COORDINATE TRANSFORMATIONS

UNI VERSAL T R A N S V E R S E M E R C A T O R / G E O G R A P H I C

by KEVI N M. KELLY

O n t a r i o M i n i s t r y of Na t u r a l R e s o u r c e s

MARCH 1 9 8 6
UNIVERSAL TRANSVERSE MERCATOR/GEOGRAPHIC COORDINATE TRANSFORMATIONS

BT KEVIN M. KELLY,
ONTARIO MINISTRY OF NATURAL RESOURCES

PREFACE

I have been requested on numerous occassions to provide

algorithms for performing geographic to UTM and vice versa
coordinate transformations. Many of these requests have come
from surveyors and engineers. Vhile there exists a multitude
of software packages that can readily handle these copmutatlons,
it seems that many individuals are acquiring their own
micro-computers or sophisticated desk-top calculators and as a
result are desiring to generate their own programs for these
computations.

In the past I have had to sift through numerous references

to complie these formulas each time I recieved a request. I
can now appreciate why these individuals called me rather than
compiling this information themselves — there are as many different
sets of algorithms to perform these basic coordinate transformations
as there are sources in which to find them. Unfortunately, the
computations are not trivial and the problem of which algorithms
to use can be confusing.

I have found the formulae outlined in reference [1], with some

formulation from references [2] and [3 ] the simplest for the
accuracies required in many applications today and these will be
presented here. Less complex formulation does exist for these
transformations, however, they usually do not provide the accuracy
and stability of the algorithms presented here.

I hope this paper will be useful to a broad spectrum of

individuals who at one time or another find the need to perform
these very popular coordinate transformations.

ABSTRACT

A simple to use set of algorithms for transforming geographic

coordinates into Universal Transverse Mercator (UTM) coordinates
and vice versa is presented. The formulation is based on series
expansions in. terms of the geodetic quantities n and t and are
accurate to 1 mm at ± 3 degrees from the central meridian.
Numerical examples showing both final and intermediate results
are also included. The alogrithms are outlined for use in the
northern hemisphere based on the Clarke 1866 ellipsoid as the
reference ellipsoid. However, with very minor modifications these
formulas could be used in the southern hemisphere or in conjunction
with another reference ellipsoid.

INTRODUCTION

The Universal Transverse Mercator projection and grid were

adopted in 1974 by the Ontario Ministry of Natural Resources as
the official standard geographical referencing grid for the
Province. It is called the "Ontario Geographical Referencing
Grid".

The UTM is the ellipsoidal Transverse Mercator (TM) to which

specific parameters, such as central meridians, have been applied.
Ontario, up to latitude 80° N. is divided into 4 zones each 6
degrees wide in longitude. Thus, Toronto is in grid zone 17 with
central meridian 81 W. longitude. Figure 1 illustrates the UTM
zones covering Ontario.

Each geographic location in the UTM projection is given x and

y coordinates, in metres, usually termed easting and northing
respectively. The meridian halfway between the zones* bounding
meridians is termed the central meridian (CM) and its scale is
reduced to 0.9996 of true scale. This reduction was choosen to
minimize scale variation in a given zone. If the geodetic ellip­
soid of reference is known, then the UTM zone number and the
(X,Y) coordinates (i.e., easting and northing) are sufficient to
define any point.

The ellipsoidal Earth is used throughout the UTM projection

system, but the reference ellipsoid may be subject to change.
At present, for all land under Ontarios jurisdiction, the
Clarke 1866 ellipsoid is used for the map projection. However,
the forthcomming Redfinition and Readjustment of North American
Geodetic Horizontal Control Networks (NAD83), once adopted in
Ontario, will change this ellipsoid. Until such official
announcement is made, however, the Clarke 1866 ellipsoid is the
reference ellipsoid for the UTM projection system in Ontario.

NOTATION AND GEODETIC CONSTANTS

The following notation and geodetic constatnts are used in

this paper:

a = semi-major axis of the reference ellipsoid,

b = semi-minor axis of the reference ellipsoid.
2 2 2 2
e = (a - b )/a = (eccentricity).
2 2 2 2
e' = (a - b )/b = (second eccentricity)

^ = geographic (or geodetic) latitude, positive north.

* = latitude corresponding to the meridional arc M=Y, i.e.,

* foot-point latitude.

\ = geographic (or geodetic) longitude; considered positive west of

Greenwich for purposes of these formulas.

^ = longitude of grid zone central meridian.

o
AX = X- X
o
 = radius of curvature in the meridian

= a((A
1- e2 \itA - e 2sin
)/(t ^ 4> \
)3/2

v = radius of curvature in the prime vertical

//. 2,2 -1/2
= a/(1 - e sin $ ) '
2 ,2 t
n = e' cos (j)

t = tan $
<f>i * Vi »ni f \ 1 andothersubscriptedquantities refer
to foot-point latitude, ^

absW = absolutevalue of theargument, W.

The following is a list of geodetic constants used in conjuetion

with the Clarke 1866 reference ellipsoid:

Ellipsoid = Clarke 1866

Associated
Geodetic datum = NAD27 (North American Datum 1927)
a = 6 378 206.4
b = 6 356 583-8
c = 6 399 902.551 59

0.006 768 657 997

0.006 814 784 946

(figure 1 goes here)

TRANSFORMATION OF GEOGRAPHIC COORDINATES

INTO UNIVERSAL TRANSVERSE MERCATOR COORDINATES

Geographic coordinates (<f> , X ) can be transformed into Transverse

Mercator (TM) coordinates (X,Y) by applying the general formulae
(CAUTION: the (X,Y) here are NOT UTM coordinates yet):

X = A(AX) + B ( A X ) 3+ C ( A X ) 5+ D ( A X )7

Y = M + F ( A X ) 2+ G(AX)*t+ H ( A X ) 6+ I(AX):

where AX = X - X is the difference in longitude from the

central meridian X in radians, A fB , C , D , F , G , H , and I are
coefficients given by

A = vcos<f>
 B = ( v/6)cos3cf> (1- t2 + n 2)

C = ( v/i20)cos5cf> (5 - I8t2 + t* ) +(14 n2+ 13 n1** 4 n6 >

- (58t2n2 + 6 4 t V + 24t2n6 )
D ( v/504Q)cos7$ (61 - 47912 + 179t** - t5)
(2)
F ( v/2)sin $cos 4>

G ( v/24)sin <f>cos3$ [(5 - t2 ) + (9 n2+ 4 ^ )]

H is ( v/720)sin $ cos5^ (61 - 5812 + )

+ (270 n2 + 445 n1*-*- 324 ^ + 88 ^ )

- (330t2n2 + 68012^* + 6 0 0 t v + '1 9 2 t v )

I = ( v /4 0 3 2 0 )sin $ co s7$ (1385 - 3111t2 + 543^ - te )

and M is the length of the meridional arc from the Equator to

latitude $ , and for the Clarke 1866 ellipsoid is given by

M = 6 335 034.502 24227 (1.00510 89203 88050 $

- 0.00510 89203 88050sin$ cos $

- 0.00002 16179 26721sin 3(f>cos$

- 0.00000 01138 17221sin5$ cos<f> (3)

- 0.00000 00006 5004lsin7$ cos$

- 0.00000 00000 03872sin9$ cos$

- 0.00000 00000 000243m1*$ cos $ )

and where in all the above formulae $ , X ,X , and AX are expressed

in radians and M in metres. 0

The expression for AX requires further explanation. If the point

to be transformed is WEST of the central meridian the AX value must
be made negative before using in equations (1). That is (see figure 2)

X -X , X < X (point is EAST of CM)

AX = | 0 0 (4)
h - X , X > Xq (point is WEST of CM).

With TM coordinates (X,Y) computed from equation (1) the

UTM coordinates E,N (Easting, Northing) are obtained through the
equations (see figure 2)
 500 000 + absC^ X) , X < Xq (point is EAST of CM)
E =
500 QOO - abs(kQ X) , X > XQ (point is WEST of CM).
(5)

The constant factor kQ s 0.9996 in equations (5) is called the

central scale factor and its function is to reduce the scale
distortion of the UTM projection system.

TRANSFORMATION OF UNIVERSAL TRANSVERSE MERCATOR COORDINATES

INTO GEOGRAPHIC COORDINATES

Since the formulas for this transformation require only the TM

X-value coordinate we must first translate and scale our UTM easting,
E into the corresponding TM X-value by

X = (E - 500 000)/k (6 )

Now the TM coordinates (X,Y) can be transformed into geographic

coordinates ( <f>,X ) by applying the general formulae

(p - (pi + p b2+ q b**+ r b6+ s 3s

(7)
AX = sec<f>5 ( B + T U B5+ V B7)

where 6 , P,Q,R,S,T,U, and V are coefficients given by

B (X/ vj)

Y ti + tinj2

P -( y/2)

Q ( y/24)[ (5 + 3ti ) + (ni2- Mui*) - 9ti2rii5

R = -( y/720) (61 + 9 0 t / + 4 5 t / )

+ (46m 2 - 3 m 4 + 100 m G+ 8 8 m e)

- (252t12ni2+ 66ti2n i 8 4 t a 2m 5+ 192ti2m 8)

- (90t/m 2- 225t/rj/)

S - ( y/40320)(1385 + 3633t j 2+ 40 951/ + 1 5 7 5 t/ )

(8)
 T = -(1/6)0 + 2tl2 +ni2)
U = (1/120) (5 + 28^* + 2 4 tjlf ) + ( 6 n i 2 - 3 m ^ 4m6)

+ ( S V m 2* 4t^m * + 2i»tl2n16 >

V = -(1 /5 0 4 0 )(6 1 + 662^2 + 1320t^ + 7 2 0 ^ e )

and*! is the latitude corresponding to meridional arc M=Y. This

foot-point latitude meridional is best solved for by successive
approximations using the following procedure:

* = Y/A c with M computed from (3) using* as latitude argument.

(*) 0 C1) <’)
* =* + (Y - M )/A0 c with M computed from (3) using
(2) (M (M (2)
* as latitude argument
(2)
* =* + (Y - M )/A oc with M computed from (3) using
<3) (z) (2) (3)
* as latitude agrgument
(3)
(9)

<t> = <Pi when M =Y ,

(n) (n)

where,

A0 = 1 + (3/4)e 2 + (45/64)e 11 + (175/256)e6

+ (11025/l6384)e8 + (43659/65536)e 10 (10)

+ (693693/1048576)e 12

Geographic coordinates ( ♦ ,*) can now be obtained using

equations (7) where (see figure 2)

X “ abs (AX ) * x > 500 000 (point is EAST of CM)

[
0 (11)
XQ + a b s ( AX ) » x < 500 000 (point is WEST of CM).

Equations (7) and (12) yield geographic coordinates ( ♦ ) in

radians accurate to five decimal places of one second.

(figure 2 goes here)

NUMERICAL EXAMPLES

Geographic to UTM

Given: 4,: 43-10-52.40864

X : 80-22-56.86602
X : 81-00-00 (UTMzone 17)
o
Convert ( <t>» X ) and X lnto radians,
o
A= 0.753654 544 701 0224
X = 1.402 938 636 430 972
X = 1.413 716 694 115 407
o
Hence, the difference in longitude given by equation (4) is

AX= -0.010 778 057 684 4350

because x is east of the central meridian.

Now using equations (2),(3), and (1) where the value for
2
e' can be obtained from the list of geodetic constants given
above or computed from the given equation we get

t = 0.938 446 031 811 1880

T\
= 0.003 623 573 115 4539

V = 6 388 338.711 516 329

A = 4 658 331.288 717 760

B = 50 753-553 301 7645

C - - 1 1 2 075-887 307 0132

D = -30 943.459 303 1409

F = 1 593 867.130 667 236

G = 293 232.014 772 5329

H = 13 292.810 740 7245
I = - 1 1 104.962 467 6480
M = 4 782 637.642 233 285

X = -50 207.826 872 6695

Y 4 782 822.800 200 157

Using equations (5) to compute the UTM coordinates (E,N)
the result is

E s 550 187.744 (because the point is east of the CM)

N = 4 780 909.671

UTM to Geographic

Given: E : 430 756.720

N : 4 718 544.799
X : 81-00-00 (UTM zone 17)
0
Translate and scale the UTM coordinates to obtain TM coordinates
(X,Y) using equation (6) and convert the central meridian into
radians

X = -69 270.988 395 3581

Y = 4 720 432.972 188 876

x = 1. 413 716 694 115 407

0
To compute the latitude corresponding to meridional arc M=Y

use equations (10),(9) and (3) where e1 can be obtained from

the list of geodetic constants given above or computed from
the given equation. Equation (9) represents an iterative or
successive approximation procedure for computing the foot-
point latitude. The following arethenumerical values for
the successive approximations forthis example. Thecomputed
value for meridional arc, M converged to the value for Y after
nine'iterations. Hence the corresponding value for foot-point
tude is 4>(9) The values after each iteration are:

*b
Z 1.005 108 920 378 586 (constant for all iterations)

0(1) = 0.733 829 804 470 0183

M( 1) z 4 656 460.445 817 601

0(2) 0.743 774 855 442 0517

M(2) -
4 719 753.872 357 474

0(3) 0.743 880 427 049 4506

M(3) = 4 720 425-797 458 893
0 (4) - 0.743 881 542 419 7422
M(4) z 4 720 432.896 391 271

4)(5) 0.743 881 554 203 0979

M(5) r 4 720 432.971 388 110

0(6) = 0.743 881 554 327 5834

M(5) = 4 720 432.972 180 416

0(7) = 0.743 881 554 328 8985

M(7) z 4 720 432.972 188 786
 <f>(8) = 0.743 881 554328 9124
M(8) = 4 720432.972 188 874

$ (9) = 0.743881 554 328 9126

M(9) = 4 720432.972 188 876

Since M(9) = Y the foot-point latitude is $(9).

Now, using equation (8)

tl = 0.920 232 618 023 3872

n1 = 0.003 689 994 240 3482

Vi = 6 388 127.327 689 574

3 = -0.010 843 708 154 5166

Y = 0.923 628 271 083 6740

P = -0.461 814 135 541 8370

Q = 0.289 249 457 104 6944

R = -0.216 321 910 514 2156

S = 0.191 382 158 964 9200

T = -0.449 557 689 464 7838

U = 0.383 076 309 631 1982

V = -0.397 982 026 367 1559

Geographic coordinates ( A) can now be obtained using

equations (7) and (11). This yields

4> = 0.743 827 255 447 8973

AX = -0.014 735 609 304 117 785

X = 1.428 452 303 419 525 (because the point is west of the CM).

The above values for ^ and x are in radians. After conversion

to degrees-ninutes-seeonds we have

(J> = 42-37-05.38473

X = 81-50-39.43759

CONCLUSION
 The above algorithms for transformations between geographic
and UTM coordinates are relatively simple, straightforward and
accurate. Simpler, although less accurate formulation is
available, however, and the author would be happy to provide
the same upon request. It is hoped, however, that the algorithms
given here have been presented in such a way as to both demystify
some of the mathematics involved in this computation and to present
a clear, methodical procedure for performing them.

References

1. Thomas, P.D. (1968). Conformal Projections in Geodesy and

Cartography. U.S. Coast and Geodetic Survey, Special
Publication No. 251.

2. Schmid, E. (1971)* The general term in the expansion for

meridian length. The Canadian Surveyor, 25(2).

3* Blachut, T.J., Chrzanowski, A. and Saastamoinen (1979).

Urban Surveying and Mapping. Springer-Verlag, New York.

4. Bomford, G. (1971). Geodesy. Third Edition. Oxford, London.

5. Snyder, J.P. (1982). Map Projections Used by the U.S.

Geological Survey. Geological Survey Bulletin 1532,
U.S. Department of the Interior.

6. Ontario Ministry of Natural Resources (1981). The Ontario

Geographical Referencing Grid (The Universal Transverse
Mercator Grid System). Ontario Ministry of Natural Resources,
Surveys and Mapping Branch.
figure 1

FIGURE 2