (which was the last occupied and known station of the traverse).

Departures and latitudes are then computed by using Equations 4.4 and 4.5

which are rewritten as

AX jk = S
jk
Sin a jk (4.7)

and

AY jk = s
jk
Cos
V < 4 8)
-

where S: is the distance between stations and k. The coordinates of all other
k j

traverse stations can be determined by adding successive departures (AX- k ) and

latitudes (AY:
k)
to the previous station's X and Y coordinates, respectively.

Using the data in Tables II, III, and IV, the azimuth and coordinate
computations at the first station are shown here by using Equations 4.6, 4.7, and 4.8.

The result for other stations can be seen in Table VII.

Computing the azimuth from station 1 to 2

Az = 18°35'27.6" + 180° + I05°21'25" = 303°56'52.6"

Computing the coordinates at station 2

X-easting = 329400.420 + [701.3 18 Sin(303°56'52.6")]

= 328818.645 m
34
 Y-northing = 3063485.483 + [701.318 Cos(303°56'52.6")]
= 3063877.126 m
When the coordinates of all stations have been computed, they are still

unadjusted coordinates and can then be adjusted by one of the two methods of Section

II.E (approximation and/or least squares method).

TABLE VII
UNADJUSTED TRAVERSE STATION POSITIONS
Stn. Angles Forwa]rd Dist. Unadjusted
Azimuths (m) Coord mates
D M S D M S X(m) Y(m)
18 35 28
1 105 21 25 *329400. 420 *3063485. 483
303 56 53 701. 318
2 243 39 18 328818. 645 3063877. 126
7 36 11 794. 090
3 168 10 15 328923. 709 3064664. 236
355 46 26 1357. 068
4 59 55 56 328823. 700 3066017. 613
235 42 22 216. 267
5 291 28 29 328645. 029 3065895. 760
347 10 51 1539. 307
6 160 43 42 328303. 493 3067396. 699
327 54 33 2017. 880
7 269 55 50 327231. 464 3069106. 259
57 50 23 504. 481
8 92 43 29 327658. 538 3069374. 789
330 33 52 489. 448
9 178 22 51 327418. 001 3069801. 054
328 56 43 644. 535
10 182 31 19 327085. 512 3070353. 210
331 28 2 719. 956
11 196 54 42 326741. 616 3070985. 723
348 22 44 414. 561
12 168 51 46 326658. 106 3071391. 785
337 14 30 1542. 415
13 161 6 42 326061. 428 3072814. 113
318 21 12 1881. 128
14 236 58 14 324811. 349 3074219. 797
15 19 26 2031. 346
15 78 37 24 325348. 180 3076178. 924
273 56 50
16
* Coordinates for station s 1 were known and held fixed.

35
D. ADJUSTMENT BY APPROXIMATION METHOD
In this thesisTthe method of Compass or Bowditch rule was used to adjusted the

data in Tables III and IV. Thus, the first step is to determine the angular error of

closure and adjust the angles to obtain the proper closing azimuth (closed azimuth at

fixed stations).

1. Angular Errors of Closure

In the closed-connecting traverse, when there are n stations of observed

horizontal angles, n-1 lines will be measured. An angular error in traverse can be

checked and obtained at the last station by comparing the computed azimuth and
closing azimuth at the known station.

The closing azimuth computed (from the known station coordinates at the

traverse end) at the station 15 is 273° 57' 12.0". But because of error in measurement,
the azimuth computed through the traverse at this station is 273° 56' 49.6", which is a

difficiency of 22.4". This amount of angular error in 15 observed stations meets the

limit for allowable error for a third-order class I traverse (allowable error from Table I

is 38.7").

The average correction (Table VIII) is distributed uniformly over all the 15

traverse angles (Table III).

2. Linear Errors of Closure

When all angles have been corrected, the process of calculating the improved

coordinates of all traverse stations may be done. The check on angular closure for a
closed traverse does not guarantee that the entire survey is correct, because there can
be considerable errors in the linear measurement of individual lines. Such errors may
not show up in the angular check. In order to check the closure of the traverse, it is

necessary to determine linear error.

The linear error (LE), the departure error (Sx), and latitude error (dy) in a

traverse are determined by equations

LE = [(6x) 2 + (8y) 2 ]
1 /2
(4 9 )
.

Sx = GE n -GE n (4.10)

36
 8y= GN n -GN n (4.11)

where 5x and 5y are the traverse closure in departure and latitude, GE n and GE n '
are

the known and computed grid easting, and GN and GN '

are the known and
computed grid northing at the closing station, respectively. By substituting the fixed

and computed values of grid easting and northing in Equations 4.10 and 4.1 1 for the

data of Tables II and VII

5x = 325348.292 - 325348.180 m = + 0.112

Sy = 3076179.297 3076178.924 = + 0.373 m

-

0.389 m
2 2 2 =
LE = [(0.112) + (0.373) ]
1 '

The relative error of closure provides a better assessment of the quality of a

traverse than the linear error of closure. Therefore, it is common practice to calculate

the relative error of closure, which is the linear error of closure divided by total

distances of traverse, and to express the result in the form of a ratio with unity as the

numerator. For the data of Tables II and VII, this computation is 0.389 / 14853.800

or 1 : 38,185.

Using the Compass or Bowditch rule, the computed traverse closures and
corrections were obtained (Tables VIII and IX) and then the adjusted station

coordinates (Table X) were computed.

E. LEAST SQUARES ADJUSTMENT BY OBSERVATION EQUATIONS

The adjustment by observation equations, shown in general form by Equation
2.13, can be done more directly than the adjustment by condition equations. To
achieve this, the explicit expression for the residual vector V from Equation 2.13 is

substituted in Equation 2.7 to obtain the following equation:

O = (F BX) T
- W
(F - BX) (4.12)

= (F T B T X T )(WF WBX)
- -

= X T B T WBX - X T B T WF + F T \VF - F T WBX

= X T B T WBX - 2F T WBX + F T WF

where X T B T \VF = F T WBX are scalar quantities.

37
 TABLE VIII
TRAVERSE CLOSURE

I) Angular error

Known azimuth at last station 273 57' 11.95"

Computed azimuth 273° 56' 49.58"
Angular error 22. 37
Angular correction per station 1. 49

II) Linear error before adjusting azimuths

Known Coordinate X(m) Y(m)
at Turtle 2 325348.292 3076179.297
Computed Coordinate
at Turtle 2 325348.180 3076178.924
Error 6x = + 0. 112 Sy = + 0. 373
Linear error of closure 0. 389 m
Total distances 14853.800 m
Relative error of closure 1 / 38,185

TABLE IX
LATITUDE AND DEPARTURE CORRECTIONS

Traverse lines X- correction (m) Y -correction (m)

1 _ 0. 031 + 0. 007
2 - 0. 035 + 0. 008
3 - 0. 061 + 0. 012
4 - 0. 010 + 0. 002
5 - 0. 069 + 0. 015
6 - 0. 090 + 0. 019
7 - 0. 022 + 0. 005
8 - 0. 022 + 0. 005
9 - 0. 029 + 0. 006
10 - 0. 032 + 0. 007
11 - 0. 018 + 0. 004
12 - 0. 069 + 0. 015
13 - 0. 084 + 0. 018
14 0. 090 + 0. 020

38
 x K
— CO r- CO '£> CT. M0 CM M0 <LT« CO r^- O LO ^ r-
CO ro -O <3< CJ> vO tT> o CO t^ H CP> 00 O cn
<tf i-i CM M0 f>- r- CO CTi H co cr> cr> CO CM CM

lo r^ «* h- LO M0 M0 <* rH CO LO rH <# o CT>

OT^» CO
^
r^ v£> rH CT> cn o r- o LO 00 cr> rH CM t^>
o) g M0 O CO
CO CO 0^ CO rH CO rH CO CO 00 CM
Pv— n ^ M0 LO
oo o o rH M0 c» <_T> a> CT> CM <#
ttJ>H MO vO M0 MO
M0 M0 M0 r^ I> r^ M0 M0
fi o o O O o o o o O o o o o o r- o
r~- r^-

H to CO CO CO co CO CO CO CO CO CO CO CO CO CO
•d

* * -d
o o f- r- r- CO CM 00 co CO M0 <tf CO «* ^ CM a)
u— CM rH LO rH CO CO CO o M0 CO o o CO LO CT> *
- < <# M0 v£> M0 CT» CO CO <tf 00 CO LO o CO CO CM -H
<H
W —K_^
'1
O CO CO CO <* CO rH CO r- LO rH ro rH rH CO
-d O rH CM 0>1 ^ o CO LO iH CO ^ LO M0 rH H 1
TJ
—1 <cx ^ 00 Ol CO M0 CO CM M0 ^ o r- M0 O 00 CO rH
D <T>
CM
CO
CM
co
CM
CO 00 00 f» I> c-~ r- M0 MO M0 «* LO <D
j%/ CM CM CM CM CM CM CM CM CM CM CM CnI rd
CO CO co co CO CO CO co CO CO co CO CO CO CO
00
OO •d
<
CI,
rd

s <* ^ ^
C
2
O a)
"d LO
M0
H
CO t^
"*
r-
M0
M0
CM rH
CO
CO
H<
00 ^
co
r^
LO cn
rH
CO
O
U 3 M0 rH
CTi
CO CO m M0 LO CM rH LO o
CT>
CO CO
cr>
O C
> 4-><-»
X
ca
p—g
•H rH r^ CO rH o CT« CO MO CM CM M0 CM LO cn
X CF> CO LO CM o O M0 CM LO CO o CM o LO CD

tu
OO
J
cd CO r^ CO
rH
rH LO r-«
rH
CM «* LO M0 H 1 <* ^ cn U
1 rH rH rH rH (1)
-J r—
CO ^ ^
<u
<*
y. u CO o O ^ rH <tf LO LO CM CM M0 * o CO LO
H 3 o <* 00 <tf LO ^ M0 CO 00 00 0^ CM CO CO rH
a -p
^^ CO o O M0 LO o O LO ^ CO ^ M0 o CT-

^ g rH LO o 00 rH CM r- O CM CO M0 o MO
T)
C
o K)
CW CO o o r-- *V r-> CM «tf CO
co
<* 00 CT« LO CO rt
o d) LO iH rH rH co o <* CM CO CO LO CM LO
u Q rH rH rH

a 1 1 1 i 1 1 1 1 1 1 1

W
w P a
i— w X 3 cr ) r- r> 3 i H X D JO H JO X JO O
bo •H rH l£ CT> > m: )
3 X n tf "O -O JO H rM * •H
D
(

—1
Q— ro O C 1 CM •o X • * * -O J> -O ^ H o P
(TJ

Q • g h ^ r- M0 LTi — * y> * rj> * :m H H -L-'

< •r->^ O CJi LO rH o H ( 3 X) tf H H ^ X •o in

T> r~- r- co cm -O 3 I -O * JO -«,
* -O X 3
< rH H N H H M
o
.co * >* c 1 CO x IM (
"0 * JO JO o r> H j0 CM in
N lO rH ro -O ^ ( ~o -O H 3 ^ "0 -f rH <D
< P
•s kO M0 M0 CM 3 ^ ( 3 ^ JO X o ^ H T> r^ CCJ

l~l i/) n ^ ^ H -O 1 f) ro -O •N N H M H LO a
-a •H
<Q ro r- io to * > 1
""»
-o X H X r- X
H
.o
h
co -d
o LO CO > M 1 -O 3 :m o sf1 ro r^ 5-i

CO CO CM ~o ro "0 ro o ro ro ro OJ o
o
d u
p iH CM CO «tf LO M0 r-« 00 CT> o rH CM CO -* LO
w rH rH rH rH rH H •K

39
 The minimization of Equation 4.12 can be done by taking partial derivative with

respected to each unknown variable (X)

<P = d<D!dX = 2X T B T WB - 2F T WB = (4.13)

Transposing Equation 4.13 and rearranging yields

(B
T WB)X = B T WF (4.14)

or

NX + U = (4.15)

T WB) T WF. Then

where N = (B is the normal matrix of dimension u by u, and U is B

X = -N _1 U (4.16)

For the adjustment in traverse, the vector V in equation 2.13 is represented the

residual of observed angles and distances. If there are n observed angles in a traverse,

there will be n - 1 observed distances and the number of residuals becomes 2n+ 1

which includes the residual of angles and distances.

There are two types of condition equations in the adjustment of observation
equations: the angle condition and distance condition.
From Figure 4.2 the angle condition can be expressed by

v
ia
= Pi -
(«ij " « lk > <
4 17 )
'

where vj is a residual of observed angle, a-: is a forward azimuth, a-^ is a backward

a
azimuth and 0- is an observed angle. Equation 4.17 is suitable when a- ^ a-^. IF cc-

< a-^, the quantities in parenthesis must be added by 360°.

The Equation 4.17 can be expressed in coordinates of the two stations as

v
ia
= Pi
-
[ tan"
i
1
M—
X. -X.
L).
1 X. -X.
tan-lfcJL—i)]
x '
(4.18)

j i k i

40
 Figure 4.2 Determination of Angle and Distance from Coordinates.

and

HX. — X. X. X. (4.19)
, -
,
-
1 L k
tan" ) - tan' ( r
_' )
Y,k - Y,
i
] '

When linearized by using liquation 2.12, liquation 4.19 yields

= g:(X:\ Xu' Yu) + a,5X:h + + ^3«^j + +

g;
: ~
i Bi^i- Y;', X:', Y;\ M"V
j»^k« 'k
u/ I ,
l
a->5Y:
i
- a,5X: - <M
u,
a45Y:
j

a 6X + a 6Y (4.20)
5 k 6 k

where Xj', Y-', X:', Y:', X^', and YR '

arc the estimated station coordinates. By
substituting the linearization of function g:, Equation 4.18 can be written as

v
ia
+ a,5X + 1
a
2
6Yj 4- a
3
oXj + a 5Yj + a 5 6X R
4
4- a
6
5Y R = U {
(4.21)

where 5X-, 6Y- ^X R , an<^ ^\ are unknown parameters (correction in X and Y

coordinates), coefliccints aj, a a^, and a^ arc the partial derivative of function gj
2
with respected to X- Y XR , and YR , respectively.

For the distance condition, it can be expressed as

2 2 l2
= d .[(X
rX + (Y
rY
[
v )
(4.22)
id [ [
) [
]

41
where

2 2 2
»H- [(X -X i) + (Yj-Yj)
1 /
]
(4.23)
j

The linerized form of Equation 4.23 is then given as

I4 - hi(Xi', Yi', Xj', Y|', X k Yk

', ') + b { 6X + b
2
6Y + b 3 6X + b^Yj (4.24)
{ i
j

Substituting the linearized form from Equation 4.24, Equation 4.22 becomes

v
id
+ bjSXj + b 25Yj + b 5Xj
3
+ b 4 5Yj = F 2 (4.25)

where Fl and F2 represent the constant form for Equations 4.21 and 4.25. Table XI

shows the partial derivatives for Equations 4.21 and 4.25.

Sometimes, the station positions which is determined by intersection from

traverse station is also to be adjusted simultaneously with traverse station positions.

In this case only the angle conditions are added to observation equations of traverse.

In the data adjusted under this thesis, there are 15 observed angles and 14

oberved distances which consisted of 13 unknown points, consequently, there will be 29

observation equations included 26 parameters of §X and 5Y. Thus, they can be

expressed in matrix form of Equation 2.13 as

V 29,l + B
29,26
X 26,1 = F 29,l ( 426 )

where Vj j, v
2 j,
Vj^ j
are the residuals of angles; Vj^ j,
Vjy v are the
j, 2q j

residuals of distances; B is the coefficeint matrix (consistings of a's and b's) of

parameter X (Table XI); and F is the constant vector.

When three intersection points with 10 observed angles were added (Figure 3.1)

to adjust the coordinates, the observation equations have 39 equations including 32

unknown parameters.

V 39,l + B
39,32
X 32,1 = F 39,l < 4 27 )
-

42
 TABLE XI
THE COEFFICIENTS OF ANGLE AND DISTANCE CONDITIONS

3 Yi - Yi Yk - Yk
= -l± = +
2
dX ± (S^) (S ik ) 2
xi - xi xk - xk
a = £ii = +
2
dY ± (s i:j ) (s ik 2 )

a7 =
3
^
axj
Y
= + —>1
Y±
r
z
-

(Sij)
xi - xi
a4 = 0*1. = .
5Y (S~~P"
j
Yk - Y i
a5 = -«Hjl =
2
^ xk (
s
ik>
d xk " xi
a6 = _lL = +
^Y
k <s lk ) a

'1 =
bn —
5h
i H
= + -^
X_- - X,
3,
^X S
i ij

>2 =
bo —
ah,
5Y
i- = +
Y,
->
S
- Y,
^
i ij
5h XH
'3 =
b, —
aXj
! = H
- -J
- X,
i-

SiJ
dh± Yi - Yi
b4 = =

By using Equation 4.16, N is the cofactor matrix and the diagonal terms of this

matrix gives the variances of the adjusted coordinates. To obtain the residuals of all

observed quantities, the reverse process must be done. The correction of X and Y
adjusted coordinates and their standard deviation (cr's) are shown in Table XII. And
the adjusted standard deviation of observed quantities were obtained by multiplying

standard deviation of unit weight [(V WV / r) '

] to a squares root of diagonal

of B(B T WB) B T
_1
element matrix (Table XIII).

43
 ^_^
cm r» o 00 O- in rH m CO CM rH CO CO
— CO
o o
LT)
rH
iH
tH rH
o CO
o
rH
U")
O
yQ
rH
yQ
iH
vD
rH
cn vD
rH
rH
rH
vD

to
o o o o O O O o o o O o O
to ^ o 00 rH ^ CO rH o CO O -* CM i£> <tf
en «* i> vD rH l> yQ O r^ m 00 m ^ rH <tf CM
CQ «* rH CsJ vD CO t> CO (Ti rH CO co CT> CO cn

t>LT) <D vD
t-~ rH
-tf CO m If)O 00 <# rH «*
o vD tH 00 cn O O- O m 00 <T>CM i> cn rH
^ CO vD o CO co rH co 00 co cn CM rH CO 00
co n ^ vD m i> cn cn o o
5^ vd vD ID vD vD vD VD VD ID o t>
<* VD <r> rH CN
O l> [> l>>
>-(o o O o O O O o o o o o O O o
O co CO CO CO CO CO
CO co CO
CO CO CO CO CO co
7: O
o U
o i> CM CO rH
<tfin CM «*IT) CO co <tf
H ^^.^^

co «* vD vD i> 1^ yQ <D o o o 00 m
< 0)*H o o o O o rH rH rH rH rH rH o o
D -Pw
o o o O o O o o o O O o o
3to
•r—
-a o vD CN CO vD o O CM CO CO rH ^ cn co o
O < CM CO [*- tH <* CO CO O vD r^ cn cn i> t"- 00
<# VD vD vd G> co CO ^ 00 CO <* cn CO co tH
H <*•>

eo CO CO ^ CO rH l> in rH r- rH rH CO
< o
CO
tH ^ o
CO
m rH ^ m vD rH <#
> «

X * CO
CNJ
en
CM
CO vD co
CO
CM yo <*
00
O l> yQ O CO co
en CO CO CO 00 CO i> i> r~- r^ yQ vD vD -* m
w CN c\j CM CM CM CM CNJ CM CM CM CM CN CM CM CM
~
i— 55
22
c
co CO co CO CO CO co CO CO CO co CO CO CO CO

X o
CO CO^ m CM LO o rH CM o rH CM rH ^
BY C>i iHCO m LO r- rH rH CM ^ VD l> CO <*
TABLE
OT) o o o O o iH rH rH rH rH iH CM CO
•H
TES 4-> o o o O o O O o o O O o O
U + + + + + + + + + + + + +
< d)
^ cnt> r- CO CM m yQ 00 co m CM cn <*
^x o CO co 00 rH CO CO co CO CM rH ^ CM
Q OT) o o o o rH rH rH rH rH rH rH o O
o o o o O o o o O o O O O
O 1 1 i 1 1 1 1 1 1 1 1 +
O
u VD ^ co o cn (T> <n <# o CO in co r-
CO CM co i-H vD cn m 00 m rH CM 00 rH cn
Q a) tH CM vd l> vD CM r~- o CM r- i> rH t«-
W -P—
H «5£ r- <# h- LO vD yQ <* iH co m rH <# cn
V) C^- i> vd H in en o i> o in CO cn rH rH
D •H>H CO vd O CO CO rH CO 00 CO cn co CO CM
— -a co «* vD o o
LO ^
o- cn cn cn rH CM
Q ^ VD \o <D l> o l>
vD I>
vD yo ID yQ l>
< O o O o o o o o o o o o o
o CO co co CO CO CO CO co CO co co CO co
u
V LO o
CT> en CO «* CO rH CM <D ID 00 cn
*
<0
+J—
«* O o CM eno iH
yQ CO rH o CN
vD t> l> o ^ ^ m o m iD iH <* CO
cue
e~ CO
H
CO co IT) co rH 00 00 in rH 00 rH rH
•hX CM CM «* o CO m rH 00 ^ in vD rH
-p co cn CO vD CO CM yQ ^ o r^ ID O CO
w CO CO CO CO CO l> l> l> t> vD vD VD ^
u CM CM CM CM CM CM CM CM CM CM CN CN CM
CO CO CO CO CO CO CO CO CO CO CO CO cn.
d
p O rH CM CO ^ m
C/l rH CM co <# IT) vD r- co cn iH rH rH rH rH tH

44
 TABLE XIII
ADJUSTED STANDARD DEVIATION OF ANGLES AND DISTANCES

Number Angles Distances

seconds meter
1
2
3
W8. 9
9. 1
0.
0.
0.
029
035
045
4 9.2 0. 024
5 9. 3 0. 048
6 9. 4 0. 056
7 8. 4 0. 030
8 8. 5 0. 029
9 7. 8 0. 032
10 7. 1 0. 034
11 9. 6 0. 028
12 9. 7 0. 048
13 9. 8 0. 054
14 6. 8 0. 058
15 6. 8

45
 V. ANALYSIS OF RESULTS

A. COMPARISON OF ADJUSTED COORDINATES

If the given coordinates of the control points are assumed to be error free, then

the accuracy of the traverse station coordinates depends only on the accuracy of
distance and angle measurements. The adjusted traverse coordinates obtained by the

approximation method are of a lower order of accuracy as only the errors in the

misclosure in azimuths and distances were determined. These errors were distributed
by assuming that all observed quantities had an equal probable occurrence.

The least squares adjustment method provides a better approximation of the true

value. Therefore, the adjusted traverse coordinates obtained by this technique provided

better estimates for position of all traverse stations and the accuracy of the adjustment
can be checked and statistically tested.

After the 13 adjusted traverse station (from stations 2 to 14) coordinates

obtained by the approximation method and by the least squares method were

compared, the difference in coordinates at each station were computed and plotted

(Table XIV and Figure 5.1). The largest difference was at station 14. Because stations
1 and 15 are held fixed, the least squares techniques adjusts simultaneously errors in

azimuths and distances, while the adjustment by approximation adjusts errors in

azimuths and distances sequentially. Consequently, the largest difference in traverse

distances occurs at the last station (station 14) before closing of traverse at the fixed

station.

When the standard deviation of observed quantities before the adjustment

(Tables III and IV) were compared with those obtained through adjustment (Table
XIII), the standard deviation of all observed quantities in Table XIII showed

increments. That means, the estimated standard deviations were optimistic.

In this thesis, the three intersection points were also adjusted. The adjusted
coordinates of these were compared to NOS results. The largest difference occurs at

point no. 3 (5x = + 0.140 m; Sy = - 0.158 m). The standard deviation of adjusted
coordinates at this point are crx
3
= ± 1.87, ay 3 = ± 0.96 m (Table XV).

46
 TABLE XIV
COMPARISON OF ADJUSTED COORDINATES/DISTANCES
OBTAINED BY APPROXIMATION AND LEAST SQUARES METHODS

Differences*
a tes-

Distances
stn. dx ?y*
_
m (m) (m)
2 0l9 mm
0. 007 0.020
3 - 0. 015 - 0. 017 0.023
4 + 0. 004 - 0. 022 0.022
5 - 0. 013 - 0. 012 0.018
6 + 0. 002 - 0. 008 0.008
7 + 0. 008 + 0. 024 0.025
8 + 0. 001 - 0. 005 0.005
9 + 0. 005 - 0. 013 0. 014
10 - 0. 008 - 0. 023 0. 024
11 - 0. 003 - 0. 034 0. 034
12 - 0. 014 + 0. 003 0. 014
13 - 0. 005 + 0. 041 0. 041
14 - 0. 019 - 0. 101 0. 103
* Approximation minus least squares solution

B. ANALYSIS OF THE REFERENCE VARIANCE OF UNIT WEIGHT

The weight matrix was set for the least squares adjustment by using Equation

2.1 1. The (Tq (a priori reference variance of unit weight) was assumed as 1. The
result of (Tq^ = (VHVV / r) was obtained after adjustment. The value of (Tq (a

posteriori reference variance of unit weight) can be used to evaluate the weighting

scheme used in the least squares adjustment.

The standard deviation of the observed angles ((7

a)
and distances (<y^) used in the

least squares adjustments (solutions 1, 2, and 3) and the corresponding a posteriori <Tq

obtained for these solutions are listed in Table XVI. As the a posteriori <Tq for the

solution no. 3 is closest to the assumed a priori Gq (= 1), the weight (or the standard

deviations) used for observed angles and distances in this case seem the most realistic.

Further statistical testing for (Tq^ done by Chi-squares or F-test was not carried out
under this thesis.

47
Figure5.1 Comparison of Adjusted Distances Obtained
by Approximation and Least Squares Methods.

48
 -
TABLE XV
COMPARISON OF COORDINATES AT INTERSECTION POINTS

X(m) dx( ±m) Y(m) <yy( ±m)

At Point No. 1
NOS 326616. 356 3071294. 683
Adjusted 326616. 470 1. 72 3071294. 635 1. 03
Sx - 0. 114 Sy + 0. 048
At Point No. 2
NOS 327056. 079 3070340. 474
Adjusted 327056. 183 1. 68 3070340. 444 1. 14
6x - 0. 104 Sy + 0. 030
At Point No. 3
NOS 325378. 101 3077263. 378
Adjusted 325377. 961 1. 87 3077263. 536 0. 96
Sx + 0. 140 Sy - 0. 158

TABLE XVI
COMPARISION OF VARIANCES OF UNIT WEIGHT

Solutions Variance used in Adj us tment V T WV/r

2 ^ 2
*a *d °o °0
1 2'* ( lOppm + 0. 5 cm) 1 15. 08
2 5" ( lOppm + 1. cm) 1 3. 98
3 10" ( lOppm + 2. cm) 1 1. 32

49
 VI. CONCLUSIONS AND RECOMMENDATIONS

A. CONCLUSIONS
By using the weight of observed quantities for the adjustment, the traverse

station coordinates computed and adjusted by the least squares observation equations

method were more accurate than those obtained by the approximation methods. Even
though the observation equations method may require a greater number of equations
than the condition equations method, the processing of the data for adjustment is

easier and the corrections in X and Y coordinates can be directly obtained through

iterative solution. This method is suitable when a computer with a memory capacity of
over 500 K bytes is available. However, for local work or a relatively short traverse, an

approximation method is commonly utilized when economic and logistic criteria are

considered.

B. RECOMMENDATION
The INDTRA Fortran program written for this thesis is automated for handling

only two kinds of survey techniques: traversing and intersection. With the computers
available at NPS, the development of adjustment programs for covering a wide range

of survey techniques should be done to use and continue analysis of the mixed kind of

survey techniques including traverse, triangulation, trilateration, resection, and

intersection.

50
 APPENDIX A
LINEARIZATIONS

This is a linearized form of m functions in n unknowns.

Yl = ^(xp x
2
,..., x
n)
y2 = f ( Xl ,x ,...,x )
2 2 n

-m =
V f (x,,
v
m1'
Xo,..-, x_)7
Z' n '

^(Xj , x , ..., x )
2 n
°
y2 2 2 '
2 '
'"'
n '

Y° =

° ' u
f (x X ° X )
m

_dy,
gyj gyj
ax, ax 2 axr
dY
yx ax

ax, ay- ax

Ax,
Ax 2
AX =

Ax,

The general form of linearized functions becomes

Y = Y° + J
yx
AX

51
 APPENDIX B
TRAVADJ FORTRAN PROGRAM

This program is used for computing and adjusting the traverse station position by

approximation method (Compass rule).

CCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCC
c c
C FORTRAN PROGRAM "TRAVADJ" C
c c
C THIS PROGRAM IS USED FOR C
C 1) REDUCE SLOPE DISTANCE TO ELLIPSOIDAL DISTANCE C
C 2) DETERMINE GRID DISTANCE C
C 3) COMPUTE CLOSE TRAVERSE C
C 4) ADJUST COORDINATE BY COMPASS RULE C
c c
ccccccccccccccccccccccccccccccccccccccccccccccccccccccc
c
c
ccccccccccccccccccccccccccccccccccccccccccccccccccccccc
c c
C INPUT DATA C
C 1) SEMIMAJOR AXIS OF REFERENCE ELLIPSOID USED C
c "a" c
C 2) VALUE OF 1/F (EX. 1/F = 294.978698 C
c "f" c
C 3) CENTRAL SCALE FACTOR C
c ''ko" c
C 4) AT STARTING AND CLOSING POINT
LAT. C
c "idl,iml,sl.id2,im2,s2" c
C 5) GRID N. AND £. OF 4 ftNOWN STATION C
c "qnl ,qel ,qn2,qe2,qn3 ,qe3 ,qn4,qn4" c
C 6) NUHBEft OF MEAGRE DISTANCF. C
c "n" c
C 7) ELEVATJON AT FIRST OCCUPIED POINT C
8
C 8) NAME OF ALL STATIONS C
C 9) INDICATOR VALUE= NO VERTICAL ANGLE
1 C
C 2 = VERTICAL ANGLE C
C 10) DIFFERENT IN ELEVATION BETWEEN TWO STATIONS C
C 11) SLOPE DISTANCE (WITH UNIT FEET OR METER) C
c "dist" c
C 12) HORIZONTAL ANGLES C
C C
CCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCC
c
c
ccccccccccccccccccccccccccccccccccccccccccccccccccccccc
c c
c sfac = scale factor coorection
c hgd = horizontal distance c
c todis= total distance in traverse c
c cegrid = grid east of traverse station c
c cnqrid = grid north of traverse station c
c difazi = azimuth misclosure c
c difdis = distance misclosure c
c coraz = angular correction per station c
ccccccccccccccccccccccccccccccccccccccccccccccccccccccc
c c
c gridaz = subroutine for computing azimuth c

52
c between two traverse stations c
c utm = subroutine for computing the grid c
c "coordinates from known distance c
c and azimuth c
c dmsr = subroutine for converting the angle c
c from degrees, minutes, and seconds to c
c radians c
c rdms = subroutine for converting the angle c
c from radians to degrees, minutes, and c
c seconds c
c c
ccccc cccccccccccccccccccccccccccccccccccccccccccccccccc
c
c
DIMENSION LAT(3).K(3),HGD(30),HAGL(30,,HANG(30)
DIMENSION DHAGL( 30),NAME(3,30),NAME1(5)
DIMENSION NAME3(2),NAME4(5),MDIST(30)

VARIABLE DECLARATION
DOUBLE PRECISION AN90R,AN180R,AN360R.AZC,AZFIR,AZIMUT o AZLAS,
CEGRHJ(301,C(JAZR(30l,C0^AZ^CN^RID(30J,
DISTAN,DIFAZI,DELTAX(30],DELTAY(30),DUMMY1,
DUMMY2,DUMDIS,DIFDIS,T0DIS
DOUBLE PRECISION NUMO NUM360 NUM1 SUMDX SUMDY STDD( 30 ) STDA( 30)
, . , . . ,

XGD,YGD,NUMf80,Nl)M90,Wto(3J
DOUBLE PRECISION ANGS(30),CFAZSi(30) j DUM1D,DUM1M,DUM1S
,EAZis,EAZ2S,C0RDXl,C0RDYl,TEMS\AZS21,AZS34
DOUBLE PRECISION sumix1,sumix2,sumfy,sunfx,sumc0y,sumc0x,
s$ta,ss^d.sigx\sigy.sigxy'semaj.$emin,
SETA, SETAA.SSlfiX.SSiGY.SSiGXY.DEGl, MINI.
SECON1,SEMAJ1,SEMIN1,TEMP01(30),TEMP02(30),
DDIFFXiDDIFFY
REAL*8 A A C,D,H A R,V,X A Y.GD A DH A ZD A CON.CUV A LAT A PHE,CUVRE,ELEV
REAL*8 HfJlS^T' SDlS^T A M!!)I^T A ESQft ,GftAMMA. AN6 A AN6V A Pi ,AGU Fl F2 , ,

REAL*8 K,KO.El,E2.0l,Q2,ftR,VV ECEN.FAC.SFAC.HfiD

REAL*8 F,CH6D A CSDiST,DDM A HH A HAGL,CHDISt,DG.[iMS.DS A DDANG,MMANG
REAL*8 GNl A GN2\GN3,GN4,GEl,6E2 A GE3.GE4 A AZ2i,AZ34.rjIST21,DIST34
REAL*8 SI §2 Dl D2 SV VANG SHA^L SANG t)D MM SS HANG RRR
, . . , , , , , , ,

REAL*8 DUM1,6UM2,DUM3 DUM4 XXXX,YYYY

REAL*8 IDD1 IMM1 IDD2 IMM2 DVV,MVV,CUVRER,CUVR,CURVR
REAL*8 ZINE,K0SE
INTEGER I N.ID1.ID2.IM1.IM2.DV,MV.DHAGL.MHAGL,DANG.MANG,UNIT
1

INTEGER ANGD730] A AN6M(36) A AD\Jl A ERR,AZD2i,AZM2i A AZD34,AZM34

INTEGER SSHAGLl3d>.CFAZDnO).CFAZM(30J.MtN2,DE62,
EAZ1D EAZlrt EAZ2D EAZ2M TEMD TEMM
, , , , ,

DATA DUM1/0. 30480061D0/ DUM2/500000. 0D0/,DUM3/0. 000001D0/

,

DATA DUM4/3600. 0D0/,SIGANG/5. 0D0/.DUM5/0. 000005D0/,DUM6/0. 005D0/

DATA TODIS/0. 0D0/,DUM1070. 0000048481368D0/

DATA NUMO/0. 0D0/,NUM360/360. 0D0/.NUM1/1. 0D0/.SUMDX/0. 0D0/,

SUMDY/0. 0D0/,NUM180/180. 0D0/,NUM90/90. OdO/
READ (5,10 A A F,,K0
READ 5 15 ID1 IM1.S1
READ 5 15 ID2, IM2.S2
READ 5 16 GN1, GE1
READ 5,16 GN2,,GE2
READ 5 16 GN3 GE3
READ (5,16 GN4,,GE4

53
 READ(5,11)N,ELEV

CALL GRIDAZ (GE1 ,GN1,GE2,GN2,AZ21 ,DIST21)

CALL GRIDAZ (GE3;GN3;GE4,GN4,AZ34,DIST34)
IDD1 = FLOAT(IDl)
IMM1 = FLOATf IM1)
CALL DMSR (IDD1 IMM1 ,Sl ,D1)
.

LATfl) =6l
IDD2 = FLOAT(ID2)
IMM2 = FLOATf IM2]
CALL DMSR (IDD2, IMM2,S2,D2)
LAT(2) =02
LAT[3) = (Dl+D2)/2.0D0
ESQR
iQR = 2.0D0*(l.0D0/FJ-(1.0D0/F)**2
X = A*DSQRTf 1. ODO-ESQR)
Y = 1.0D0-(ESQR*(DSIN(LAT(3)))**2)
K — X/
C
c
C DETERMINATION OF THE SCALE FACTOR FOR UTM.
C
C
C
El = DABS(DUM2-GE2)
E2 = DABS(DUM2-GE3)
C
ECEN = ESQR/(1. ODO-ESQR)
c
01 = DUM3*E1
02 = DUM3*E2
QPRIME = ((Q1**2)+(Q1*Q2)+(Q2**2))/3.0D0
c
DOl M=l,3
RR=A/(1.0D0-ESQR*(DSIN(LAT(M)))**2)**0. 5

Fl = fl.0D0+ECEN*DC0S(LAT(M)))*(10.0D0**12)
F2 = 2.0D0*(RR**2)*(KO**2)
FAC= F1/F2
C
C
KfMj)=KO*(l. ODO+FAC*QPRIME+(0. 00003D0*(QPRIME**2)))

C
SFAC = K(3)
C SFAC = 6. ODO/( ( 1. 0D0/K( 1 ) )+( 4. 0D0/K(3))+(1. 0D0/K(2)))
c
READ(5,20) UNIT
-'Mr
READ(5,18)(NAME1(L),L=1,5)
C DETERMINATION OF THE HORIZONTAL DISTANCES
C
DO 1000 J=1,N

READ(5,14)(NAME(L,J),L=1,3)
READf5 12jI.DH.DV MV SV
READ[5,13) ^DI^T
MDIST(J) = SDIST
C
C IN CASE OF THE LENGTH'S UNIT IS IN FEET, THEN CONVERSE TO METER

IF(UNIT. EQ. 1) THEN

SDIST = SDIST*DUM1

54
 DH = DH*DUM1
ELEV = ELEV*DUM1
END 4F
C
READ(5,17)DHAGL(J),MHAGL(J),SHAGL(J)
DD = DFLOATfDHAGLfJ))
MM = DFLOATfMHAGL(J))
SS = SHAGL(J)

CALL DMSR ( DD,MM^SS,RRR )

HANG(J) = R&R
STDA(J) = SIGANG
C
c
c
C DETERMINATION OF HORIZONTAL DISTANCES WHEN THE DIFFERENT IN
C ELEVATION IS APPROXIMATELY KNOWN.
C
IF(I.EQ.1)THEN
IFCDH.NE. O.OJTHEN
DDM = SDIST
IFCDDM.GE. 3300.0)THEN
DO 100 KK=1,3
CURV =(0.016192D0*DDM)
CALL DMSR (NUM0,NUM0.CURV,CURVR)
D=DSQRT(DDM**2-(DH*D£0S(CURVR)r*2)-DH*DSIN(CURVR)
C
100 CONTINUE
C
HDIST = DDM
ELSE
HDIST = DSQRT(DDM**2-DH**2)
END IF
ELSE
HDIST = SDIST
END IF
END IF
C
C DETERMINATION OF HORIZONTAL DISTANCE WHEN ZENITH DISTANCE IS KNOWN
C
IF(I.EQ. 2)THEN
DVV = FLOAT(DV)
MVV = FLOAT(MV)
CALL DMSR (DVV'MVV,SV,VANG)
CALL DMSR (NUM§0 a NUM0'NUM0'AN90R)
CALL DMSR (NUM180\NUM6,NUMO\AN180R)
ZD = ANQOR^VANG
DH = SDIST^DSINCZD'
ds = sdist*dcos(zd;
IFf SDIST. GE. 3300.01 THEN
CUVRE = 0. 013925TJ0*SDIST
CUV = 0.016192D0*SDIST
CALL DMSR (NUM0,NUM0,CUVRE,CUVRER)
CALL DMSR (NUMO'NUMO CUV.CUVR)
PHE = (AN90R-ZD)+CUVRER
C = (AN90R+CUVR)
GRAMMA = AN180R-(C+PHE]
HDIST = (SDIST*DSIN(GRAMMA))/DSIN(C)
ELSE
HDIST = DS
END IF
END IF

55
 HH= (ELEV+DH)
C
C DETERMINATION OF THE HORIZONTAL DISTANCE ON GEOID
C
GD = R*HDIST/(R+H)
C
C DETERMINATION OF THE GRID DISTANCES
HGD(J) = GD*SFAC

TODIS = TODIS+HGDfJ)
STDD(J) = HGD(J)*DUM5+DUM6
WRITE(6,*) 'GRID DISTANCE =' ,HGD(J)
WRITE(6;*)' Y
ELEV = HH
1000 CONTINUE
READ(5,19)(NAME3(L),L=1,2),DANG,MANG,SANG
READ(5;18)(NAME4(L);L=1;5)
CALL 0UTPUT(N,AZ21.AZ34.GN2,GE2,GN3.GE3,HGD,DHAGL,MHAGL,SHAGL,
*DANG.MANG,SANd NAM^l.NAME3,^AME4,NAME,MdlST,GNl,G^l,GN4,GE4)
>
dDANG = DFLOAT(DXNG)
MMANG = DFLOAT(MANG)
CALL DMSR (DDANG, MMANG, SANG, RRR )
NN = N+l
HANG(NN) = RRR
STDA(NN) = SIGANG

CALL DMSR ( NUM360,NUM0,NUM0,AN360R )

CALL DMSR ( NUM180,NUM0,NUM0,AN180R )

DO 2000 ADJ1 = 1,2

XGD = GE2
YGD = GN2
AZFIR = AZ21
DO 200 I = 1,N
AZIMUT = AZFIR+HANG(I)
IF (AZIMUT. GE.AN360R) THEN
AZIMUT = AZIMUT-AN360R
END IF
IF(AZIMUT.GE.AN180R) THEN
AZIMUT = AZIMUT-AM80R
ELSE
AZIMUT = AZIMUT+AN180R
END IF
COAZR(I) = AZIMUT

CALL RDMS ( AZIMUT,DUM1D,DUM1M,DUM1S )

CFAZD(I) = l6lNT(DClMlD)
CFAZM(I) = IDINT(DUMIM)
CFAZS(I) = DUM1S

DISTAN = HGD(I)
CALL UTM ( XGD, YGD, DISTAN, AZIMUT, DUMMY1 ,DUMMY2 )

WRITE(6,*)'I =' ,1

56
 WRITE(6,*)'GRID E=' .DUMMYI GRID N=',DUMMY2
'

DELTAXd' = DUMMYl-XGd
.

DELTAYf Ii = DUMMY2-YGD
WRITE(6 *i'DEPARTure =' + DELTAX(I), LATITUde =',DELTAY(I)
ZlNE = DSIN(AZIMUt)
KOSE = DCOS(AZIMUT)
SUMDX = SUMQX+DE|_IAX(
= SUMDY+DELTAY(I
n
SUMDY
CEGRID(I) = DUMMYI
CNGRID(I) = DUMMY2
CALL GRIDAZ ( XGD, YGD, DUMMYI, DUMMY2,AZC,DUMDIS )

CALL ROMS ( AZC,DUM1D,DUM1M,DUM1S )

CBAZD(I) = IDINTfDUMID*
CBAZM(I) = IDINT(DUM1M'
CBAZS(I) = DUM1S

IF (ADJ1.EQ. 1 THEN
TEMPOKI
TEMP02(I) = CNGRID
Dm
END IF
AZFIR = AZC
XGD = DUMMYI
YGD = DUMMY2
200 CONTINUE
IF (ADJ1.NE. 1) THEN

CORDX1 = ( SUMDX-(GE3-GE2) )/TODIS

CORDY1 = f SUMDY-(GN3-GN2) )/TODIS
XGD = GE2
YGD = GN2

DO 230 1=1,
DISTAN = HGD(I)
XXXX = C0RDX1*HGD(I)
YYYY = CORDYl*HGDm
DELTAX(I) = DELTAX(I) - XXXX
DELTAY(I) = DELTAY(I) - YYYY

CEGRID(I) XGD+ DELTAX(I

CNGRID(I) YGD+ DELTAYfT
DDIFFX CEGRIDfl)-TEMPOl(I)
DDIFFY CNGRIDC I )-TEMP02( I

XGD = CEGRIDtn
YGD = CNGRID(I
230 CONTINUE

END IF
AZLAS = AZFIR+HANG(NN)
IF ( AZLAS. GE.AN360R ) THEN
AZLAS = AZLAS-AN360R
END IF
IF (AZLAS. GT.AN180R) THEN

57
 AZLAS = AZLAS-AN180R
ELSE
AZLAS = AZLAS+AN180R
END IF
C
CALL RDMS ( AZLAS,DUM1D,DUM1M,DUM1S )
TEMD = IDINt(DUMlD)
TEMM = IDINT(DUMIM)
TEMS = DUM1S
C
C
DIFAZI = AZLAS-AZ34
WPRED(l) = DIFAZI
CORAZ = DIFAZI/DFLOAT(NN)

CALL RDMS ( DIFAZI. DUM1D.DUM1M.DUM1S )

EAZ1D = IDINt(DABS(DUMlD))
EAZ1M = IDINTfDABS(DUMlM))
EAZ1S = DABS(DUMIS)

CALL RDMS ( CORAZ ,DUM1D.DUM1M,DUM1S )

EAZ2D = IDINT(DAB$(DUM1D))
EAZ2M = IDINT[DABS(DUM1M))
EAZ2S = DABS(DUMIS)

IF(ADJ1.EQ. 1)THEN
WPRED(2) = SUMDX - ( GE3-GE2 )
WPREDC3) = SUMDY - f GN3-GN2 )
DIFDIS = DSQRT( WPRED(2)**2 + WPRED(3)**2 )
ERR = IDINTC TODIS / DIFDIS )
END IF
C
C
C PRINT RESULTS OF TRAVERSE COMPUTATION
C
C
CALL RDMS (AZ21.DUM1D.DUM1M.DUM1S)
AZD21 = iDINTffiUMID)
AZM21 = IDINT(DUMIM)
AZS21 = DUM1S
CALL RDMS (AZ34.DUM1D.DUM1M.DUM1S)
AZD34 = iDINTfDUMID)
AZM34 = IDINT(DUMIM)
AZS34 = DUM1S
IF (ADJ1.EQ. 1) THEN
WRITE(8,29)
WRITE(8 27
WRITE(8,30'
WRITE(8 31"
ELSE
WRITE(8,28'
WRITE(8 27'
WRITE(8 30"
WRITE(8 40
END IF
WRITE(8,32'
WRITE(8,33'
WRITEC8 30"
WRITE(8,34) AZD21,AZM21,AZS21
WRITE(8;35) GE2.GN2
DO 240 1=1,
II = 1+1
CALL RDMS[HANG(I),DUM1D,DUM1M,DUM1S)
ANGD(I) = IDINT(DUMID)
ANGMflj = IDINT(DUMIM)
ANGSfl) = DUM1S
SSHAGL(I) = IDINT(SHAGL(I))

58
 IF (ADJ1.EQ. 1) THEN
WlTE(8.3^^
W
I^^I fc
g^L(I).CFAZD(I).

WRITE(8,37^ANGDn) ANGMri),ANGS(I),CFAZD(I),CFAZM(I),
^
END IF
WRITE(8,38) II,CEGRID(I),CNGRID(I)
240 CONTINUE
c
CALL RDMS ( HANG(NN) ,DUM1D,DUM1M,DUM1S)
IF (ADJ1.EQ. 1) THEIsl
DHAGL(NN) = DANG
MHAGL(NN) = MANG
SSHAGL(NN) = IDINT(SANG)
WRITE(8,36)DHAGL(NN),MHAGL(NN),SSHAGL(NN),TEMD,TEMM,TEMS
CALL RDMS (HANG(NN).DUM1D,DUM1M,DUM1S)
ANGD(NN) = IDlNKDUMip
IDINT(DUMID)
ANGMfNN) = IDINT(DUMIM)
ANGS(NN) = DUM1S
WRITE(8,37)ANGD(NN)
ITE( ,ANGM(NN) ,ANGS(NN) ,TEMD,TEMM,TEMS
END IF
WRITE(8,30)
-1)
WRITE(8,39)AZD34,AZM34,AZS34,GE3,GN3
V
WRITEC8 *) '

IF (ADJl.EQ. 1J THEN
WRITE(8,4I)EAZ1D,EAZ1M,EAZ1S
WRITEf 8 42 1 EAZ2D EAZ2M EAZ2S
, , ,

WRITEC8 43JT0DIS
WRITEC8 44]DIFDIS
WRITE(8,45) ERR
END IF
C
C CORRECTED OBSERVED ANGLES
C
DO 250 1=1, NN
HANG(I) = HANG(I)-CORAZ
250 CONTINUE
c
SUMDX = O.ODO
SUMDY = O.ODO
C
C
2000 CONTINUE
C

SUMDX = GE3-CEGRID(N)
SUMDY = GN3-CNGRIDCN)
DIFDIS= DSQRT ( SUMDX**2 + SUMDY**2 )

EAZ1D.EAZ1M.EAZ1S
WRITE(8,41) EAZ1D.
WRITE(8,44) DIFDI$
10 FORMAT(5X,F15. 5 4 5X^F15, 10, 15X.F12. 8)
11 F0RMAT(26X,I4,3^X,F7.3)
12 F0RMAT(20X II !4X,F9. 3, 11X,2I3, F10. 4)
13 F0RMAT(25X'F15\ 6)
14 F0RMAT(6X,3A4)
15 F0RMAT(5X,2I3'f14. 8)
16 FORMATflO^.FlS. 5,15X^15. 5)
17 FCRMAT(25X,2I3,Fl0.4)
18 FORMAT? 5A4)
19 FORMAT(2A4,17X,2I3,F10. 4)
20 FORMAT(H)
C
27 FORMAT( 16X ,':::::::::::::::::::::::::::::' ,///)

59
 28 F0RMAT(16X,' ADJUSTED TRAVERSE COMPUTATION') x
29 F0RMATQ5X 'UNADJUSTED TRAVERSE COMPUTATION')
FORMAT( --
r
30
* i \

31 FORMAT STN OBSERVED FORWARD GRID COORDINATE (M

40 FORMAT TN CORR. OBS. FORWARD GRID ADJUSTED COORDINA
*TE(M1 )
32 FORMAT HOR. ANG. AZIMUTHS DIST.

33 FORMAT( 0. D M S D M S (M.) GRID EAST GRID :

*NORTH
34 FORMAT( ,13X,I3,1X,I2,1X.F5.2,34X I')
'

35 FORMATf'I 1. '36X:Fl6.3!lX!Fli.3'lVlh
36 F0RMAT('I',3X,' Y ,i3,lX,i2,iX,I2,4x,i3,lX,I2,lX,F5.2,lX,F8. 3.25X
37 F6RMATC 'I' ,3X, ',13, IX, 12, IX, F5. 2, IX, 13, IX, 12, IX, F5. 2, IX, F8. 3,
*25X I '
'

38 FORMATfl' 13 '

36X,F10.3,1X,F11.3,1X,T)
C39 FORMATflX, ^KNdWN dATA p ,7X;iJ,l)(,I2!l)(.Fl2,ilX,F10.
2,11X, 3,1X,F11.3,//)
C41 FORMATfWANGULAR ERROR 1 '

= f ,2x, 13, IX, D' IX, 12, IX, 'M' IX, F5. ' , ,

C *2 IX S ' )
C42 FdRMAjr; ANGULAR CORR./ STATION =' ,2X, 13, IX, 'D' , IX, 12, IX, 'M' , IX, F5.
C *2
C43 f6rmAt 'TOTAL DISTANCES
C44 FORMAT 'LINEAR ERROR = 3}
2X F10.3)
C45 FORMAT 'LINEAR CLOSURE = ;3x;'i / r ,i6,///////////)
STOP
END
CCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCC
c c
C SUBROUTINE TO CALCULATE GRID AZIMUTH AND DISTANCE BETWEEN C
C TWO KNOWN STATIONS WITH THEIRS GRID NORTH AND GRID EAST. C
c c
cccccccccccccccccccccccccccccccccccccccccccccccccccccccccccccc
c
SUBROUTINE GRIDAZ (X1,Y1,X2,Y2,ANGLR,DIST12)
c
REAL*8 X1.X2.Y1,Y2.ANGLR.ANG0,ANG90,ANG180,ANG270,DIFX,DIFY,DIST12
REAL*8 ANG90ft,AN180R,AN270R
DATA ANGO/0. 0D0/.ANG90/90. 0D0/,ANG180/180. 0D0/,ANG270/270. ODO/
CALL DMSR ANG90,ANG0.ANG0 A ANG90R )
CALL DMSR ANG186,ANG6,ANG0,AN180R )
CALL DMSR ANG270,ANG0,ANG0,AN270R )
DIFX = X2-X1
DIFY = Y2-Y1
IF((DIFX. EQ. ANGO). AND. (DIFY. EQ. ANGO)) THEN
ANGLR = ANGO
ELSE IFf DIFX. EQ. ANGO) THEN
IF(DIFY.GT.ANGO) THEN
ANGLR = ANGO
END IF
ANGLR = AN180R
ELSE
IF(DIFX.GT.ANGOJ THEN
ANGLR = ANG90T<-DATAN(DIFY/DIFX)
v '
ELSE
ANGLR = AN270R-DATAN(DIFY/DIFX)
END IF
END IF

DIST12 = DSQRT(DIFX**2+DIFY**2)

60
 RETURN
END

(3(3(3(3(3(3(3(3(9(3(3(3(3(3(3(3(3(3(3(3(3(3(^
(3

(3 SUBROUTINE OUTPUT TO PRINT OUT THE RESULT FOR USING IN (3

(3 THE NEXT COMPUTATION. (3
(3 (3
(3(3(3(3(3(3(3(3(9(3(3(30(3(3(3(3(3(9(3(30^

SUBROUTINE OUTPUT(N.A.B.G2.E2.G3.E3,RD,DH,MH,SH,DA,MA,SA,NAMEl,
*NAME3 NAME4 NAME MDf Sf ,(h £l Gl4 £4)
, , , , , ,

DIMENSION RD(N),DH(N),MH(N),SH(N),NAME(3,N),NAME1(5),NAME3( 2)
,NAME4(5),MDlSTfN)
REAL*8 Al6.DG,G2 A G3 A E2.E3.RD.AA.BB,HH.SH,SA,SSA,SIGA,SIGD
REAL*8 MDI$T,MDIS\RRD,Gl,G4,F.l,£4,DUMi
INTEGER N,DH,MH,DA,MA
DATA SIGA/5. 0D0/,DUMl/0. 30480061D0/
WRITE(7 ,240'
WRITE(7 250'
WRITE(7 ,260'
WRITE(7 250'
WRITE(7 270'
WRITE(7 280'
WRITE( 7 290'
WRITE! 7 280'
WRITE! 7 240'
WRITE! 7 280'
WRITE! 7 300' NAME1(L),L=1.5).G1.E1
WRITE 7 ,310' 'NAME(Ll5,L=i.3).G2.E2
WRITE( 7 310' NAME3(L),L=1,3),G3,£3
WRITE! 7 300' NAME4(L) L=1,5),G4,E4
WRITE! 7 ,280'
WRITE! 240'

WRITE (6 230)
WRITE 6 100)
WRITE 6 110)
WRITE 6 120)
WRITE 6 110)
WRITE 6 125)
WRITE 6 140)
WRITE 6 130)
WRITE
WRITE (6
U 140)
100)
WRITE (6 140)

Nl = N+l

WRlfEC'6,150)(NAME(L,I-l),L=l,2),(NAME(L,I),L=l,2),DH(I-l),MH(r
* l).SH(l-l)
10
WRITE(6,150)(NAMEfL,N),L=1.2)XNAME3(L),L=1.2).DHfN).MH(N) SH(N)
>

WRITE 6 150 (NAME3(L) L=l ,2) ,(NAME4(L L=l,2) DA, MA $A

WRITE(6 140)
WRITE(6,100)

WRITE(6,230)
WRITE(9;i70)

61
 WRITE (9 180)
WRITE 9 490)
WRITE 9 180)
WRITE 9 170)
WRITE 9 200)
WRITE 9 210)
WRITE 9 200)
WRITE 9 170)
WRITE 9 200)
DO 20 J=1,N-1
mdis = md:
MDIS MDISTf J)*DUM1
RRD = RD(J) 7D0M1
SIGD = (RD(J)*0.00001D0)+0.01D0
WRITE(9 220)(nAME(L
V J),L=1,2),(NAME(L,J+1),L=1,2),MDIS,
)

MDIST(b) RD(5) RRD S

20 CONTINUE
MDIS = MDIST(N)*DUM1
RRD = RDCNVDOMl
WRITE(9,220)(NAME(L,N),L=1,2),(NAME3(L),L=1,2),MDIS,MDIST(N)
* RD(N)\tfRD
SIGD = (RD(N)*0.00001D0)+0.01D0
WRITE(9,200r
WRITE(9,170:
C
C
40 F0RMAT(2I3,F6.2)
100 FORMAT 9X ****************** *************** **************** *******
(

* * *
110 F0RMAT£9X,'*
120 FORMAT J^X,'* MEASURED HORIZONTAL ANGLE * * * %

125 F0RMAT(9X '******************************* DEGREES * MINUTES * SEC

*ONDS ™y
130 FORMAT £?X,'* AT TO
*
140 FORMAT^X, 1

* *
150 F0RMAT(9x\ ,

*'*' .2X.FS. 2,5X, '*•

*\3X f
2A4,3X, *',3X,2A4,3X,
, , ,

,3X,I3,3X,
, ,

,4X,I2,3X,
170
*########to#### (

)
180 F0RMAT(5X,'# # #
#')
190 F0RMAT(5X '# DISTANCES # MEASURED DISTANCES # REDUCE
r
*D DISTANCES # )
200 F0RMAT(5X/# # # # #
* # # )
210 F0RMAT(5X '# FROM # TO # METERS § FEET # METERS
*
# FE£T #Y )
220 F0RMAT(5X '#' ,2A4; '#' 2A4 ,'#\2X,F8. 3,2X, '#' ,2X,F8. 2,2X,'# ,2X,
.
.
I

*F8.3.1X ,'r ,2XF8\2,2X,'£')

230 FORMAt('i')
240 FDRMATiQX *******************************************************
* *
250 F0RMAT(9X,
260 F0RMAT(9X, NAME COORDINATES
270 F0RMAT(9X, OF *********************************
* * *
280 FO RMAT(9X,

290 F0RMAT(9X, KNOWN POSITIONS *

GRID NORTH * nDTn
GRID CACT
EAST *
* \

300 F6RMAT(9X,'*',5A4 '*' .2X.F11.3.2X '*' .2X.F11. 3,2X. '*'

310 FORMAThx, *' 4X,5A4,4x, ut ,2X Fli. 3,2X, '*' ,2X Fll. 3,2X,
1
'*'
RETURN

62
 END
C
C
C 0000000000000000000000000000000000
C
C CONVERT DEGREES, MINUTES,
C (3 AND SECONDS TO ftADIANS (3
C (3

C 0(3(3(3(3(3(3(3(3(3(3(3(3(3(30(3(3(3(3(3(3(300(3(3(3(3(3(3(3(30
C
c
SUBROUTINE DMSR ( D1,M1,S1,R1 )

REAL*8 D1,M1,S1,R1,PI1,N60,N3600,N180
DATA N60/60. 0D0/,N3600/3600. 0D0/,N180/180. ODO/
PI1 = 4. ODO*DATAN(1.0DO)
Rl = ((D1+(M1/N60)+(S1/N3600))*PI1)/N180

RETURN
END
C
C
C 000000000000000000000000000000000
C
C COMPUTE THE COORDINATES
C FROM ONE KNOWN POINT WITH
C MEASURED DISTANCE AND ANGLE
C
C 000000000000000000000000000000000
C
SUBROUTINE UTM (XI Yl ,DIS12 ANG12R X2 Y2
, , , ,
)

REAL*8 X1,Y1,X2,Y2,DIS12,ANG12R,DIFX,DIFY
DIFX = DIS12*(DSIN(ANG12R))
DIFY = DIS12*(DCOS(ANG12R))
X2 = Xl+DIFX
Y2 = Yl+DIFY
C
RETURN
END
C
C 000000000000000000000000000000000
C
C CONVERT ANGLE FROM RADIAN
C TO DEGREE, MINUTES, SECONDS
C
C 000000000000000000000000000000000
C
SUBROUTINE RDMS ( RR1,D2,M2,S2 )

I
REAL-8 RRl. M ,M2.$2.PIl.H60.H180.TOEQ.DIF.™iH
DATA N60/60.0D0/,N180/180.0D0/
C
PI1 = 4. ODO*DATANf 1. ODO)
TDEG= fN180*RRl)/Pll
D2 = DFLOATflDINT(TDEG))
DIF = TDEG-D2
TMIN= DIF*N60
M2 = DFLOATflDINT(TMIN))
DIF = TMIN-M2
S2 = DIF*N60

RETURN
END

63
 APPENDIX C
INDTRA FORTRAN PROGRAM

INDTRA Fortran program is used for computing and adjusting traverse station

position by least squares method of observation equations (indirect observation) and

simultaneously adjust the intersection points which are observed at each traverse

station.

C (3(3(3(3(9(3(3(3(3(3(3(3(3(3(3(3(3(3(3(3(3(3(3(90^
C (3(3 (3(3

C (3(3 LEAST SQUARE ADJUSTMENT OF (3(3

C (3(3 OBSERVATION EQUATIONS METHOD (3(3

C (3(3 (3(3

C 0P(3(3(3(3(B(3(3(3(3(3(3|B(3(3(3(3(3(3(3(3(3(cl(3(B(3(3(3(3(30(3(3P(300@
c
c
C (3(3(3(3(3(3(3(3(3(3(3(3(3(3(3(3(3(3(3(3(3(3(3(3^^
C (3

C (3 N.NA = NUMBER OF OBS. ANGLES (3

C (3 Nftl = NUMBER OF RESECTION PT. (3

C (3 ND = OBS. DISTANCES (3

C (3 NT = OBS. DIST + OBS. ANG. (3

C (3 NC = TRAV. STN. + RESECT. PT, (3

C (3 (NOT INCLUDE KNOWN STN.) (3
C (3 NZ1 = NUMBER OF RESECT. ANGLE AT EACH TRAV. STN. (3

C (3 EGRID = GRID EAST OF KNOWN POINT (3

C (3 NGRID = GRID NORTH OF KNOWN POINT (3
C (3 D.M.S = READ ANGLE IN DEGREES, MINUTES, SECONDS (3

C (3 RANG, DIST = GRID DISTANCES (3

C (3 NCO.ECO = GRID NORTH AND EAST OF INTERSECTION PT. (3

C (3 STDA = STANDARD DEVIATION OF ANG. (3

C (3 STDD = STANDARD DEVIATION OF DIST. (3
C (3 NP2 = NUMBER OF UNKNOWN DX A DY (3

C (3 CEGRID,CNGRID = TRAV. STN. C(!)ORDI. (3

C (3 INDLSQ = SUBROUTINE ADJUSTING THE COORDINATES (3
C (3 BY OBSERVATION EQUATIONS (3
C P CALFM = SUBROUTINE FOR COMPUTING CONSTANT VECTOR (3

C (3 CALAM = SUBROUTINE FOR COMPUTING THE COEFFICIENTS (3

C P OF UNKNOWN PARAMETERS (3
C (3 (3

C (3(3(3(3(3(3(3(3(3(3(3(3(3(3(3(3(3(3(3(3(3(3^^
C
C
c
DOUBLE PRECISION NGRID(4) .EGRID(4J .AZ21 .DIST21 ,AZ34,
* DIST34 XdD YGD^AZf^I R NIJM360.NL)M0
, , .

AN360R A DD A MM A S$\R A ANGlf 40) JTDA[40)

DUM10 a R^NG\d1ST(46),NC0,EC0 a CNGRID(40),
CEGRID(40) A STDD(40),AZIMUT A DIS(40K
COAZR(4O),0UMMYl.DUl5|MY2.AZC A CFAZS[4O) >

DOUBLE PRECISION SUMDX ,SUMDY' CBAZSi40) ,TTCOX(40KTTCOYjf 40)

AZLAS.DIFAZf.WPREDfSKEAZlS'CORAZ.DIFDIS,
EAZ2S,TODIS,ASNY(40),ASEX(46),DUMil,
DUM12
DOUBLE PRECISION WM( 80.80J.STDO(80J .FMf 80 ) .BMf 80^80 ) .

BMt(8d.80j,BMTWM(8d,80).Nrt(80.8d),TM(80),
DELTA(§0),NMI(80,80),WKi0(428§),VM(l,80),

64
 stdad,xcox(80),ycoy(80)

integer nt.na,np.np2.nd,ni,pn.asp,printd.n.nsta.
d.M a s.cfXzd(4o).cfAzm(40):cbazd(4oj,k,ki(,
K4 £bAZM( 40) EAZ1D EAZ1M; FJ\Z2D EAZ2M 4 N2
, . ,
*
NUM(40),ERR,NRl,NZi(40),K10,KON,II,lll,NC

DATA NUM360/360. ODO/,NUMO/0. 0D0ADUM10/2. ODO/

DATA DUMll/O. 00001D0/.DUM12/0. 0260/,SUMDX/0. ODO/
DATA SUMDY/O. ODO/, TODlS/O. ODO/
C
DATA PW/O/.PRINTD/l/
C(3(3 PW = 1 UNWEIGHT
C(3(3 PRINTD = NOT PRINT DETAILS OF MATRIX
C
C
C DETERMINE BACK AZIMUT AT STARTING STATION AND FORWARD
C AZIMUTH AT CLOSING STATION
C
C
DO 1=1.4
1
READ(5,*) NGRID(I)
READ(5 *) EGRID(I)
1 CONTINUE
C
CALL GRIDAZ (EGRID(2) ,NGRID(2) ,EGRID( 1) ,NGRID(1) ,AZ21 ,DIST21)
CALL GRIDAZ (EGRID(3) ,NGRID(3) ,EGRID(4) ,NGRID(4) ,AZ34,DIST34)
C
COMPUTE APPROXIMATED COORDINATE OF TRAVERSE STATION POSITIONS
C
C
C@(3(3 READ NUMBER OF OBSERVED ANGLES
C
READ(5,*) N

03(3(3(3 READ NUMBER OF RESECTION POINT

C
READ(5,*) NR1
C
NSTA = N-l
XGD = EGRID(2)
YGD = NGRID(2)
AZFIR = AZ21
C
CALL DMSR (NUM360,NUM0,NUM0,AN360R)
C
K =
C
DO 100 1=1, NSTA
C READ ANGLE IN DEGREE, MINUTE, SECOND
READ(5, 105J D.M.S
Dti = DFLdAf(D)
MM = DFLOAT(M)
SS = DFLOATfSJ

CALL DMSR (DD,MM,SS,R)

ANG1(I) = R
STDACI) = DUM10
C READ REDUCE GRID DISTANCES
READ(5.M RANG
DfST(I) = RANG

65
 READ(5,*) NZ1(I)
C
C

IF (RANG. EQ. 0.0) THEN

C READ APPROX. GRID COORDINATES OF RESECTION STATION

READ(5,*) NCO
READ(5,*) ECO
C
CNGRID(I) = NCO
CEGRID(I) = ECO
GO TO 100
END IF
C
K = K+l
DISfK) = DIST(I)
STDD(K) = (DIS(K)*DUM11) + DUM12
AZIMUT = AZFIR + ANGl(I)
IF (AZIMUT. GT.AN360R) THEN
AZIMUT = AZIMUT - AN360R
END IF
C
COAZR(K) = AZIMUT
03(3(3(3(3 FOR PRINTING
C
CALL RDMS (AZIMUT,DUM1D A DUM1M,DUM1S)
CFAZD(K) = iDINT(rjUMlD)
CFAZM(K) = IDINT(DUMIM)
CFAZS(K; = DUM1S
C(3(3(3(3(3
C
CALL UTM (XGD,YGD, RANG, AZIMUT, DUMMY1 ,DUMMY2)
DELX(K) = DUMMY1-XGD
DELY(K) = DUMMY2-YGD
C
SUMDX = SUMDX + DELX(K)
SUMDY = SUMDY + DELY(K)
C
CEGRID(I)= DUMMY1
CNGRID(I)= DUMMY2
C
C
CALL GRIDAZ (DUMMY1 ,DUMMY2,XGD,YGD,AZC,DUM1S)
C(3(3(3(3(3 FOR PRINTING
CALL RDMS (AZC,DUM1D,DUM1M,DUM1S)
CBAZD(K) = IDINT(DUMID)
CBAZM(K) = IDINT(DUMIM)
CBAZS(K) = DUM1S
C(3(3(3(3(3
c
AZFIR = AZC
XCOX(K) = DUMMY1
YCOY(K) = DUMMY2

XGD = DUMMY1
YGD = DUMMY2
TODIS = TODIS + RANGE
C
C
C
c
100 CONTINUE

66
c
c
K10 = K-T.
C
DO 9 I =1,K10
TTCOX(I) = XCOX(I)
TTCOY(I; = YCOY(I)
9 CONTINUE

KK = K+l

READ (5,105) D A M A S
DD = DFLOAt(D)
MM = DFLOAT(M)
SS = DFLOAT(S)

READ (5 *) RANG
DiST(N) = RANG

CALL DMSR (DD,MM,SS,R)

ANG1(N) = R
STDA(N) = DUM10
AZLAS = AZFIR + ANG1(N)
IF (AZLAS. GT.AN360R) THEN
AZLAS = AZLAS - AN360R
END IF
C
DIFAZI = AZLAS - AZ34
WPRED(1)= DIFAZI
CORAZ = DIFAZI / DFLOAT(KK)
03(3(3(3(3 FOR PRINTING
C
CALL RDMS (DIFAZI ,DUM1D,DUM1M,DUM1S)
EAZ1D = IDINK ABS (QUMID^ )
EAZ1M = IDINTf ABS (DUM1M'
EAZ1S = ABS (DUM1S)
CALL RDMS (CORAZ, DUM1D,DUM1M,DUM1S)
EAZ2D = IDINTf ABS (DUM1D) )
EAZ2M = IDINTf ABS (DUM1M) )
EAZ2S = ABS (DUM1S)
C(3(3(3(3(3
C
WPRED(2) = SUMDX - (EGRID(3) - EGRID(2))
WPRED(3) = SUMDY - (NGRID(3) - NGRID(2))
DIFDIS = DSQRTf fWPREDf 2})**2 + (WPRED(3))**2 )
ERR = IDINT( TODIS / DIFDIS )
C
C READ NUMBER OF MEASURED ANGLES AT EACH TRAVERSE STATION
C
DO 101 Jf_1.KK
READ(5,'*) N2
J)"= N2
101 CONTINUE
C
C
DO 108 I = 1.NR1
II = K10+I
READ(5,*) DUMMY1
READ(5!*1 DUMMY2
TTCOY(ll) = DUMMY1

67
 TTCOX(II) = DUMMY2
"
108 CONTINUE
C
c
c
ASP = NSTA - 1
C
CO(3 KEPP APPROX. GRID COORDINATES OF EACH STATION IN
C(3(3 ASNY AND ASEX
DO 102 1=1, ASP
ASNY(f)' = CNGRID(I)
ASNY"
ASEX (I) = CEGRID(I)
CONTINUE
C
CPP NUMBER OF OBSERVED ANGLES PLUS OBSERVED DISTANCES
C
NT = N+K
C
03(3 NUMBER OF OBSERVED ANGLES
C
NA = N
C
03(3 NUMBER OF OBSERVED DISTANCES
C
ND = K
C
03(3 NUMBER OF TRAVERSE STATIONS NOT INCLUDED KNOWN STATIONS
C
NP = ASP
C
03(3 NUMBER OF UNKNOWN DX AND DY
C
NP2 = K10*2 + NR1*2
C
C
03(3
NI = NP2**2 + 3*NP2
C
C(3(3 NUMBER OF TRAVERSE STATION (NOT INCLUDE KNOWN STATION)
C PLUSE NUMBER OF INTERSECTION POINT
C
NC = K10+NR1
C
03(3(3(3(3 CALL SUBROUTINE TO ADJUST STATION POSITIONS
C BY LEAST SQUARE METHOD OF OBSERVATION EQUATION

CALL INDLSQ (NC.NRl.NZl.NUM.KK.K.DIST.TTCOX.TTCOY.NT.NA.ND.NP,

NP2\NI,AZ2l\AZ34 A N\^ID,E£RID A AtfGl,STdA,dlS,STdD,
* PRINTD,PW,W^.STD6,FM.BM,BMT,6MTWM!NM,Tf5l,
DELTA, NMI.WKiO, ASNY, ASEX,VM,STDAD)
c
03(3(3(3 PRINTING DETAILS
C
WRITE(6,*V STANDARD DEVIATION OF UNIT WEIGHT =',STDAD
WRITE[6 *]' '

WRITE(6,*)'STANDARD DEVIATION OF ADJ. ANGLES'

DO 103 1=1. NA
CALL RtfMS(VM(1 A I) J DUM1D A DUM1M,DUM1S)
DUMMY1 = 6UMiD*36O0. ODO + DUM1M*60. ODO + DUM1S
WRITE(6,*)'I =',I, Y ADJ. STDA. =', DUMMY
103 CONTINUE
C
WRITE(6,*)' STANDARD DEVIATION OF ADJ DISTANCES

68
 DO 104 1=1, ND
K4 = NA+I
WRIT€(6,*)
WRITE(6 *) I =',I,'ADJ. STDD. =',VM(1,K4)
104 CONTINUE

105 FORMAT( 13 IX 12 IX 12)

, , , ,

STOP
END

(3(30(3(3(3(3(3(3(30(3(3(3(3(3(3(3(3(3(3(3(3(3(3^(30(3(3(90(3(3(3(3(3^0
(3(3 (3(3
(3(3 SUBROUTINE LEAST SQUARE (3(3
(3(3 ADJUSTMENT OF INDIRECT (3(3
(3(3 OBSERVATIONS (3(3
(3(3 (3(3

PP(3(3(3(3(3(3(3(3(3(3(3(3(3(3(3(3^(3P(3(3(3^(3PP0(3(3(3(3@(3(3@0

SUBROUTINE INDLSQ (NCI .NR.NZZ.NUMM.KK.K.DISTl ,TCOX.TCOY,NT,NA,

"ND.I^P.^P2,^I.AZ^l.XZJ4.NGRfD.EGSlD,A^G,
*
STflA,fllSl'STflD,PRlNTD,PXWM.$TDO.Ffl,
BM BMT.BMtWM NM TM, DELTA NMl ,WKl6
, , . ,
*
ASflG,A$EG,VM STAND)

DOUBLE PRECISION DUMMY1,NGRID(4),EGRID(4).DUM1D,DUM1M,

DUM1S,ANG NA),D]!SHND),W^('NT,Nf),STDd(NT),
ASNG(rtP)J\SEG(NP),FM£Nt),BM(NT,NP^ JESTl!
TEST2,BMt(NP2 NT] BMTWM(nP2,NT3,NM(NP2,NP^),
"\0(Ni)
STDA(NA),STDD(ND)^rjUMMY2,NUM0 AZ21 ,AZ34
;
; )
"1(1." STAND, DI$T1(NA);TC0X(NC1);
tcoy(nci)
:6y(

INTEGER IJ,K1 J IER,PRINTD J PW J CHECK,NUMM(KK),K,KK,

Nrt,^ZZ(NA),KON,Ili,N(!l,NNDi

DATA K1/0/.TEST1/0. 00010000000D0/.NUM0/0. ODO/

NC4 = 4
NP5 = 6
Nl = 1
C
C(3(3 SET WEIGHT MATRIX
C
DO 30 1=1, NT
DO 20 J=LNT
WM(I.Ji = O.ODO
).J) 1THEN
IF (t.EQ.J)
Wl,Jj = l.ODO
END IF
20 CONTINUE
30 CONTINUE
C(3(3 FOR STD. ANGLE
C
IF (PW. NE. 1) THEN
1)0 40 1=1, NA
DUMMY1 = STDA(I)

69
 CALL DMSR (NUM0,NUM0.DUMMY1 .DUMMY2)
STDO(I) = l.ODfl / ( (DSIN(DUMMY2) )**2)
40 CONTINUE
C
CP(3 FOR STD. DISTANCE
C
DO 50 1=1,
J = NA+I
DUMMY1 = STDD(I)
STDO(J)= 1.0D0 / (DUMMY1**2)
50 CONTINUE
C
C(3P SET UNEQUAL WEIGHT
DO 70 1=1, NT
DO 60 J=1,NT
IF (I.EQ.J) THEN
WMJXJJ = STDO(I)
END IF
60 CONTINUE
70 CONTINUE
C
END IF
C
CPP PRINT WEIGHT MATRIX
C
IF (PRINTD. NE.O) THEN
WRITE(6,*) 'WEIGHT
r
MATRIX'
WRITE(6,*1 '

CALL USWFM ( R-C. ,NC4,WM,NT,NT,NT,NP5)

'
'

END IF
C
C
99 CONTINUE
C
C
Kl = Kl+1
C
IF (Kl .GT. 2) THEN
GO TO 999
END IF
C
CHECK =
C
CPP CALL SUBROUTINE TO CALCULATE "F" MATRIX
C
C
CALL CALFM (NUMM.KK^DISTl ,TCOX .TC0Y.AZ21 .AZ34.NA.ND,
NT.NCaRld^GRID.XNG.DfSl.NfJ.ASNd.ASEd.FM)
c
c
c
C(3(3 CALL SUBROUTINE TO CALCULATE "A" MATRIX (COEFFICEINT DX,DY)
C
CALL CALAM (NR.NZZ.NUMM.KK.K.DIST1.TC0X.TC0Y,NT,NA,ND,NP,
NP2\NGtflD,E(3RID\ASNG,ASEG,BM)
c
C@(3 TRANSPOSE OF "A" MATRIX
C
DO 120 1=1, NT
DO 110 J=1.NP2
BMT(J,f) = BM(I.J)
c
110 CONTINUE
120 CONTINUE
C
C(30 "A" TRANSPOSE * W

70
c
CALL VMULFF (BMT,WM,NP2,NT,NT,NP2,NT,BMTWM,NP2,IER)
C(3(3 "A" TRANPOSE * W * A
C
CALL VMULFF (BMTWM,BM,NP2,NT,NP2,NP2,NT,NM,NP2,IER)
C(3(3 "A" TRANSPOSE * W * F
C
CALL VMULFF (BMTWM,FM,NP2,NT,N1,NP2,NT,TM,NP2,IER)
w
C
c
CP@ (INVERSE OF ("A" TRANSPOSE W * A) )

CALL LINV2F (NM,NP2,NP2,NMI ,N1 ,WK10, IER)

c
C(3(3 CALCULATE DELTA MATRIX (COORECTION VECTOR)
C (INVERSE OF ("A" TRANSPOSE W * A] * "A" TRANSPOSE * W * F )
CALL VMULFF (NMI ,TM,NP2,NP2, N1,NP2,NP2, DELTA, NP2, IER)
CPP UPDATE APPROX. VALUE
C
C
C KON = K-l+NR
DO 130 1=1, NCI
TCOX(I) = TCOX(I) + DELTA((I-1)*2 + 1)
TCOY(I) = TCOY(I) + DELTA((I-1)*2 + 2)
130 CONTINUE
C
c
III = (K-l)*2
DO 131 1=1, NP
C
IF (NZZ(I) EQ. CO THEN
.

ASEG(I) = ASEG(I) + DELTAf (1-1*2 + 1)

ASNG(I) = ASNG(I) + DELTA( (I-l)*2 + 2)
ELSE
DO 132 J =1,NR
IF (NZZ(I) .EQ. JJ THEN
ASEG(r) = ASEG(I) + DELTA( II1+(J-1)*2 +1
ASNGf I) = ASNG(I) + DELTA( II1+(J-1)*2 +2)
GO TO 131
END IF
132 CONTINUE
END IF
C
131 CONTINUE
C
c
IF (PRINTD .NE. 0) THEN
WITE(6,1020) Kl

CALL U$WFM ('R-C. ',NC4,FM,NT,NT,N1,NP5)

WRITE(6,1000)
WRITE(6 1040)
CALL USWFM ( R-C. ',NC4,BM,NT,NT,NP2,NP5)
'

WRITE(6,1000l
5
CALL USWFM ('R-C. ',NC4,BMT,NP2,NP2,NT,NP5)
WRITER, 1000)
6
CALL ulwFM ('R-C.
'

,NC4,BMTWM,NP2,NP2,NT,NP5)
WRITEf 6,1C005
7
CALL U$WFM ('R-C. ',NC4,NM,NP2,NP2,NP2,NP5)

71
 WRITE(6,1000)
WRITE(6 1080] ,,
CALL USWFM ('R-C. ',NC4,TM,NP2,NP2,N1,NP5)
WRITE(6,1000)
CALL USWFM ( ' R-C. *
,NC4,NMI,NP2,NP2,NP2,NP5)
WRITE(6,1000)
WRITE(6 1100]
CALL USWFM f ' R-C. '
,NC4, DELTA, NP2,NP2,N1,NP5)
WRITE(6,1000)
WRITE(6 1110)
DO 140 J=I,NP
WRITER,*) I, N = ASNG(I),' E = '
,ASEG(I)
f
WRITE(6,*) '

140 CONTINUE
C
END IF
C
C
C(3@ CHECK DELTA MATRIX
C
C
DO 180 1=1, NP2
TEST2 = DABS (DELTA(I))
IF ( TEST2 .GE. TEST1) THEN
CHECK = 1
END IF
180 CONTINUE
C
C
IF (CHECK .NE. 0) THEN
GO TO 99
END IF
C
CP@ CALCULATE "A" * X MATRIX
C
CALL VMULFF (BM, DELTA, NT, NP2,N1 ,NT,NP2,STD0,NT,IER)
C(3(3 CALCULATE "V" MATRIX
C
DO 190 1=1. NT
STDO (f) = STDO(I) - FM(I)
VMC.1,1) = STDO(I)
190 CONTINUE
C
C
C CORRECT OBSERVED ANGLES AND DISTANCES
C
DO 139 I = 1,NA
ANG(I) = ANG(I)+(STDO(I))
139 CONTINUE
DO 149 I = l.ND
NND1 = NA+^
DIS1(I) = DIS1(I)+STD0(NND1)
149 CONTINUE
IF (PRINTD . NE. OJ THEN
WRITE(6,1000T
WRITE(6 1120]
CALL USWFM ('R-C. ',NC4,STD0,NT,NT,N1,NP5)
WRITE(6,1000)
WRITE(6,1130J
CALL USWFM ('R-C. ,NC4,VM,N1,N1,NT,NP5) '

END IF
C
C
CP(3 CALCULATE "V" TRANSPOSE * W * V MATRIX

72
 CALL VMULFF (WM,STD0,NT.NT,N1,NT.NT.FM.NT.IER]
CALL VMttLFF (VM,FM,Ni ,Nt,Ni ,Ni Nt StAND Nl IER) , , , ,

IF (PRINTD NE. 0) THEN

.

WRITE(6,1000)
WRITE(6 1140'
CALL r USWFM ( R-C. '
'
,NC4,FM,NT,NT,N1,NP5)
WRITE(6,11501
WRITE(6 1160 STAND
END IF

CALL VMULFF (NMI ,BMT,NP2,NP2,NT,NP2,NP2,BMTWM,NP2, IER)

STAND = DSQRT ( STAND / (DFLOAT(NT-NP2)) )

DO 240 1=1, NP2

DO 230 J=1,NP2
NMI (I, J) = STAND (DSQRT(DABS (NMI(I.J)) ) )

IF (LEO. J) THEN
NMI(I.J)
END IF
230 CONTINUE
240 CONTINUE

CP(3 CALCULATE "A"*( INVERSE "A"TRANSPOSE*W*A) "A"TRANSPOSE

C
CALL VMULFF (BM,BMTWM,NT,NP2,NT,NT,NP2,WM,NT, IER)
IF fPRJ
/RITEC6, 1000T
WRITEC6 1170)
CALL USWFM ( <R-C. '
,NC4,WM,NT,NT,NT,NP5)
END IF

DO 280 1=1, NT
DO 270 J=1.NT
WM (I, J) = STAND * ( DSQRT( DABS(WM( I ,J)) ) )

IF I ..EQ.J) THEN
VMir.ij = wm(i,j)
END IF
270 CONTINUE
280 CONTINUE
C
999 RETURN
C
c
1000 F0RMATC1H1 1

)
1020 FORMATf ///,5X 12,' ITERATED ,/)
1

1030 FORMATf/// 5X [F matrix!,/)

1040 FORMATf/// 5X A MATRIX 1
;/'
1050 FORMATf/// 5X 'TRANSPOSE 01 A MATRIX' ,/)
1060 FORMATf/// 5X 'A TRANSPOSE W MATRIX',/) '

1070 FORMATf/// 5X 'A TRANSPOSE * W * A MATfti:

IX /)
1

1080 FORMATf/// 5X * w
'A TRANSPOSE * W **
F MATRIX'!/)
1090 FORMATf/// 5X 'THE INVERSION OF A TRANSPOSE' W * A',/)
1100 FORMATf///, 5X 'X MATRIX OR DELTA MATRIX*/)
1110 FORMATf/// 5X 'UP-DATE THE APPROX. VALUES',/)
1120 FORMATf/// 5X 'V MATRIX'/)
1130 FORMATf/// 5X 'V TRANSPOSE MATRIX',/)
1140 FORMATf/// 5X W * V MATRIX' ,/)
* $ * V
1150 FORMATf///, 5X 'V TRANSPOSE MATRIX',/)

73
1160 F0RMAT(///,5X,F20. 15,/} , s , ,

1170 FORMAT?/// 5X'A*(IN$ERSE A TRANSPOSE *W*A)*A TRANSPOSE',/)

C
END
C
C
c
c
c
C (30(3(3(3(3(3(3(3(3(3(3(3(3^(3^(3(3(3^(3(3(3(3(3(3(3(3(3(3(3(3(3(3(3(3(30
C (9(3(3 POP
C (3(3(3 SUBROUTINE FOR COMPUTING (3(3(3

C (3(3(3 F MATRIX (3(3(3

C (3(3(3 (3(3(3

C p@@(30(3(3(3(3(3(3(3(3P(3(3(3(3(3P(3(3(3(3(3pp(3(3(3(3(3(3(3(3(3p^
C
C
C
SUBROUTINE CALFM (K0UNT,KK,K.DIS C0X C0Y.AZ21.AZ34, 1 >

NNA,NtfD,rtNf,GRY,GRX\OBANG,OgDIS,
*
NOS,ASY,ASX,F)

DOUBLE PRECISION GRY (4) J GRX(4),0BANGi(NNA) J 0BDIS(NND) J

N36a.N360R.DUMl A NUM0,CDIS7l001,TEMPI,
TEMP2.TEMP3\TEMPXASY(NOS).AZF AZB,
ASX(n6S] a F(NNT],a221,AZ34,6IS(NNA)
COX(K),C6Y(K),DiSTAN,ADIS(l00)

INTEGER Il,I2,K,KK,KKK,N,STN,KOUNT(KK)

DATA N360/360. 0D0/,NUM0/0. 0D0/,N/0/,STN/l/

CALL DMSR (N360,NUM0,NUM0,A360R)

DO 500 1=1, NNA
DISTAN = DIS(I)
IF (STN .EQ. 1) THEN
TEMPI = GRX(2)
TEMP2 = GRY(2)
AZB = AZ21
TEMP3 = ASX(I)
TEMP4 = ASY(I)
CALL GRIDAZ (TEMPI ,TEMP2,TEMP3,TEMP4,AZF,DUM1)
CDISm = DUM1
IF (AZF .GT. AZBJ THEN
F(I) = OBANG(T)-(AZF-AZB)
ELSE
F(I) = OBANG(I)-(A360R+AZF-AZB)
END IF
N = N+l
GO TO 400
END IF
IF (STN .EQ. 2) THEN
TEMPI = COXCSTN-:
COX(STN-l)
TEMP2 = COYCSTN-1)
Tl

CALL GRIDAZ (TEMPI, TEMP2,TEMP3,TEMP4, AZB, DUM1)

74
 IF (KK EQ. 3) THEN
.

TEMP3 = GRX(3)
TEMP4 = GRY(3)
END IF
TEMP3 = ASX(
TEMP4 = ASYl'(I)
CALL GRIDAZ (TEMPI ,TEMP2,TEMP3,TEMP4,AZF,DUM1)

CDIS(I) = DUM1
IF (AZF .GT. AZBJ THEN
F(I) = OBANGCI)-(AZF-AZB)

F(I) = OBANG(I)-(A360R+AZF-AZB)

N = N+l
GO TO 400
END IF

IF (STN .EQ. KK) THEN

AZF = AZ34
TEMPI = GRX(3)
TEMP2 = GRYC3)
C
C IF (DISTAN .EQ. 0. ) THEN
C TEMP3 = ASX(I)
C TEMP4 = ASY(I)
C ELSE
TEMP3 = COX(K-l)
TEMP4 = COY(K-l)
C END IF
c
CALL GRIDAZ (TEMPI ,TEMP2,TEMP3,TEMP4,AZB,DIS1)
IF (AZF .GT. AZB) THEN
F(I) = OBANGCI)-(AZF-AZB)
ELSE
F(_I) = 0BANG(I)-(A360R+AZF-AZB)

N = N+l
GO TO 400
END IF

IF (STN .EQ. K) THEN

TEMPI = COX(K-r
TEMP2 = COYfK-l'
TEMP3 = C0X(K-2'
TEMP4 = C0Y(K-2;
CALL GRIDAZ (TEMPI ,TEMP2,TEMP3,TEMP4,AZB,DUM1)
IF (DISTAN .EQ. 0.) THEN
TEMP3 = ASX(I)
TEMP4 = ASY(I)
ELSE
TEMP3 = GRX(3)
TEMP4 = GRY(3)
END IF
CALL GRIDAZ (TEMPI ,TEMP2,TEMP3, TEMP4, AZF, DUM1)
CDIS(I) = DUM1

75
 IF (AZF .GT. AZB) THEN
F(I) = OBANGO)-(AZF-AZB)
ELSE
F(_I) = 0BANG(I)-(A360R+AZF-AZB)

N = N+l
GO TO 400
END IF

IF (STN .GT. 2) THEN

TEMPI = COXYSTN-1
TEMP2 = COY(STN-l'
TEMP3 = COX(STN-2'
TEMP4 = COY(STN-2'
CALL GRIDAZ (TEMPI, TEMP2,TEMP3,TEMP4,AZB,DUM1)
TEMP3 = ASX(I)
TEMP4 = ASY(I)

CALL GRIDAZ (TEMPI ,TEMP2,TEMP3, TEMP4, AZF, DUM1)

CDISfll = DUM1
IF (AZF .GT. AZB) THEN
F(I) = OBANGCI)-(AZF-AZB)
ELSE
F(_I) = 0BANG(I)-(A360R+AZF-AZB)

N = N + 1

END IF

400 IF ( (DISTAN .NE. 0.) .AND. (STN LE. K) ) THEN

.

ADISfSTN) = CDIS(I)
ADISCSTN)
WRITE(6,*) STN=' ,STN,'
,

CDIS=' ,CDIS(I)
END IF

KKK = KOUNT(STN)
IF (N LT. KKK) THEN
.

STN = STN
ELSE
STN = STN+1
N =0
END IF
C
C
500 CONTINUE
C
C
c
II = NNA
DO 600 12=1, NND
II = 11+1
F(I1) = OBDIS(I2) - ADIS(I2)
600 CONTINUE
C
RETURN
END

76
C (3(9(^(3(30(3(3(2^^(30(300^^0^(30^(3(300(3(3300(3(3(3(3
C 5(3 (3(3
C ,:.r SUBROUTINE FOR CALCULATING
C TO A MATRIX ;=?
C : = (3(3
C :::^::?^::::^:?:?:::::::::^^?^::^
c
c
c
SUBROUTINE CALAM (NR.NZ, COUNT. KK.K, DISS, COXX.COYY, NT, NA,
ND,NP,NP2,KMy,|(N)(,ANY,ANX,XM)

-
DOUBLE PRECISION KNY(4) KNX(4^ANY(NP^ANX(NP^DISTAN,
,

AM(NT,N'P2).DIS S(NA),CdXX(k),CdYY(K),
du'i,du2,du3,du4;du$;du6

INTEGER 1,11,12,13 J4, 15. J1.J2,K,KK.K1, K2.K3,

K4,N.NN,STN^.T^M.tEMi,T^,jJ.JJl,Jj2.
NOft,£OUNT(Kfc),Tt,TTl TEM3,NU$ED,NZ(nA),
NR.L.I10

DATA STN/l/,N/l/,NN/0/,TT/0/,JJ/l/

NP1 = NP2-1
Kl = NA-1
K2 = K-I
K3 = K+l
110 = ((K-l)*2) - 1

C (3(3(3(3t3@(9@(3^(3(3(3(3@P(3(3P(p(3(3(3(3P(3!3P(3(3
C P (3

C (p COEFFICEINT OF COORDINATES (=»

C (3 FOR ANGLE CONDITION (3

C := (3

C
c
DO 800 I = l.NA
DISTAN = DISS(I)
DO 700 Jl = 1,NP1,2
J2 = Jl+1
AM(I,J1) = O.CDO
AM(I,J2) = O.ODO
C
IF (STN . EQ. 1) THEN
C
C
IF (DISTAN .EQ. 0) THEN

DO 231 L=1,NR
IF (NZ(I) .EQ, L) THEN
JJ = I10+2*L
GO TO 232
END IF
231 CONTINUE
C
ELSE
JJ = 1
TEM = JJ
GO TO 232
END IF
C

232 IF (Jl .EQ. JJ) THEN

77
 DU1 = ANY(I)-KNY(2)
DU2 = ANXfll-KNXm
DU3 = (DU1**2)+(DU2**2)
AM(I,J1) = DU1/DU3
AM(I,J2) = -DU2/DU3
N = N+l
GO TO 700
END IF
GO TO 700
END IF
********************

IF (STN . EQ. 2) THEN

IF (Jl .EQ. J J) THEN

TDU1 = ANY(I) - COYY(STN-l)
DU2 = ANX(I) - COXX(STN-l)
DU3 = (DUl**2)+rDU2 2) u
DU4 = KNY(2) - COYY(STN-l)
DU5 = KNX(2) - COXX(STN-l)
DU6 = (DU4**2)+(DU5**2)
AM(I,J1) = -(DU1/DU3WDU4/DU6)
AM(I,J2) = (DU2/DU3)+(DU5/DU6)

IF (DISTAN .EQ. 0) THEN

W
233 I = l.NR
IF ( NZ(f) EQ. L) THEN .

JJ1 = I10+2*L
GO TO 700
END IF
233 CONTINUE
c
ELSE
JJ1 = TEM + 2
TEM1 = JJ1
GO TO 700
END IF
GO TO 700
END IF
IF (Jl .EQ. JJ1) THEN
AM(I,J1) = (DU1/DU3)
AM(I,J2) =-(DU2/DU3)
N = N+l
GO TO 700
END IF
GO TO 700
END IF
****************

IF ( (STN .GT. 2) .AND. (STN . LE. K2) ) THEN

K4 = STN - 2

IF (Jl .EQ. JJ) THEN

DU1 = ANY(I) - COYY(STN-l)
DU2 = ANX(I) - COXX(STN-l)
DU3 = (DU1**2)+(DU2**2)
DU4 = C0YY(K4) - COYY(STN-l)
DU5 = C0XX(K4) - COXX(STN-l)

78
 DU6 = (DU4**2)+(DU5**2)

AM(I.Jl) = -DU4/DU6
AM(I,J2) = DU5/DU6
JJ1 = TEM1
GO TO 700
END IF
IF (Jl .EQ. JJ1) THEN
AM(LJl) = "(DU1/DU3) + (DU4/DU6)
AM(I,J2) = (DU2/DU3) - (DU5/DU6)
IF fDISTAN .EQ. 0) THEN
DO 234 L = 1,NR
IF NZ I .EQ. L) THEN
JJ2 = I10+2*L
GO TO 700
END IF
234 CONTINUE
c
ELSE
JJ2 = TEM1 + 2
TEM2 = JJ2
GO TO 700
END IF

GO TO 700
END IF
IF (Jl .EQ. JJ2) THEN
AM(I.Jl) = DU1/DU3
AM(I,J2) = -DU2/DU3
N = N+l
GO TO 700
END IF
GO TO 700
END IF
*******************

IF (STN .EQ. K) THEN

IF (Jl . EQ. JJJ THEN
1)1)1 = COYYfK-2) - COYY(K-l)
DU2 = C0XX(K-2) - COYY(K-l)
DU3 = (DU1**2) + (DU2**2)
AM(I,J1) = -DU1/DU3
AM(I,J2) = DU2/DU3
JJ1 = TEM1
GO TO 700
END IF
IF (Jl EO. JJ1) THEN
.

IF (DISTAN .EQ. 0. ) THEN

DU4 = ANY(I) - C0YY(K-1'
DU5 = ANX(I) - COXXfK-1'
DU6 = (DU4**2) + (D05**;
ELSE
TEMP2 = Jl
DU4 = KNY(3) - C0YY(K-1'
DU5 = KNX(3) - COXXfK-1'
DU6 = (DU4**2) + (DG5**;

79
 AM(I,J1) = -(DU4/DU6) + (DU1/DU3)
AM(I,J2) = (DU5/DU6) - (DU2/DU3)
N = N + 1

NUSED = Jl
JJ2 =
GO TO 700
END IF
NN = NN + 1

AM(I,J1) = -(DU4/DU6) + (DU1/DU3)

AM(I,J2) = (DU5/DU6) - (DU2/DU3)
DO 235 L = l.NR
IF ( NZ(f) EQ. . L) THEN
JJ2 = I10+2*L
GO TO 700
END IF
235 CONTINUE

GO TO 700
END IF
IF (Jl .EQ. JJ2) THEN
AM(I,J1) = DU4/DU6
AM(I J2) = -DU5/DU6
N = N + 1
GO TO 700
END IF
GO TO 700
END IF
c
c
******************
C
IF (STN EQ. KK) THEN
.

IF (Jl EQ. JJ) THEN

.

TJU1 = COYYfK-1) - KNY(3

DU2 = COXX(K-l) - KNX(3
DU3 = (DU1**2) + (DU2**
c
AM(I,J1) = -(DU1/DU3)
AM(I,J2) = (DU2/DU3)
c
N = N + 1
GO TO 700
END IF
c
GO TO 700
END IF
c
700 CONTINUE
C
NOM = COUNT(STN)
c
IF (N . LE. NOM) THEN
c
STN= STN
c
ELSE
c
STN = STN + 1
JJ = TEM
N = 1

80
 IF (STN .GT. 3) THEN
JJ = TEMI
TEM1 = TEM2
END IF
END IF
C
800 CONTINUE
C
C 0000000000000000000000000(3
C COEFF. OF COORDINATES
C FOR DISTANCE CONDITION P
C 0(3(3(9(3(3(3(3(3(3(3(3(3(3(3(3(300(3(3(3(3(3(3(3
C
C
DO 3000 II = l.ND
12 = NA + fl
13 = II - 1
14 = 13 + (13 -
15 = 13 + (13

DO 2000 Jl = 1.NP1.2

J2 = Jl + 1
AM(I2,J1) = O.ODO
AM(I2,J2) = O.ODO
IF (II EQ. 1J THEN
.

IF (Jl EQ. 1J THEN

.

DU1 = KNYC2) - COYY(l)

DU2 = KNX(2) - COXXC 1)
DU3 = DSQRT ( DU1**2 + DU2**2)
AM(I2,J1) = -(DU2/DU3)
AM(I2,J2) = -(DU1/DU3)
END IF
GO TO 2000
END IF

IF (II .EQ. ND) THEN

IF (Jl EQ. NUSED) THEN
.

T»U1 = COYY(K-I) - KNY(3'

DU2 = COXX(K-l) - KNX(3'
DU3 = DSQRT ( DU1**2 + DU2**2)
AM(I2,J1) = (DU2/DU3)
AM(I2,J2) = (DU1/DU3)
END IF
GO TO 2000
END IF

IF (14 . EQ. J1J THEN

VUl = C0YYCI3) - C0YY(I1
DU2 = C0XXfI3) - C0XX(I1'
DU3 = DSQRT ( DU1**2 + D02**2
AM(I2,J1) = (DU2/DU3)
AM(I2!J2) = (DU1/DU3)
GO TO 2000
END IF
IF (15 EQ. J13 THEN
.

I)U1= C0YYCI3) - COYYfll

DU2 = C0XX(I3) - C0XX(I1
DU3 = DSQRT ( DU1**2 + D02**2
 - AM(I2,J1) = -(DU2/DU3)
AM(I2 J2) = -(DU1/DU3)
C
END IF
C
2000 CONTINUE
C
3000 CONTINUE
C
RETURN
C
C
END
C
C
C
C
C 000000000000000000000000000000000000000
C
C
00
(3(3 SUBROUTINE TO CALCULATE GRID
W
P@
C 00 AZIMUTH AND DISTANCE P(3
C 00 @@
C 000000000000000000000000000000000000000
C
SUBROUTINE GRIDAZ (XI ,Y1 ,X2,Y2, ANGLR, DIST12)

REAL*8 XI .X2. Yl Y2.ANGLR.ANG0.ANG90.ANG180.ANG270

,

REAL*8 ANG\90R,AN18(!)R,AN2?0R,DiFX,DIFY,DISTi2
DATA ANGO/0. 0D0AANG90/90. 0D0/,ANG180/180. 0D0/
DATA ANG270/270. 6D0/
CALL DMSR ( ANG90 A ANG0 A ANG0 A ANG90R )
CALL DMSR ( ANGlS^ANGd.ANG^ANlSOR
CALL DMSR ( ANG270 ANGO ANGO AN270R
DIFX = X2-X1
DIFY = Y2-Y1
IF((DIFX. EQ. ANGO). AND. (DIFY.EQ.ANGO)) THEN
ANGLR = ANGO
ELSE IFi DIFX. EQ. ANGO) THEN
IF(DIFY.GT.ANGO) THEN
ANGLR = ANGO
END IF
ANGLR = AN180R
ELSE
IF(DIFX.GT.ANGO) THEN
ANGLR = ANG90R-DATAN(DIFY/DIFX)
ELSE
ANGLR = AN270R-DATAN(DIFY/DIFX)
END IF
END IF

DIST12 = DSQRT(DIFX**2+DIFY**2)
RETURN
END
C
C
C
C 000000000000000000000000000000000000000
C 00 00
C 00 SUBROUTINE FOR CHANGING THE 00
C 00 ANGLE FROM DEGREE, MINUTE, AND 00
C 00 SECONDS TO RADIAN 00

82
C 00 00
C 00000000(3000000000000000000000000000000
SUBROUTINE DMSR ( D1,M1,S1,R1 )

REAL*8 D1,M1,S1,R1,PI1,N60,N3600,N180
DATA N60/60. 0D0/,N3600/3600. 0D0/,N180/180. 0D0/
PI1 = 4. ODO*DATANfl.ODO)
Rl = ((D1+(M1/N60)+(S17N3600))*PI1)/N180

RETURN
END
C
c
c
C 000000000000000000000000000000000000000
C 00 00
00 SUBROUTINE FOR COMPUTING GRID 00
C 00 COORDINATES FROM KNOWN AZIMUTH 00
C 00 DISTANCE AND STARTING COORDI- 00
C 00 NATES 00
C 00 00
C 000000000000000000000000000000000000000
C
C
SUBROUTINE UTM (XI ,Y1 DIS12 ,ANG12R X2 Y2
, , ,
)

REAL*8 X1,Y1,X2,Y2,DIS12,ANG12R,DIFX,DIFY
DIFX = DIS12*(DSIN(ANG12R))
DIFY = DIS12*(DCOS(ANG12R))
X2 = Xl+DIFX
Y2 = Yl+DIFY
C
RETURN
END
C
c
c
C 000000000000000000000000000000000000000
C 00 00
C 00 SUBROUTINE FOR CHANGING THE 00
00 ANGLE FROM RADIANS TO DEGREE, 00
C 00 MINUTES, AND SECONDS 00
C 00 00
C 000000000000000000000000000000000000000
c
SUBROUTINE ROMS ( RR1,D2,M2,S2 )
C
REAL*8 RR1,D2,M2,S2,PI1,N60,N180,TDEG,DIF,TMIN
DATA N60/60. ODO/ N180/180. ODO/
,

C
PI1 = 4.0D0*DATAN[1.0D0)
TDEG= [N180*RR1)/PI1
D2 = DFLOAT(IDINT(TDEG))
DIF = TDEG-D2
TMIN= DIF*N60
M2 = DFLOATflDINT(TMIN))
DIF = TMIN-M2
S2 = DIF*N60
C
RETURN
END

83
 LIST OF REFERENCES

Allan, Arthur L v J.R. Hollwev, and J.H.B. Maynes, Practical Field Surveying and
Computations, Fifth Edition, Heinemann Publishing Co., 1968.

Bomford, G., Geodesy, Fourth Edition, Clarendon Press, Oxford, 1980.

Brinker, Russell C, and Paul R. Wolf, Elementary Surveying, Sixth Edition, Harper &
Row, Publishers, 1977.
Clark, Late David, and J.E. Jackson, Plane and Geodetic Surveying, Volume II Higher
Surveying, Sixth Edition, Constable &
Company Ltd., 1976.
Davis, Ravmond E., Francis S. Foote, James M. Anderson, and Edward M. Mikhail,
Surveying Theory and Practice, Sixth Edition, McGraw-Hill Company, 1981.

Department of the Army Technical Manual, TM 5-241-8. Universal Transverse

Mercator Grid, Headquarters, Department of the Army, 1958.
Evett, Jack B., Surveying, First Edition, John Wiley & Son, 1979.

Federal Geodetic Control Committee. Standards and Specifications for Geodetic Control
Networks, Rockville, Maryland, 1984.

Mikhail, Edward M., and Gordon Gracie, Analysis and Adjustment of Survey
Measurements, First Edition, Van Norstrand Reinhold Co., 1981.

Mikhail, Edward M., and F. Ackermann, Observations and Least squares, First
Edition, Dun-Donnelley Publisher, 1976.
Schmidt, Milton O. and William H. Ravner, Fundamentals of Surveying, Second
Edition, D. Van Nostrand Company, 1978.

Torge, Wolfgang, Geodesy, First Edition, Walter de Gruyter Berlin, New York, 1980.

84
 INITIAL DISTRIBUTION LIST

No. Copies
1. Defense Technical Information Center 2
Cameron Station
Alexandria, Virginia 22304-6145
2. Library Code 0142
Naval "Postgraduate School
Monterey, California 93943-5002
3. Chairman, Department of Oceanography
Code 68
Naval Postgraduate School
Monterey, California 93943-5000
4. NOAA Liaison Officer
Post Office Box 8688
Monterey, California 93943-0688
5. Dr. Muneendra Kumar
Defense Mapping Agencv
Hydrographic; Topographic Center
Geodesy and Surveys Department
6500 Brookes Lane
Washinton, D.C. 20315
6. Office of the Director
Naval Oceanosraphv Division (OP-952)
Department ofthe Navv
Washington, D.C. 20350
7. Commander
Naval Oceanographv Command
NSTL, Mississippi 39529
8. Commanding Officer
Naval Oceanographic Office
Bav St. Louis
NSTL, Mississippi 39522
9. Commanding Officer
Naval Ocean Research and Development Activity
Bav St. Louis
NSTL, Mississippi 39522
10. Chief of Naval Research
800 N. Quincy Street
Arlington, Virginia 22217

11. Chairman, Oceanography Department

U.S. Naval Academy
Annapolis, Maryland 21402

12. Topographic Sciences Department

CodeTSD-MC
Defense Mapping School
Ft. Belvoir, Virginia 22060-5828

85
13. CAPT Glen R. Schaefer
Code 68 Sc
Naval Postgraduate School
Monterey, California 93943-5000
14. Supote Klangvichit
Hydrographic Department
Royal Thai Navy
Banckokyai, Bangkok
Thailand

15. Hvdrographic Department Library

Hydrographic Department
Roval Thai Navy
Bariskokyai, Bangkok
Thailand

16. Yu, Ta-Te

SMC 1343, NPS
Monterey, California 93943
17. Wang, Chih-Ping
SMC* 1342, NPST
Monterey, California 93943

17 898
86 \/
 LIBRARY
2ft?*™ SCHOOL ^

Thesis
k
,

C•
282
-L
^Sr^ adjustment.

1363
31 HW «?

297
Thesis
K57282
~
Klangvichit
c i
Traverse adjustment