UTM & MGRS Coordinate System History

In 1994 a friend of mine made a request to the Pentagon for information on the origin of the UTM and MGRS coordinate systems. My friend handed off the documents he received to me in 2015. I've scanned them into pdf files and had them retyped so they would be both more readable and searchable. I've also replaced a missing page and line that's not in the pdf files.

The documents received date from the 1940's and clearly show that the requirements of artillery gunners were the driving force in the design of map coordinate systems.

Response from John W. Hager Geodesist with the Defense Mapping Agency

A link to a pdf of the original document

              DEFENCE MAPPING AGENCY
          HYDROGRAPHIC/TOPOGRAPHIC CENTER
            SCIENTIFIC DATA DEPARTMENT
             GEOPOSITIONING DIVISION
             GEOPOSITIONING BRANCH V


Details of the origin of the Military Grid Reference
System (MGRS) have become obscure with the passage of time.
The earliest information available in this office, the Study
and Discussion of Military Grids, Probably early 1947,
contains all the specifications of the Universal Transverse
Mercator (UTM) grid.  This study does not mention the MGRS.

A memorandum from the Commanding Officer of the Army Map
Service to the Chief of Engineers, 6 December 1946, (Encl 2
to the above study) gives examples of grid references.  It is
noted that the 100,000-meter grid square letters, an integral
part of the MGRS, are not here used in the grid references.

The letter from Brigadier Martin Hotine, Directorate of
Military Survey , (Encl 3 to the above study) states that the
specifications for a six-degree Transverse Mercator world-
wide system were communicated to the British in October 1945.

On 27 February 1948, the Joint Mapping Photography Committee
Ad-hoc Committee on Universal Military Grid Referencing
System proposed a grid referencing system containing all
the specifications of the MGRS as it is currently used.

The Joint Intelligence Committee (JIC) of the Joint
Chiefs of Staff issued JIC Papers 410/1 and 410/2, approved
14 November 1949, prescribing the use of the Universal
Transverse Mercator Grid, the Universal Polar Stereographic Grid
and the Military Grid Reference System by all branches of the
Armed Forced for joint operations use.

AGAO-S 061.3 (28 Dec. 49) CSGID-M, Issued 29 December
1949 by the Office of the Adjutant General, Department of the
Army, prescribed the use of those grids and grid reference
system for the Department of Army.

Army Map Service Technical Manual No.36, Grids and Grid
References, January 1950, elaborated the details of the
military grids and grid reference system.  This was subsequently
replaced by Department of the Army Technical Manual 5-241-1 ,
same title.  The latest elaboration of these details is
contained in Defense Mapping Agency Technical Manual 8358.1,
Datums, Ellipsoids, Grids, and Grid Referencing Systems.


JOHN W. HAGER
Geodesist

Brigadier Hotine's letter, London 1945

This is one of the early documents articulating the desire for a world-wide grid system.

A link to a pdf of the original document

                                                     Directorate of Military
                                                          Survey, War Office.
                                                        12th November, 1945.

To:-

Colonel A.G. Matthews,
Chief, Intelligence Division,
c/o ST & D (RE)
1801 K Street N.W.
Washington, D.C.

Dear Matty,

Many thanks for your letter of October 26th regarding proposed new world-
wide grid systems.

I think there can be no doubt that the polyconic is a bad military grid
because it is not orthomorphic and does not therefore give the required
degree of accuracy in rapid "plane" computations for range and bearing; or the
same facility as an orthomorphic projection in rapid small instrumental
surveys interpolated by "plane" methods.

The present haphazard plaster of grids which has grown up all over the
world (particularly in British areas of responsibility) is also a headache
even though these are orthomorphic.  They frequently lead to junctions
in awkward places, although there is no reason to suppose that the junction bug-
bear will be overcome by any cast-iron 6-degree system; or indeed by any grid
system at all.

The main point I think is that before we change at all we want to make
quite certain that we are changing in the direction of the far future and
not merely to meet some transient consideration.  It is bound to be many years
before we get on to a new system and we shall merely have had the vast labour
and confusion for nothing if by then we have changed our minds again.

We have been considering this matter at this end and have decided at any
rate to try out a "mesh" system based upon the graticule rather that the use
of a "plane" grid at all.

The advantage of a mesh system are as follows:

(a) It would obviate all grid junctions everywhere at any rate on
the smaller scale maps.  There would, however, be come
discontinuity remaining on large scale maps across the boundary of
disparate survey systems (e.g.:- the frontier of two countries)
where fundamental geographic positions are not in sympathy.

(b) It would vastly facilitate inter-service cooperation.  The Navy,
for instance, always work on some sort of graticule system of
reference if they can and are only induced to accept "plane"

grids for bombardment purposes under protest.  The Air Force similarly
hate grid junctions which always fall awkwardly for such purposes as
fighter defense.  Many anti-air defence systems cannot in any case
operate across a grid junction, e.g.:- the use of "fruit machines"
for vectoring defending aircraft.

The question has for instance arisen in particularly acute form in
relation to Coast Artillery which may otherwise be forced to switch at
very short notice between no less than three grids; one for landward
firing in support of ground operations; one for seaward firing in
conjunction with the Navy; and one at the shortest notice for employment
in an A.A. role.

The disadvantages of a mesh system may be summarized as :-

(a) Computations for range and bearing will not be simple,
although it is likely that for the most rapid purposes the
introduction of a scale factor in one direction will be sufficient:
the value of the scale factor being suitably broadcast by, for
instance, marginal information on maps.

(b) All trig. lists would have to be cast into the graticule.  We
should no longer be able to use the results of foreign surveys
neat in their own native projection.

(c) The vast cost and probable confusion over a long period of time
involved in any change.

We are, however, trying out a mesh system in experimental areas in
conjunction with the Artillery.  I do not know what the answer will be
but I certainly think that we must go into this much broader question
in detail before we make any alteration whatever.

If it is decided to stick to grid and to introduce a world-wide
system then I think the six-degree Transverse Mercator proposal is as
good as any.  One advantage of it which has probably been brought to your
notice is that the Russians do this and have adopted the six-
degree belts of the International 1/1M map.  The Germans were also
proposing to do so, We get a considerable portion of the globe already
covered for us on this system therefore (always assuming that we can get
any data out of the Russians, which is doubtful).  Conversely it may be
considered an advantage to the Russians for us to facilitate their study
of our surveys and to utilise them.  This aspect of question very
soon runs into deep water.

Although the introduction of a world system of grids such as the
Transverse Mercator proposal looks very tidy i doubt very much whether
it would work out quite as tidily.  A meridian boundary, in the nature
of things must always ignore such factors as a grid junction falling
awkwardly in a possible battle area and also such factors as the
straight utilisation of National surveys, which of course are placed on
a National projection rather than a purely geographic one.  We might
accordingly find that we had undertaken all the disadvantages of change
for very little if we were to adopt such a stereotyped proposal.

Another question that arises is the choice of a unit.  We are
likely to standardize on the metre of the grid systems.  The main reason
is that the British Army must accustom itself to training, and even
maybe fighting overseas where it will frequently get foreign maps dished
up with the least possible alteration in the shortest possible time.
In the majority of the cases foreign material and trig. data would be in
metres.  There is moreover a growing world tendency (not as yet very
evident in America) to get on to such an International unit as the
metre for basic surveys, even though the common linear units of the
country may be different.  For instance the new surveys of Great Britain
are coordinated in metres and all post-war O.S. maps will carry a metric
grid.  It would of course be a great advantage if we could both do the
same but I do not know how you view the chances of getting the metre
adopted for such purposes in America.  At first I should not have
thought the chances were very great.


If we go on to a "mesh" system the question of degrees or grades or
mils or possibly some other systems altogether will arise.  It is necessary
to have some decimal sub-division of angle for this purpose but the
centesimal system works out too small and the mil works out too large
as applied to latitude and longitude on the earth's surface.  The
sexagesimal minute is about right but is not decimally sub-divided, nor does
it spring decimally from a parent unit.  The answer may be to adopt, as
the unit tens and decimals of sexagesimal minutes.  We are adopting the
latter for our preliminary trials.

I will let you know this question progresses and shall be grateful
for any further developments your side.  I think it is important
that we should keep in close touch with one another oven though we may
not finally be able to adopt the same system.  In fact I feel a little
guilty about not having briefed you sooner.

M.HOTINE
Brig. D. Survey


*COPY*


STUDY AND DISCUSSION OF MILITARY GRIDS
by
Army Map Service
and
Military Intelligence Division, Office, Chief of Engineers, U.S. Army

A link to a pdf of the original document

                      STUDY AND DISCUSSION OF MILITARY GRIDS
                                        by
                                Army Map Service
                                       and
     Military Intelligence Division, Office, Chief of Engineers, U.S. Army

                              1. Purpose of Study
     The purpose of the following study is to determine the characteristics
required in a military grid and to select a system most nearly answering these
requirements. Marked disadvantages are inherent in most grid systems now used.
These disadvantages are complicated by the existence of many incompatible systems.


                              2. Existing Conditions
     2.01.  The polyconic military grid is prescribed by Section VII, AR 300-15,
for use on all military maps of the United States. This system is laid out in
zones 9 degrees wide in longitude with 1 degree of overlap between zones.  It
is so inaccurate at long ranges in certain directions that it cannot be used
satisfactorily for the control of the fire of coast artillery weapons or heavy
field artillery.

     2.02.  Because of this inaccuracy, the coast artillery harbor defense grid
for area in the neighbourhood of the harbor defences in the continental United
States is also authorized in paragraph 28, Change No.4, AR 300-15.  This harbor
defense grid system is a Lambert conic conformal designed particularly to
serve the guns of the harbor defense concerned.  It is not only not connected
to the military grid system in the same area but is incompatible therewith.

     2.03. Many other grid systems are in use not only in the Unites States
but also in the rest of the world.  Twenty states of the Unites States have
adopted the state plane coordinate systems measured in feet and especially
designed to serve the particular state in question.  Each such system is hardly
extensible beyond the borders of the state without the introduction of mater-
ial inaccuracies.  The enclosed map shows the overall picture of the grid
systems used in allied military operations during the recent campaigns.  (Incl. 1)

     2.04.  It is obvious that the presence of so many systems complicates map
preparation and impose material confusing handicaps on actual combat operations.
The presence of more than one grid system covering one area presents no particular
problem in peace time or on manoeuvres involving but one arm.  When the
fog of war confuses men's minds, the presence of several coordinate systems in
one area for use of different arms in fraught with potent opportunity for
disasters resulting from uncoordinations attributed to mistakes in using the
military grid.  Involved in this matter are branch pride and branch stubbornness,
each branch feeling justified in having a special grid system designed to the
particular capabilities and needs of that branch.  An example of this occurred
in the United Kingdom whore the British coast artillery, Navy and Air Force
covered the coastal area with three incompatible, incommensurable grid systems.
Intolerable confusion which resulted from the use of these grids during
the numerous German air raids in the Battle of Britain makes it highly probable that
these conflicting systems would have led to at east a few local disasters
has there been an invasion of Great Britain.

                               3. Basic Requirement
     3.01.  Primary Purpose of a Grid.  The primary purpose served by the
military grid on a map is to provide quick solutions to problems of distance and
azimuth for the firing of weapons.  It provides a quick simple means for
referring to spot locations and for designating targets.  It is an essential tool
in coordination of military operations.

     3.02. The coordination of the efforts of the many arms used on land, sea,
and air, is a problem so complex as to make mandatory a single simple solution
for problems of target designation and determination of range and azimuth.  This
requirement is believed to be so important in war that the use of a single system
of limited but adequate accuracy is held to be better than the simultaneous
use of two incompatible but otherwise more accurate systems.

     3.03. The characteristics of the using arms and weapons which affect the
design of the system to the adopted involve relatively little research.  As a
general rule, it has been assumed that permanently emplaced batteries will be
more accurate in their fire than batteries temporarily emplaced in the fields.
Therefore, the following table appears to provide sufficient criteria to
determine the desirable accuracies of the grid system adopted.

                      Probable Errors of Different Caliber
                   Permanently Emplaced Guns at Ranges Shown
 ------------------------------------------------------------------------
                                                  Minimum Probable
                Probable Errors in Yards          Relative Errors
           --------------------------------------------------------------
Range in      6" Gun      8" Gun    16" Gun
  Yards    Range Defl. Range Defl. Range Defl.    Range     Deflection
 10,000     22    2     68    3       18  3       1:555     1:5,000
 15,000     35    4     70    5       28  4       1:535     1:3,750
 20,000     52    6     73    8       40  6       1:500     1:3,333
 25,000     68    8     77   13       52  7       1:480     1:3,571
 30,000                 83   19       63  9       1:476     1:3,333
 35,000                              73  10       1:479     1:3,500
 40,000                              80  10       1:500     1:4,000
 45,000                             *77  *7      *1:584    *1:6,428
 ------------------------------------------------------------------------
* These values are appear unusual


                             4. Desirable Characteristics

     4.01. Primarily a grid system should be accurate enough for all weapons
and all military uses other than for very long distance missiles, should be
quickly applicable to any previously ungridded native, map, should yield readily
to simple computing methods and should provide simple numerical designators for
location of targets.

     4.02. Plane System. The system of coordinates desired is one with which
 all computations for the most accurate artillery firing can be simply yet
accurately performed and especially one in which the integrity of angles is
preserved.  A mathematically exact graticule, such as that presented by the
meridians and parallels, requires the use of geodetic functions to solve the
spherical triangles involves, and entails a long, time-consuming complicated
computation.  Moreover, due to the convergence of the meridians, the arc of the
parallel intercepted between any two meridians becomes shorter as the latitude
increases.  Complicated geodetic formulas would be necessary in the computation
of any distance except one along a meridian.  Complex fire control instruments
would be needed, named by personnel highly trained in a branch of advanced
mathematics.  Neither the personnel, the instruments, nor the time are normally
available.  As a consequence, the system adopted should be one in which plane
trigonometry can be employed in the solution of triangles.  In such a plane
system for general application to large area, the simplest and quickest
computations can be secured through use of a grid network of equally spaced
parallel and mutually perpendicular lines.

     4.03. Grid Accuracy. A high degree of accuracy is, of course, desirable.
However, grid accuracies which are greatly in excess of the accuracies for the
most precise weapon using the grid appear to be neither necessary nor practicable.
By reference to paragraph 3.03 above, it will be noted that the probable
errors of artillery weapons are much greater in range than are their probable
errors in deflection.  The minimum probable error of permanently emplaced
guns rarely is less than 1/555 in range and 1/5,000 in deflection.  Consequently,
a suitable military grid should be one designed to conform to these minimum
probable errors.

     4.04. Adaptability to Various Projections.  The grid system selected
should be adaptable for use of native maps without complicated recomputation
or redrafting of that map.  There are many map projections used in the making
of large scale maps throughout the world.  It is desirable to be able to
overprint the adopted grid system on any or all of these projections without
the introduction of errors in range and azimuth beyond that probable in good
artillery practice.

     4.05. Unit of Measure. Three general systems of linear measure are
commonly encountered, on maps and in grids: the metric system, the so-called
English system, and the nautical system.  Mixtures of these systems unfortunately
are prevalent.  This matter is further complicated by the fact that three
differing elements are involved - map quantities, grid quantities, and the
quantities employed by using arms and weapons.

a. Map quantities include azimuths, horizontal distances, contour
intervals, and underwater depths.  To the map user, the unit of measure
in which horizontal distances on a map are expresses is not particularly
important,  as the conversion from map distances to ground distances is frequently
done graphically against either an appropriately graduated bar scale on the
map or range scale.  However, the unit of linear measure used in the basic
survey of the map may complicate the computation and compilation of trig lists
for fire control.  This latter operation is already quite complex due to the
differing origins of longitude, datum planes, spheroids and schemes of projection,
and  other variations encountered in the native surveys of the world.
Thus, the conversion from one unit of linear measure to another incommensurable
unit adds an operation subject to mistakes and affecting final accuracies.

b. Contour intervals and spot elevations should, but may not, be
in the same unit of measure as the horizontal distances, in order to provide
for the ready calculation of true slant ranges, defilade, mask, profile, etc.

c. Underwater depths are generally expressed either in meters
or in fathoms, although shoal water depths may be also expressed in feet.
It is highly desirable that those units of measure be the same as the horizontal
unit in order to be readily useful in the computation fo underwater profiles
and beach gradients.

d. The military grid, while essentially concerned with angles
and horizontal distances, must be precisely related to the computed geographic
positions.  The necessary correlation between the vertical unit of measure and
the horizontal unit of measure on the grid as indicated in b above, is essential
for the quick solution of problems involving defilade, mask and true gun target
distance.

e. The using arms and weapons are not entirely coordinated in
the units of measure employed in the laying of the piece. The Coast Artillery
measures azimuths in degrees and hundredths of a degree from grid south as the
origin.  The Field Artillery measures angles in mils, with filed orientation
of base circles.  The Coast Artillery measures its range in yards, while the
Field Artillery may measure it in yards or meters.  Due to the tangent relationship
of the mil, Field Artillery can readily transpose from angular measure to
linear distance in either meters or yards.  Each artillery weapon is served
and laid by employing a multiplicity of tabular information, plotting tools,
and gunner's instruments.  All these things must be related to the unit of
measure selected for map and grid quantities.  At the present moment, due to
the recent tremendous concentration of field artillery weapons in the European
campaigns and to the use of the metric system throughout in that area, our
Field Artillery is well equipped and trained in the use of the metric system.
The Seacoast Artillery of the United States, including the Panama canal and
Oahu, are not so equipped or trained.  They use the yard-hundredths of a degree
system, except in the coast defenses of San Diego where the coast Defense
grid is graduated in feet rather than in yards.

f. Existing Conditions.  The majority of the large scale maps
of the world are made on the metric system.  Exceptions to this rule are the
United States, Canada, Australia, United Kingdom, Union of South Africa,
India, Melanesia, and Middle East.  The following table shows the unit of
measure of native maps and military grids employed in operational areas of
the recent campaigns.


                               MAP UNITS                  GRID UNITS
            -------------------------------------------   -------------------
AREA        HORIZONTAL UNIT    VERTICAL    DEPTH UNIT     GRID   GRID TYPE
            (Map Bar Scale)    UNIT (Map   (Bathymetric   UNIT
                               Contour)    Contour)
-----------------------------------------------------------------------------
1. France        Meter         Meter       Meter          Meter  Lambert
2. Germany       Meter         Meter       Meter          Meter  Trans. Merc.
3. Italy         Meter         Meter       Meter          -----  -----
4. Tunisia       Meter         Meter       Meter          -----  -----
5. Libya         Meter         Meter       Meter          Meter  Trans. Merc.
6. Egypt         Meter         Meter       Fathon         Meter  Trans. Merc.
7. Okinawa       Meter,Cho     Meter       Meter          Meter  Unknown
8. Burma         Miles         Foot        Fathom         Yard   Lambert
9. China         Li,meter      Meter       Fathom         Yard   Lambert
10. Russia       Meter         Meter       Foot           Meter  Trans. Merc.
11. Hawaii       Mile          Foot        Fathom         Yard   Polyconic
12. Philipines   Mile          Foot        Fathom         Yard   Polyconic
13. Poland       Meter         Meter       Meter          Meter  Stereographic
14. Holland      Meter         Meter       Meter          Meter  Stereographic
15. Belgium      Meter         Meter       Meter          Meter  Bonne
16. Japan        Meter,Cho     Meter       Meter          Meter  Trans. Merc.


     4.06. Width of Zone.  An inherent fault of any system of plane coordinates
applied to the surface of a spheroid is the fact that  inaccuracies increase
as the zone is extended east and west or in other cases, as the belt is extended
north and south.  It is also desirable, although not entirely necessary, to keep
at a minimum the deviation of grid north from true north.  This deviation likewise
increases materially as the distance from the central meridian increases.
These inaccuracies can be kept within reasonable limits by the adoption of narrow
grid zones.  It has been stated only semi-facetiously that there are three
military engineering axioms:

    a. It always rains in war.
    b. It's always too cloudy to get aerial mapping photographs.
    c. Battles are always fought on grid junctions.

This Inst axiom is spoken from the depths of bitter experience, rendering it
obvious that a grid zone should include as much area as possible so as to
obviate too frequent junctures between grid zones on the battlefield.  However,
this desirable criterion cannot be widely applied without including intolerable
inaccuracies of the grid.  Elimination of the undesirable features consequent
upon fighting a battle on a grid juncture can be partially accomplished
by providing for overlap between grid zones.  The 9 degree width of the military
grid system presently prescribed for use in the United States introduces
appreciable inaccuracies near the edges.  A reduction in width to 6 degrees betters
this situation materially.  The computation of ranges and azimuths where the
gun position is located in one grid zone and the target in another can be
provided for by a half degree overlap between grid zones, enabling the coordinates
of the gun zone to be extended a half a degree into the grid zone in which the
target is located.

                                   5. Comparison of Grids

     5.01. Polyconic Grid.  For military purposes, a grid may be regarded as
a set of perfect squares ruled on a plane map, scale 1:1, and then transferred
to the earth's surface.  Evidently after being transferred to the earth's
surface the squares will no longer be perfect; and distortions they will have
received in being put on the surface of the earth will reflect the distortions
of the projection used for the map.

           a. The polyconic projection is defined as one which the central
 meridian and all parallels are mapped to scale and with true curvature.  All
 other lines are stretched, the amount increasing as the square of the distance
 from the central meridian, and being greatest for north-south lines.  Angular
 errors also appear, increasing with distance from the central meridian.  Of
 course it is possible to compute these errors, at least roughly, and to allow
 for them, and this is regularly done by engineer survey troops.  But the
 corrections are generally considered beyond what is to be expected of artillery
 units in the field, and for that reason all mention of them is omitted in
 artillery technical manuals, even when survey procedures are discussed.
 It is proper, therefore, to compare the errors of the projection as shown in the
 following table, directly with the errors of the guns (Section 3.03).


                      Errors of U.S.Military Grid at 4°30'
                      from Central Meridian, and 30° Latitude
-------------------------------------------------------------------------------------------------
         |                            |                             |             True
         |       True  Azimuth 0°     |      True Azimuth 45°       |          Azimuth 90°
True     |----------------------------|-----------------------------|----------------------------
Range    | Range  Rela- | Defl. Rela- | Range  Rela- | Defl.  Rela- | Range  Rela- | Defl.  Rela-
(yds)    | Error  tive  | Error tive  | Error Defl.  | Error  tive  | Error  tive  | Error  tive
         | (yds)  Error | (yds) Error | (yds) Error  | (yds)  Error | (yds)  Error | (yds)  Error
---------|--------------|-------------|--------------|--------------|--------------|-------------
10,000   |  23    1:435 |  0      0   |  12   1:836  |  12    1:836 |  0       0   |  0       0
30,000   |  70    1:428 |  4   1:7500 |  35   1:859  |  35    1:859 |  0       0   |  0       0
40,000   |  70    1:428 |  4   1:5000 |  47   1:850  |  47    1:850 |  0       0   |  0       0

It is evident that at 40,000 yards, and an azimuth of 45°, the error of the grid
in deflection is almost five times the probable error of a 16-inch gun.  Errors
of the above amount are characteristic of the s0-called non-conformal projections,
i.e., those in which the shaped, as well as the scale of small areas is distorted.
Such non-conformal grids were widely used prior to World war I, but most of them
have been abandoned in recent years, chiefly, no doubt, because of the breakdown
of the old French Bonne grid during the War.  In addition to its lack of comformality,
the polyconic projection possesses the disadvantage that it has not been
studied thoroughly from a mathematical standpoint.  Hence the small corrections
needed for precise surveys are not known; the transformation from other
grids to be polyconic is not known; not even the transformation from one
polyconic belt to another has been studied.

           b. So far as grid junctions are concerned, the polyconic
is theoretically excellent; it can be extended indefinitely both north
and south, so that the world can be divided up into meridianal strips.
In practice, the present World Polyconic has two latitudinal junctions,
one near 24°, due to failure to put the origin of the old U. S. grid
sufficiently far south; the other is near 49°, and is due to inaccuracy in
the old tables which amounts to 1.1 meters.

           c. The polyconic grid is not well suited for foreign maps
due to its lack of conformality.  By the older laborious hand methods
of grid plotting, this introduced no difficulties other than laborious
calculations; but the use of the coordinatograph, for rapid plotting of
grids and projections, requires a conformal grid.

     5.02. Cassini-Soldner Grid.  This is a non-conformal grid,  very
similar to the polyconic, and open to the same objections.  It differs
only in that the grid east-west lines,  rather than the parallels are
represented to true scale and with true curvature.

     5.03. The Bonne Grid.  Both meridians and parallels are represented
to true scale on the Bonne projection; the error shows up on lines which
run NE to SW, or NW to SE.  It is just as bad as in the case of the other
non-conformal projections; and this projection has been generally abandoned.

     5.04. The Stereographic Grid.  The stereographic grid may be briefly
defined as a conformal grid (that is, one having zero angular distortions
for small distances, and very small angular distortions for any distance)
in which the scale error is Zero on a standard circle.

This grid is conformal  insures  that  within about 200 miles of the center-
point, the error in deflection will be less than the error of the most
accurate guns.  The range error is considerably larger than the deflection
error , but is considerably less  than  the range error of permanently
emplaced artillery.  Unfortunately,  this  grid cannot be extended more
than 300 miles from its center point and grid zones are circular in
Shape.  It is quite suitable for small roughly circular countries such
as Poland, Holland or Romania, which use it ; but on a world-wide basis
It would lead to a great multiplicity of grid junctions, and  points where
three or more grids meet could not be avoided.

     5.05. Lambert Grid.  Like the  stereographic, the lambert is a
conformal projection.  It may be defined as a conformal projection in which
the scale errors are zero along two parallels .  It is well suited to the
mapping  of moderately large areas , and has been extensively used by the
British and French, especially in recent years.  The numerical values
of the errors are similar to those for the Transverse Mercator given in
the next paragraph.  The errors are very small in deflection, as in necessary
for artillery grid.  It is readily adaptable to gridding maps
On other projections.  In general a slight change of scale only would be
Required.  On the other hand, the rapidly increasing grid declination
makes it impracticable to extend the grid more than 15° from the
central meridian (except along the Equator , where grid declination is
always zero).  For this reason, the British and French were forced to
introduce numerous zones, and to permit junctions of three grids at the
same point.  These numerous grid zones necessary, both north-south and
east-west, make the Lambert undesirable for extensive coverage

     5.06. Transverse Mercator Grid.  The Transverse Mercator grid may
be defined as a conformal grid in which the central meridian is represented
by a straight line at true scale.  it is well suited to largo areas, and
is being used by the Germans, Russians, British, and Japanese.  The errors
at a given point vary little in all azimuths and average values for
different ranges are given in the following table for 30° north latitude and
for 3 1/2° from the central meridian.


         -------------------------------------------------------------------
         |        | Transverse                  |   Transverse             |
         |  True  | Mercator Grid               |   Mercator Grid          |
         |  Range |     Range        Relative   |   Deflection    Relative |
         |  (yds) |     Error          Error    |   Error (yds)     Error  |
         |        |     (yds)                   |                          |
         |--------|-----------------------------|--------------------------|
         | 10,000 |      10           1:1000    |       0             0    |
         | 30,000 |      30           1:1000    |       3          1:10000 |
         | 40,000 |      40           1:1000    |       5.8        1:6900  |
         |--------|-----------------------------|--------------------------|


Obviously, the grid is well suited to artillery purposes since the
inevitable errors ere thrown into range rather than deflection.  The grid
can be extended indefinitely in latitude like the polyconic.  Hence it
never necessary to have a grid junction involving more than two grids
(except, of course, near the Poles).  Transformation of coordinates from
one bolt to another can be done by a formula already worked out.  The
formula is always the same, and is very simple in character.  The grid
declination will remain moderate throughout the belt.  The grid can be
readily adapted to use on other projections.  Much theoretical work has
already been done on this subject by a large group of mathematicians,
including especially Professor W. K, Hristow.  Extensive computations,
especially for the Balkan countries, were done by the German High
Command (O. K. H. ) during World War II which could, in an emergency, be
promptly utilized if the proposed projection is adopted.  In addition,
much geodetic data of foreign areas on file at Army Map Service are on
this system.  The projection is well suited for converting data on
various spheroids to a common basis.  Transverse and lower orders of
triangulation may be computed and adjusted directly on the grid due to its
conformality.  This feature, which is a large saving in field and office,
is not practicable where a non-conformal projection such as the present
Polyconic is used.

     5.07.	Tabular Comparison of Grids.
    ____________________________

   --------------------------------------------------------------------------------------
   GRID SYSTEM           APPLICABILITY          GRID           MAXIMUM         MAXIMUM
                         TO FOREIGN MAPS**   JUNCTIONS         RELATIVE        RELATIVE
                                                             RANGE ERRORS*   DEFLECTION
                                                              (45,000 yds)      ERRORS*
                                                                             (45,000 yds)
   --------------------------------------------------------------------------------------
   Polyconic             Very Poor          Few & simple       1/1228          1/2454
   Cassini-Soldner       Very Poor          Few & simple       1/1228          1/2454
   Bonne                 Very poor          Many & complex     1/2454          1/615
   Stereographic         Good               Many & complex     1/2454          1/15,360
   Lambert               Good               Many & complex     1/2416          1/7,663
   Transverse Mercator   Excellent;
                         much work
                         already done       Few & simple       1/2416          1/7,663
   ---------------------------------------------------------------------------------------
** A grid system is considered applicable to a foreign map if it can be put on
most maps without changing map or grid except in scale.

* Range and deflection errors are maximum values within 160 miles from the center
of the projection, whether the center is a line, as in the Transverse Mercator,
Lambert, Cassini-Soidner, and Polyconic, or a point as with the other two.  The
figures are based on GSGS "Survey Computation".   160 miles is the approximate
distance (at 40° of latitude) from the center of the proposed Transverse
Mercator zones to the junctions, about 3° of longitude.  The actual maximum errors
of the present world Polyconic grid are considerably larger, since the grid
zones are 9° in, width.

                           6. Conclusions as to System
                                  to be Adopted.

     6.01. The Lambert Orthomorphic projection is conformal but is not
suitably as it requires grid zone junctures both north and south and east
and west.  The polyconic grid system now prescribed for use as military
grid on all maps of IJ. S. is inaccurate in both azimuth and distance.
The greater inaccuracy is in azimuth and is more than the probable error
in deflection of permanently placed guns.  The transverse mercator grid
is conformal and is immediately applicable without plottable error, to
the majority of the map projections commonly encountered on the native
maps of the world.  The transverse mercator grid reduces inaccuracies
to a point where they are compatible with the accuracies required by all
modern artillery weapons.  This grid is sufficiently accurate to
eliminate the necessity for a special Coast Artillery grid in the
vicinity of coast defense locations.

     6.02. In view of the foregoing, a military grid system based on.
the transverse mercator projection applied to the local spheroid and
measured in meters, or in the standard unit of the country concerned,
should be applied in zones running from Latitude 80° N to Latitude 80°
S, 6 degrees of longitude wide, with one degree of overlap (1/2 degree
each side).  The latitude of the origin is the equator.  (Incl #2).
The false casting to be applied for each zone would be 500,000 meters
or yards.  Scale factor on the C. M. should be 0.9996.  The zones should
be numbered, commencing with Zone 1, with its western edge at 180°
longitude, running east to 174° west longitude.  Consecutively numbered zones
continue eastward by successive steps of 6 degrees until reaching the
point of beginning; these number designations being identical to the
I.M.W 1:1,000,000 layout.  It will be noted that in certain countries
where the native maps use the English units in elevations.  and contours,
such as the U. S., Canada, Australia, India, the proposed grid system
should be graduated in yards rather than meters.  In certain training
areas in the U. S., both metric and yard grids will be required for
training purposes.  Where the metric Grid is used in the domestic U.S.,
the spot elevations should be in meters and the contours should be
converted to a metric interval, provided that such conversion of contours
shall be limited to those maps to be used for metric training purposes.


                           7. Proposed Specifications

Projection:                Transverse Mercator.

Spheroid:                  Same as that used to compute the triangulation
                           of the area.

Unit:                      Meter in most areas; yard in U. S. and other
                           areas where the English system is firmly
                           established.

Central Meridians:         3°E (or W) of Greenwich and every 6° thereafter.

Latitude of Origins:       0°

False Easting:             500,000 meters (or yards) .

False Northing:            0 for Northern hemisphere
                              10,000,000 for southern hemisphere.

Scale Factor:              0.9996

Zone Width:                6° of longitude (plus 1/2° overlap at each edge)

Limits of Zones:           North  80° latitude
                              South: 80° latitude

Zone Numbering;            Commencing with Zone 1 east of 180° longitude, and
                           continuing easterly around the earth.  (Identical
                           to I. M. W. system designation)

Limits of Tables:          North: -80°  latitude
                              South: -80° latitude

                     8. Implementing Actions and Costs

     8.01. It should be noted that the application of the transverse
mercator grid system should be progressive rather than instantaneous.

Priorities for conversion are indicated as follows:

            a. Military areas in the United States,
            b. Mapping and map revision of foreign areas embraced in
the 20-year strategic mapping plan approved by the War Department.
            c. General areas of the United States as stocked by the Army
Map Service.
            d. Other maps of foreign areas as reissued.
Upon the issue  of the new map in any area, the new grid will normally
be shown in full, but to safeguard against the event of the occurrences
of an emergency while a series is still in a state of partial conversion,
the maps will carry marginal marks to permit easy plotting and
overprinting of the old grid.

     8.02. Since Amy Map Service and map depots in the theaters now
hold extensive stocks of maps carrying the expedient war-time grids, It
is to be expected that the proposed standardization of the transverse
mercator grid automatically renders obsolescent these stocks.  The cost
of a conversion in this respect is estimated as follows:

            a. Cost of conversion of points and correction of drafting copy

               (1) United States areas 	$99,500
               (2) European areas 		$103,000
               (3) Other overseas areas 	$22,000

            b. Cost of replacement of stocks to be retired
               (1) United States areas  	$ -0-
               (2) Overseas areas 		$10,000

Since all United States maps are to ho converted to military scales
(1/25,000, 1/50,000), new stocks are to be prepared in any event.
Overseas accumulation of war time maps, it has been reported, are being
salvaged except for small reserves which will probably not be replaced.  The
$10,000 figure should be ample to provide for all requisitions directly
attributable to the change in grid.  The cost of now authorizing the
proposed conversion is properly to be weighed against the much greater
cost that would be borne should circumstances require the conversion
ten years hence.  The conversion must ultimately be made in view of the
inadequacies of the present medley of grids to suffice for the anticipated
requirements of another war.  It is considered unquestionable
that the cost of the conversion should be accepted now if it be agreed
By the General Staff that the transverse mercator grid is in fact the
correct design for the future, based on what can now be discerned as to
future characteristics requirements.

     8.03. It is suggested that the views of the Navy Department be
obtained prior to final standardization in view of the application of the
military grid to maps and charts for amphibious operations.  It would
also be desirable to coordinate the design so far as possible with the
Director of Military Survey, War Office, London, who has expressed his
general views in an informal letter (Incl. #3).

ARMY MAP SERVICES
Recommendations for Military Grids
6 December 1946

A link to a pdf of the original document

                                 ARMY MAP SERVICES
                           Corps of Engineers, U.S. Army
                                                             6 December 1946

MEMORANDUM TO:  Chief of Engineers
SUBJECT:        Recommendations for Military Grids

      1. Included herein are certain recommendations for grid numbers and
grid references.  A pertinent discussion and background information follows
each recommendation.

      2. The Universal Transverse Mercator Grid supersedes approximately
85 previously used grids which made up an undesirable heterogeneous system.
This modernization should be further expanded to revise the outmoded standards
for grid numbers and grid references which were designed primarily for fire
control purposes only.  The inadequacy of the present system for use in making
general grid references became apparent during the past war.  To satisfy
their needs it was necessary for the individual theater commanders to devise
new methods to fill this deficiency.  There was an unfortunate lack of
consistency for the systems used varied with the theater.  Before formulating
these recommendations an exhaustive study was made of all the war-time
provisional systems.  The best of each is incorporated within these recommendations,
which if adopted would assure a standard but simple fool-proof system, designed,
at the same time, to accommodate changing techniques in warfare.

      3. GRID INTERVALS

         a. RECOMMENDATION - It is recommended that grid intervals be:

              Maps 1:5,000 and larger               1,000 yards (or meters) with
                                                    grid lines ticked at 100 yard
                                                    (or meter) intervals.

              Maps 1:10,000)
                   1:25,000)                        1,000 yards (or meters)
                   1:50,000)

              Maps 1:100,000)
                   1:250,000)                       1,0000 yards (or meters)

         b. The grid intervals authorized at present are:

              Maps larger than 1:5,000              100 yards

              Maps 1:5,000 to 1:63,360
               inclusive      			                   1,000 yards

              Maps smaller than 1:63,360 to
               larger than 1:100,000                5,000 yards

              Maps 1:100,000 to larger than
               1:400,000		                          10,000 yards

              Maps 1:400,000 to 1:500,000
               inclusive      		                    50,000 yards


         c. The intervals of British grids are:

              Maps larger than 1:5,000		            100 meters (or yards)

              Maps 1:5,000 to 1:100,000
               inclusive*	 		                      	1000 meters (or yards)

              Maps smaller than 1:100,000* to 	     10,000 meters (or yards)
               1:500,000

         d. It is noted that according to the new edition of AR 300-15, authorized
            map scales are:

              Small scale                           (1:1,000,000

              Medium Scale                          (1:250,000
                                                    (1:100,000

              Large scale                           (1:50,000
                                                    (1:25,000
                                                    (1:10,000
                                                    (1:5,000

         e. The war proved that generally the British grid intervals were
            superior.   The U.S 5,000 yard interval was awkward and confusing
            in as much as the abbreviated reference for a common point on maps
            of different scales would be dissimilar in all instances.  Authorization
            should be granted to revise the grid intervals to overcome
            this defect and to make them more compatible to the revised
            authorized map scales.

      4. GRID REFERENCES

         a. RECOMMENDATION - To satisfy particular needs two types of grid references
            should be made standard: general references, and fire control references.

            (1) General reference - Such a reference should generally consist
                of the grid zone designation by a group of numbers
                expressing the E and N coordinates of the referred point;
                examples:

                30 NC 80432864       27 SF 69143872     (1,000 unit reference)

                30 NC 804286         27 SF 691387       (10,000 unit reference)
-----------
      *In all except Europe, AMS sheets of 1:100,000 falling in British Grid
      areas were gridded at 10,000 meter (or yard) intervals

                (a) Grid zones

                    1. Zones for the Universal Transverse Mercator grid are
                       identified with the IMW column (6° E-W) numbers,
                       starting at the international date line (180° meridian)
                       and reading 1 to 60 in an easterly direction.  (See
                       attached index).  To prevent similar references for
                       points 1,000,000 units apart (north-south) the IMT
                       row letters preceded by N (for north) or S (for south)
                       should be incorporated within the system and added to
                       the zone number designation.  Under the IMW plan, each
                       row (4° N-S) is assigned a letter of the alphabet
                       starting from the equator, preceding in both directions.

                    2. To assure proper identification each sheet should carry
                       in its grid reference box its complete zone identification.

                    3. Within an area assigned to an army the grid zone
                       designation may be deleted at the discretion of the
                       Commanding General for reporting within the grid zone
                       providing sender and receiver are not more than 500
                       miles apart in a N-S direction.  (The zone designation
                       is necessary in such a case since the numerical reference
                       is the same at 1,000,000 units in a N-S direction).
                       For reports to higher headquarters, however, the complete
                       reference must be given.

                (b) Numerical reference - To facilitate making such references
                    from a 1,000 unit grid, a reference should be simply an
                    eight digit number; for example: 80472866.  The "804"
                    represents the 100,000, 10,000 and 1,000 digits of the
                    easting grid line to the west of the referenced point,
                    the "7" represents the estimated tenths from the easting
                    grid line to the point, the "286" represents the 100,000,
                    10,000 and 1,000 digits of the northing grid line south
                    of the referenced point, and the "6" represents the
                    estimated tenths from the northing grid line to the point.
                    To maintain a relationship between similar grid references
                    from different scale maps, a reference from a 10,000 unit
                    grid should be a six digit number; for example: 804286.
                    The "80" represents the 100,000 and 10,000 digits of the
                    easting grid line to the west of the referenced point,
                    the "4" represents the estimated tenths from the easting
                    grid line to the point, the "28" represents the 100,000
                    and 10,000 digits of the northing grid line south of the
                    referenced point, and the "6" represents the estimated
                    tenths from the northing grid line to the point.

            (2) Fire control references to be used within the sphere of the
                equivalent of one adjacent 1:50,000 sheet in all directions
                (two 1:25,000 sheets, four 1:10,000 sheets, etc.) - Existing
                methods for determining grid references for fire control
                as outlined in FM 6-40, Part Four, Chapter 2, should be
                retained with but one modification: Sheet name designations
                should never be used.  When a grid reference is being sent
                to a station outside the sphere of the equivalent of one
                adjacent 1:50,000 sheet in all directions (for example: a
                long range gun), then the full grid reference should be
                sent preceded by the zone designation (see (1) above); for
                example:

                                  30 NC (804.72-1286.68)

         b. Existing regulations (See FM 6-40, Part Four, Chapter 2) designate
            the following methods for reading a grid reference:

            (1) Designation of sheet, parenthesis, X coordinate, decimal,
                location to nearest yard, a dash, Y coordinate, decimal,
                location to nearest yard, parenthesis.

                Example: Annapolis (804.729-1286.684)

            (2) When the map is definitely understood, its designation may be
                omitted.

                Example: Annapolis (804.729-1286.684)

            (3) If the location to the nearest 10 or nearest 100 yards only
                is desired, or if the measurements cannot be made with greater
                accuracy, the digits indicating units or tens may be omitted.

                Examples: (a) (804.72-1286.68) to nearest 10 yards

                          (b) (804.7-1286.7) to nearest 100 yards

            (4) For expediency it is permissible to include only two digits to
                the left of the decimal point (10,000 and 1,000 digits),
                omitting any preceding digits.

                Examples: (a) (04.729-86.684)

                          (b) (04.72-86.68)

                          (c) (04.7-86.6)

            (5) If the point is fixed within an area 10,000 yards square, only
                one digit need be given before the decimal point of each
                coordinate.

                Examples: (a) (4.729-6.684)

                          (b) (4.72-6.68)

                          (c) (4.7-6.7)


            (6) If a large number of points are being designated by the
                abbreviated coordinates shown in example (c), the decimals
                and dashes may be omitted and the reference given as (4767).

         c. References for British Grids are read according to the following
            methods:

            (1) Maps bearing a 10,000 unit interval (1:100,000 to  1:500,000);
                letter of 500,000 unit square (written as a small capital
                letter), letter of 100,000 unit square (written as a large
                capital letter), 10,000 digit of easting line to the left of
                the point, estimated tenths (1,000 units) eastward to point,
                10,000 digit of northing line south of the point, estimated
                tenths (1,000 units) northward to the point.

                Example:	    cA1428

                This locates point to nearest 1,000 units.

            (2) Maps bearing a 1,000 unit interval (1:5,000 to 1:100,000 inclusive):
                Letter of 100,000 units square, 10,000 and 1,000 digits of easting
                line to left of point, estimated tenths (100 units) eastward to
                point, 10,000 and 1,000 digits of northing line south of the
                point, estimated tenths (100 units) northward to the point.

                Example : 			A143286

                This location point to nearest 100 units.

         d. During the war, the Pacific and Southwest Pacific Commands found it
            feasible to use a system for reading general grid references
            similarly to that used with British Grids.  Apparently, a broad
            interpretation of existing regulations was made to find authority
            for the change.  The name of the map is not mentioned (authority:
            see 4 b (2) above); digits to the left of the 10,000 and 1,000
            unit digits are omitted (authority: see 4 b (4) above); decimals
            and dashes are omitted (authority: see 4 b (6) above).

            (1) To read a reference point on a map employing a 1,000 unit
                interval read: the 10,000 and 1,000 digits of the casting
                line to left of point, estimated tenths (100 units) eastward
                to point, the 10,000 and 1,000 digits of the northing line
                south of the point, estimated tenths (100 units), northward
                to the point.  Write as a 6 digit continuous number.

                Example: 		143286

            (2) A similar procedure is followed in reading a reference on a
                map using a 10,000 unit interval, except that the digits for
                the grid lines are for the 100,000 and 10,000 units (the
                last four digits being omitted) and the estimated tenths
                represent 1,000 units.  Thus, a reference for the same point
                cited in (1) above, might read: 214128

         e. Discussion of recommendation (par. 4 a (1), above) for:  General
            references.

            (1) Experience in the Pacific Theater proved that general grid
                references were frequently used.  Usually it was unnecessary
                that these general grid references possess the same accuracy
                as that required for fire control purposes.  It was deemed
                sufficient to identify any general grid reference to the
                nearest 1/10th of the grid interval (i.e., 100 units at a
                1,000 unit grid interval; 1,000 units at a 10,000 unit grid
                interval).  The system which employed a continuous six
                digit number as a grid reference (example: 143286) proved
                highly successful.  Its principal merits wore simplicity and
                intelligibility.  The standard method for reading grid
                references )see par. 4 b (1), above) was primarily designed for
                fire control purposes and when used for general purposes becomes
                very awkward.  This was the experience in the Pacific Theater
                which found that sheet name designations, parentheses, decimal
                points and hyphens were superfluous and only increased the time
                necessary to reading and sending general grid references.

            (2) There were two faults with the Pacific system: danger of
                confusion between a reference taken from a map bearing a 1,000
                unit interval and from one of a 10,000 unit interval, in as much
                as both were six digit numbers; and lack of connection between
                references for a common point taken from a 1,000 unit grid and
                from a 10,000 unit grid, in as much as in reading a reference from
                a 1,000 unit grid the principal digits were the 10,000 and 1,000
                ones and for a reference from a 10,000 unit grid the principal
                digits were the 100,000 and 10,000 ones.  Thus, references for
                a common point might read: 047866 (from a 1,000 unit grid) and
                804286 (from a 10,000 unit grid).

            (3) Under the ANS proposal these faults would be eliminated.  An
                eight digit reference would immediately be recognized as
                being from a 1,000 unit grid, and a six digit as being from a
                10,000 unit grid.  Further, a coordination would exist between
                references from different unit grids for common points, as:

                80432863 	           (reference from 1,000 unit grid)

                804286		             (reference from 10,000 unit grid)

            (4) Normally, under the system as proposed it is required that
                in referring a point the entire reference be given - grid zone
                designation and numerical coordinates.  In reporting in a
                single grid zone between points not more than 500 miles apart
                in a N-S direction designation of the zone is unnecessary.
                Consequently, if he is certain that no confusion will result,
                the theater commander should be permitted to issue instructions
                to omit the zone designation from grid references.  However,
                in reporting to higher headquarters, between grid zones, and
                between points in the same grid zone, more that 500 miles apart
                in a N-S direction, the grid zone designation should never
                be omitted.  The 500 mile rule is required since a reference
                will read the same for points which are 1,000,000 units apart
                in the same grid zone.

            (5) With the new type of warfare in which activities are far-flung
                it is important that references given in communications identify
                the area.  The use of a sheet name as presently required by
                regulations is inadequate for the receiver would generally
                expand too much time searching map catalogues and indices to
                identify the locale of the sheet.  To introduce such a reference
                with only the designation number of the grid zone would
                require the use of 1,000,000 digits in the numbers.  This is
                not desirable as it would mean that the numerical reference
                would differ from that used for a local general reference,
                and would also require the use of decimal points and hyphens
                since the 1,000,000 digit might occur only with one coordinate.
                The solution is to introduce such a grid reference with the
                designation number of the grid zone followed by a sub-zone
                letter designation. (See paragraph 4 a (1) above).  This
                makes an absolute identification.  Its use would simplify
                the overall grid reference system in that the numerical
                reference would be the same for both an abbreviated general local
                reference and for a reference used in official communications
                to higher headquarters.  The use of the sub-zone letter
                designation does not create a new system but makes complete
                utilization of the entire IMW numbering system whose row
                numbers are the basis of the numbering of the zones of the
                Universal Transverse Mercator Grid.  (See attached diagram).

         f. Discussion of recommendation (par. 4 a (2), above) for: Fire
            control references - The system presently in use is generally
            quite adequate for its purpose.  However, it is deemed more
            desirable to use grid zone designations instead of sheet
            identifications.

            (For arguments see paragraph 4 e above).

      5.GRID REFERENCE BOX

         a. RECOMMENDATION - It is recommended that the grid reference boxes
            used in foreign areas on AMS maps be made standard practice for
            use on all maps including areas in the United Stats.  The grid
            reference box should contain instructions for determining a general
            reference.

         b. The inclusion of a grid reference box in the margin will assure
            standard renditions of grid references, eliminating any reference
            to military manuals by personnel unfamiliar with grids.

         c. It is felt that grid references for fire control come within the
            category of special purpose and are not as widely used as general
            grid references as they generally are limited to artillery use.
            Consequently, the method of determining such references need not
            appear in the grid reference box but should be explained in proper
            military manuals (see par. 7).  If considered necessary, reference
            to such manuals could be included in the grid reference box.

      6. GRID NUMBERS

         a. RECOMMENDATION - It is proposed that a modification of the so-called
            Canadian Grid Numbering System be made standard on all maps published
            by the Army Map Service.  Under the system, grid numbers would appear
            on all four sides of a sheet labelling each grid line, and "principal
            digits" would be shown on the face of the map labelling each grid
            line, appearing east or north of every accentuated grid line (every
            even tenth line - 10,000 on a 1,000 unit grid and 100,000 on a 10,000
            unit grid).  On a 1,000 unit interval grid except for the values
            shown in the southwest corner the last three digits of each grid
            number are omitted and the principal digits (100,000, 10,000 and
            1,000) appear larger than the 1,000,000 digit; for example:

              (corner)
            1 276 000 yds.	1 277 	1 278

            Numbers for a 10,000 unit grid appear in a similar manner, except
            that the last four digits odf wach number are omitted; for example:

              (corner)
            1 27 0000 yds.	1 28	 1 29

         b. Existing standard practice requires that on grids of intervals of
            1,000 units the last three digits be omitted and that on grids of
            10,000 units the last four digits be omitted.  Regulations do not
            specifically limit the appearance, frequency or location of the numbers.

         c. The advantages of this numbering system are apparent: the numbers
            on the face of the map materially aid the map user in reading the
            grid and in determining references; the use of superior type around
            the border accentuates the principal digits (100,000, 10,000 and
            1,000) materially aiding the map user in making general grid
            references.

         d. Under the system used in the Pacific for 1,000 unit grids, numbers
            appeared on all four sides of the sheet labelling each grid line.
            The last three digits of each number were omitted and the principal
            digits (10,000 and 1,000) appear larger than the 1,000,000 and
            100,000 digits, as for example:

            (corner)

            12 76 000 yds.	12 77	12 78


            The principal digits also appeared on the face of the map labelling
            each grid line appearing at 10,000 unit intervals east or north of
            every 10,000 unit grid line (which are accentuated in weight).
            The modification to this system recommended in a, above, is
            necessary since the 100,000, 10,000 and 1,000 digits would appear
            in a grid reference as recommended in par. 4 a (1), above.

      7. GRID MANUALS

         a. RECOMMENDATION - Subject to approval of the recommendations appear-
            ing in the preceding paragraphs, a recommendation is made that
            the Army Map Service be directed to prepare a new military manual
            covering the subject of grids, and to prepare the text necessary
            for any revisions to existing manuals.

         b. Investigation reveals that no military manual covers the subject
            of military grids completely.  This is a serious omission and
            should be remedied.

         8. It is felt that the above recommendation will materially improve the
            use of our grid system.  The needs of the various grid users are provided
            for; a standardization is effected; and full use is made of knowledge gained
            through experience during the past war.



                                                    W.H. MILLS
                                                    Colonel, Corps of Engineers
                                                    Commanding Officer.

Chairman of JMPC Ad-hoc Committee on Universal Military Grid Referencing System

A link to a pdf of the original document

       27 February 1948

FROM: Chairman of JMPC Ad-hoc Committee on Universal Military Grid
      Referencing System

To: Colonel Northrup, U.S. Army
    Captain Hobbs, U.S. Navy
    Colonel Tison, U.S. Air Force

Reference: (a) JIC 410/M of 23 January 1948

Enclosure: (A)


     1. In accordance with directive contained in par. 1 of reference (a),
a committee consisting of Colonel Mills, Commanding Officer, Army Map
Service (Chairman), Mr. Bloom of the Aeronautical Chart Service and Mr. Medina
of the U. S. Hydrographic Office has devised a universal military grid
referencing system which it considers suitable for the Armed Forces.

     2. This is submitted for consideration.  Enclosure (A) contains complete
information on the system.  It is requested that this proposal be
transmitted to the operating forces for study.

     3. Should the Army, Navy or Air Force believe that there is another
system more suitable than that described in Enclosure (A), complete information
regarding such a system should be submitted to this ad-hoc committee
for investigation and consideration.

     4. In view of the urgency for an early decision, action should be
expedited. Approval, or any recommended changes, are desired by 15 April
1948.



                                           W. H. MILLS
                                           Colonel, Corps of Engineers
                                           Chairman, JMPC Ad-hoc Committee
                                           on Universal Military Grid
                                           Referencing System

Copies:

Navy: 25
Army: 50
AF: 25
Lt. J. R. Phillips, JMPC: 1

                    Joint Mapping Photography Committee
                            Ad-hoc Committee on
                 Universal Military Grid Referencing System


        A PROPOSED STANDARD UNIVERSAL MILITARY GRID REFERENCING SYSTEM

1. INTRODUCTION
   There is a mandatory and urgent need for a standard referencing system,
   for use by the Armed Forces.Such a system must have the following
   outstanding characteristics:

   a. It must meet the individual and collective requirements of the Army,
      Navy and Air Forces in such a way that all services speak the same
      language.

   b. It must insure positive identification of any point in the world,
      without danger of ambiguity,particularly at spheroid and datum junctions.

   c. It must be simple of understand, brief and capable of being abbreviated
      either for large-scale or for local operations; and adapted
      to the requirements of fighter pilots traveling at high speed.

   d. It must be suitable for rapid computations of range and azimuth,
      with the required accuracy.

   e. It must be usable in Polar areas.

   f. It must avoid difficulties arising from the reversal of sign at
      the 0 and 180th meridians and Eqautor.


2. PRESENT SITUATION

   a. Section IV, W.D.Circular No.33, 5 February 1937, establishes the
      Universal Transverse Mercator Grid as the official grid for the
      Department of the Army. Accordingly, all maps published by the
      Army since 5 February 1947 bear the UTM Grid.  In the interests of
      consistency, the U. S. Hydrographic Office has also surprinted the
      UTM Grid on all approach and bombardment charts, and on all
      hydrographic charts designed for use in amphibious training exercises.
      The UTM Grid supersedes an undersirable heterogeneous system which
      included the World Polyconic Grid, the U. S. Polyconic Grid, the
      Panama Grid and approximately 80 so-called British Grids.  Since the
      projection of the grid is conformal, the grid can normally be
      applied to map constructed on any the standard projections. (I)
----------------
I
The merits of the UTM Grid and deficiencies of the others are
discussed in Appendix I.

   b. The following are the grid referencing systems which are authorized
      and/or presently in use by the various armed forces(2); none can
      satisfactorily fill the requirements for a universal system as outlined
      in paragraph 1, above:

      (1) Fire Control Referencing (FM 6-40, 1 June 1945, Chapter 2, Section
          1 , pp.163-164: also, see Appendix II-I): This is the system
          presently authorized for use by the Army.


          (a) Advantages: The system is suitable for rapid computations
              of range and azimuth;no difficulties arise from the
              reversal of sign at the 0° and 180° meridians and at the
              Equator.

          (b) Deficiencies: Designed originally for fire control use by
              the Army, it does not fill the needs of the Air Force and
              the Navy; it is not adaptable to world wide use; it does
              not provide for spheroid junctions; while it allows for
              abbreviation for large scale activities; it does not
              provide for referencing in the polar areas; it is lengthy
              and awkward due to use of sheet-names, decimals, hyphens
              and parentheses.

      (2) Pacific World War II Practice (see Appendix II-2): Finding the
          authorized referencing system inadequate, the commanders of
          the Pacific Theaters devised a new system to fit their needs.
          This was a compromise between the Fire control and British
          Grid referencing systems. To a certain degree, this alleviated
          but did not eliminate the deficiencies of the authorized
          system.


          (a) Advantages: The system is suitable for rapid computations
              of range and azimuth;no difficulties arise from the
              reversal of sign at the 0° and 180° meridians and at the
              Equator; a reference is brief and not complicated with
              decimals,hyphens and parentheses.

          (b) Deficiencies: While fairly satisfactory for the local needs
              of the Army an Navy, it did not fill the needs of the Air
              Force: it is not adaptable to world wide use; it does not
              provide for spheroid junctions; it does not provide for
              abbreviation; it does not provide for referencing in polar
              areas; while the awkwardness of the authorized system was
              eliminated, ambiguities in reporting outside grid zones
              could result.

      (3) British Grid Referencing (TM 44-225, section IX, pp. 70-72; also,
          see Appendix II-3): By agreement, U. S. Forces used the British
          Grids wherever they existed.

______________
2
Regulations also provide for the Point Designation Grid (TM 44-225, 30
June 1944, Section V, pp. 63-64), the Jan grid (TM 44-222, Section VI,
pp. 65-66), and the Thrust Line Method (TM 44-222, Section VII, p. 67).
These are special purpose systems and their continued use will not be
affected by the adaption of a standard referencing system.



          (a) Advantages: The system is suitable for rapid computations of
              range and azimuth; no difficulties arise from the reversal of
              sign at the 0° and 180° meridians and at the Equator; a
              reference is brief and not complicated with decimals, hyphens and
              parentheses.

          (b) Disadvantages: It does not fill the needs of the Air Force
              and Navy; it is not easily adaptable to world wide use due
              to lack of order and homogenity; it does not provide for
              spheroid junctions; while it permits slight abbreviation in
              local areas, no provision is made for similar abbreviation
              for large scale activities; it does not allow for referencing
              in polar areas; the number of grid and datum junctions was
              excessive resulting in frequent difficulty in rapid computation
              of range and distance when origin and destination were in
              different grid areas.

      (4) Air Defense Grid (TM 44-225, Section X, pp.72-79)

          (a) Advantages: The Air Defense Grid referencing partially
              satisfies large scale use of the Air Force.

          (b) Disadvantages: The system does not satisfy the requirements of
              the Army and Navy; the system does not lend itself to the use
              of the pilot in fighter support; while it is fairly satisfactory
              in reporting areas in world wide activities it is not
              satisfactory for reporting spot positions; while it provides
              for abbreviation, in certain areas if the first letter of
              the reference is dropped, the nearest similar reference is
              near enough to create ambiguity; it is not suitable for rapid
              computation of range and azimuth; it does not provide for
              polar areas; it is ambiguous at datum junctions.


3. PROPOSED SYSTEM
   a. General

      The proposed system is based on the Universal Transeverse Mercator Grid
      between 80° N and 80° s; in the polar areas it is based on polar
      Stereographic grids.

      (1) In conjunction with these grids, the reference system meets all
          the requirements for a referencing system as outlined in
          paragraph 1, above.

          (a) It provides for the individual and collective needs of the
              Army, Navy and Air Force. The basic principles followed by
              any of the armed forces in giving a reference are alike; a
              reference cited by any one branch will be readily recognized
              by all and will permit easy conversion when required.


          (b) It provides a positive and unambiguous identification for any
              point on the globe especially at spheroid and datum junctions.

          (c) Since the referencing in all cases follows the basic and simple
               system of reading "right-up", a reference is easily understood.
               A reference is brief and is capable of being abbreviated
               either for local or large scale activities, thus satisfying
               any and all needs.

          (d) It is well suited for rapid computation of range and azimuth. (3)

          (e) It is suitable for polar areas. Azimuth will agree with that
              now in use by polar aviators; this avoids the difficulties
              of meridian convergence which are a necessary feature of
              computations in latitude and longitude.

          (f) There is no reversal of sign at any meridian or any other
              junction of zones.

      (2) The system has the following additional merits:

          (a) The coordinates always increase to the east and to the north,
              except in the polar regions where the direction of positive
              coordinates is umambiguously indicated.

          (b) The system conforms to the spherical surface of the earth.
              The grid line can be considered as a system of coordinates
              whose position on the earth's surface is as exactly fixed
              as meridians and parallels.

          (c) As the grid is conformal, it can easily be applied to a map
              constructed on any of the conventional projections, unless
              they are so extended as to have unusually large distortions.

   b. First division - 8° NS X 6° EW rectangles

      (1) Between 80° N and 80° S the world is divided into rectangles 8°
          NS X 6° EW (See Exhibit E-H). The columns (6° wide )are identified
          by the Universal Transverse Mercator(UTM) Zone numbers --
          that is,starting at the 180° meridian the columns are numbered
          from 1 to 60 consecutively proceeding easterly. The rows (8°  NS)
          are identified by letters; starting from 80° south and proceeding
          northerly the rows are lettered consecutively from C to X (I and
          O omitted). The designation (called the grid zone designation)
          of such an 8° NS X 6° EW rectangle is determined by reading
          (right-up) first the column designation (as 54) and second row
          designation (as U); as: 54U

      (2) The north polar area above the 80° parallel is divided into two
          parts by the 0° and 180° meridians; the half beginning west of
          0° is identified as Y; the half beginning east of 0° is
          identified as Z
__________
3
See Appendix I.


          (See Exhibit I). Similarly,the south polar area below 80° is
          divided into two halves by the 0° and 180° meridians; these halves
          are identified by A and B respectively.

   c. Second division - 100,000 meter squares

      (1) Between 80° N and 80° S each 8° X 6° rectangle is divided into
          100,000 meter squares based on the UTM grid for zone. Every
          column of squares is identified by a letter; likewise, every row
          of squares is identified by a letter. (See Exhibits A and B).  On
          the equator, starting at the 180° meridian, and proceeding easterly
          for 18°,the 10,000 meter columns,including partial columns
          (caused by convergence), are lettered A to Z (I and O omitted)
          consecutively. The 100,000 meter rows are labelled from A to V
          (I and O omitted) reading from south to north, with the partial
          alphabet being repeated every 2,000,000 meters. Every odd numbered
          6°  wide UTM zone will have the alphabet of the 100,000 meter row
          letters beginning at the equator; the even numbered 6° wide UTM
          Zones will have the alphabet of the 100,000 meter row letters
          beginning at the 500,000 meter northing grid line north of the
          equator. This staggering will considerably lengthen the distance
          between duplicating letters and will permit necessary manipulation
          along spheroid junctions. Below the equator the 100,000
          meter row letters will continue consecutively following the plan
          of the letters above in the same zone. The designation of a
          100,000 meter grid square is determined by reading (right-up)
          first its column designation (as X) and second its row
          designation (as Q): as: XQ.

      (2) Under this system a 100,000 meter square designation will not be
          repeated in an aera 18° NS 18° EW. This will normally eliminate
          the necessity of preceding grid references within such an area by
          the grid zone designation (54U in b (1) above) even though report
          is being made from as many as two grid zones away.

      (3) In the polar areas the 100,000 meter columns at the right angles to the
          90° - 90° meridians are lettered from J to Z in zone designation Y
          and A to R in zone designation Z (I and O omitted; also omitted are
          D, E, M, N, V and W to avoid confusion with 100,000 meter
          squares in adjoining UTM zones.Starting  at the 80° line the
          100,000 meter rows at right angles to the 0° - 180° meridians
          are labelled A to Z cosecutively (I and O omitted). The identification
          of a 100,000 meter square consists of two letters,
          reading right-up (See Exhibit I).

   d. Grid references for U.S. Army

      (1) A U.S. Army reference shall consist of a number and a letter (the
          grid zone designation) followed by two letters (identifying the
          100,000 meter square in which the point of reference lies),
          followed by a group of numbers expressing to the required
          accuracy the E and N coordinates of the referred point within
          the 100,000 meter square; examples:


          (a) 54UXQ         (locating a point within 100,000 meters)

          (b) 54UXQ55       (locating a point within 10,000 meters)

          (c) 54UXQ5354     (locating a point within 1,000 meters)

          (d) 54UXQ539544   (locating a point within 100 meters)

      (2) Normally, a general reference is seldom located to an accuracy of
          more than the closest 100 meters. To provide for the needs of
          surveying and to anticipate any contingency, the following
          references are provided:

          (a) 54UXQ53925443      (locating a point within 10 meters)

          (b) 54UXQ5392354432    (locating a point within 1 meter)

          (c) 54UXQ539234544321  (locating a point within 0.1 meters)

      (3) Normally,all elements of a grid reference shall not be used. Those
          to be omitted will depend upon the size of the area of activities.
          Thus:

          (a) If activities are confined to an area not exceeding 18°  EW X
              18° NS,the grid zone designation (54U) usually will be
              omitted. In such an area, a reference to the closest hundred
              meters usually will read: XQ539544

          (b) If activities are confined to an area not exceeding 100,000
              meters NS X 10,000 meters EW,in addition to omitting the
              grid zone designation, the 100,000 meter square identification
              (XQ) will be omitted; the point will be referenced only by
              numbers. Thus, in such an area, a reference to the closest
              hundred meters will read: 539544

          (c) Numerical reference: The numerical part of a reference taken
              from a 1,000 meter grid, will be a six digit number; for
              example: 539544. The "53" represents the 10,000 and 1,000
              digits of the easting grid line to the west of the referenced
              point, the "9" represents the estimated tenths from the easting
              grid line to the point, the "54" represents the 10,000
              and 1,000 digits of the northing grid line south of the
              referenced point, and the "4" represents the estimated
              tenths from the norhthing grid line to the point. (See
              Exhibit C). To maintain a relationship between similar grid
              references from different scale maps, the numerical part of
              a reference taken from a 10,000 meter grid will be a four
              digit number; for example: 5354. The "5" represents the
              10,000 digit of the easting grid line to the west of the
              referenced point, the "3" represents the estimated tenths
              from the easting grid line to the point, the second "5"
              represents the 10,000 digit of the northing grid line south
              of the referenced point, and the "4" represents the estimated
              tenths from the northing grid line to the point. (See Exhibit
              D).

e. Tad Grid Reference

   (1) A TAD reference shall consist of two letters (identifying the
       100,000 meter square containing the point of reference (see
       paragraph 3 c, above), followed by four numerals (the 10,000
       and 1,000 digits of the northing grid line south of the point
       and the 10,000  and 1,000 digits of the easting grid line west
       of the point, (the identification of the 1,000 meter grid
       square containing the point), followed by a letter (identifying
       the 200 meter grid square containing the point). (See Exhibit
       F). Example: XQ5354J

   (2) The first two letters (XQ) normally shall be  omitted; they will not
       be used unless the reference is being reported more than 100,000
       meters away. Thus, the usual TAD reference will be written simply as: 5354J

   (3) The p1an for lettering the 200 meter squares follows:


                        A B C D E
                        F G H I J
                        K L M N O
                        P Q R S T
                        U V W X Y

f. Air Defense references

   (1) A reference shall consist of the grid zone designation (as 54U),
       followed, if necessary, by two letters (identifying the 100,000
       meter square containing the point), followed, if necessary, by a
       group of numbers expressing to the required accuracy the E and N
       coordinates of the referred point within the 100,000 meter square;
       (See Exhibit G).   Examples:

       (a) 54U        (locating a point within a 8° NS X 6° EW)
       (b) 54UXQ      (locating a point within 100,000 meters)
       (c) 54UXQ55    (locating a point within 10,000 meters)
       (c) 54UXQ5354  (locating a point within 1,000 meters)

   (2) If reports are being confined to an area not exceeding 18° NS X
       18° EW, the grid zone designation may be omitted and the
       reference read as:

       (a) XQ     (locating a point within the 100,000 meters)
       (b) XQ55   (locating a point within 10,000 meters)
       (c) XQ5354 (locating a point within 1,000 meters)

4. CONCLUSION

   It is felt that the system outlined above is a homogeneous solution of
   the problem and will satisfactory fill the needs of the Navy, the Air
   Force and Army. The needs of the various grid users are provided
   for and a standardization is effected.


                                      APPENDIX 1


                       STUDY AND DISCUSSION OF MILITARY GRIDS
                                         by
                                 Army Map Service
                                        and
      Military Intelligence Division, Office, Chief of Engineers, U.S. Army

1. Purpose of Study
   The purpose of the following study is to determine the characteristics
   required in a military grid and to select a system most nearly answering these
   requirements. Marked disadvantages are inherent in most grid systems now used.
   These disadvantages are complicated by the existence of many incompatible systems.


2. Existing Conditions
   2.01. The polyconic military grid is prescribed by Section VII, AR 300-15,
         for use on all military maps of the United States. This system is laid out in
         zones 9 degrees wide in longitude with 1 degree of overlap between zones.  It
         is so inaccurate at long ranges in certain directions that it cannot be used
         satisfactorily for the control of the fire of coast artillery weapons or heavy
         field artillery.

   2.02.  Because of this inaccuracy, the coast artillery harbor defense grid
          for area in the neighbourhood of the harbor defences in the continental United
          States is also authorized in paragraph 28, Change No.4, AR 300-15.  This harbor
          defense grid system is a Lambert conic conformal designed particularly to
          serve the guns of the harbor defense concerned.  It is not only not connected
          to the military grid system in the same area but is incompatible therewith.

   2.03. Many other grid systems are in use not only in the Unites States
         but also in the rest of the world.  Twenty states of the Unites States have
         adopted the state plane coordinate systems measured in feet and especially
         designed to serve the particular state in question.  Each such system is hardly
         extensible beyond the borders of the state without the introduction of mater-
         ial inaccuracies.  The enclosed map shows the overall picture of the grid
         systems used in allied military operations during the recent campaigns.  (Incl. 1)

   2.04.  It is obvious that the presence of so many systems complicates map
          preparation and impose material confusing handicaps on actual combat operations.
          The presence of more than one grid system covering one area presents no particular
          problem in peace time or on manoeuvres involving but one arm.  When the
          fog of war confuses men's minds, the presence of several coordinate systems in
          one area for use of different arms in fraught with potent opportunity for
          disasters resulting from uncoordinations attributed to mistakes in using the
          military grid.  Involved in this matter are branch pride and branch stubbornness,
          each branch feeling justified in having a special grid system designed to the
          particular capabilities and needs of that branch.  An example of this occurred
          in the United Kingdom where the British coast artillery, Navy and Air Force covered
          the coastal area with three incompatible, incommensurable grid systems.
          Intolerable confusion which resulted from the use of these grids during the
          numerous German air raids in the Battle of Britain makes it highly probable that
          these conflicting systems would have led to at least a few local disasters had
          there been an invasion of Great Britain.

3. Basic Requirements

   3.01. Primary Purpose of a Grid: The primary purpose served by the
         military grid on a map is to provide quick solutions to problems of
         distance and azimuth for the firing of weapons. It  provides a
         quick simple means for referring to spot locations and for designating
         targets.  It is an essential tool in coordination of military
         operations.

   3.02. The coordination of the efforts of the many arms used on land, sea,
         and air, is a problem so complex as to make mandatory a single
         simple solution for problems of target designation and determination
         of range and azimuth.  This requirement is believed to be so
         important in war that the use of a single system of limited but
         adequate accuracy is held to be better than the simultaneous use
         of two incompatible but otherwise more accurate systems.

   3.03. The characteristics of the using arms and weapons which affect the
         design of the system to be adopted involve relatively little
         research. As a general rule, it has been assumed that permanently
         emplaced batteries will be more accurate in their fire than
         batteries temporarily emplaced in the fields. Therefore, the
         following table appears to provide sufficient criteria to determine
         the desirable accuracies of the grid system adopted.


                      Probable Errors of Different Caliber
                   Permanently Emplaced Guns at Ranges Shown
 ------------------------------------------------------------------------
                                                  Minimum Probable
                Probable Errors in Yards          Relative Errors
           --------------------------------------------------------------
Range in      6" Gun      8" Gun    16" Gun
  Yards    Range Defl. Range Defl. Range Defl.    Range     Deflection
 10,000     22    2     68    3       18  3       1:555     1:5,000
 15,000     35    4     70    5       28  4       1:535     1:3,750
 20,000     52    6     73    8       40  6       1:500     1:3,333
 25,000     68    8     77   13       52  7       1:480     1:3,571
 30,000                 83   19       63  9       1:476     1:3,333
 35,000                              73  10       1:479     1:3,500
 40,000                              80  10       1:500     1:4,000
 45,000                             *77  *7      *1:584    *1:6,428
 ------------------------------------------------------------------------
* These values are appear unusual

4. Desirable Characteristics

   4.01. Primarily a grid system should be accurate enough for all weapons
         and all military uses other than for very long distance missiles,
         should be quickly applicable to any previously ungridded native
         map, should yield readily to simple computing methods and should
         provide simple numerical designators for location of targets.

   4.02. Plane System. The system of coordinates desired is one with which
         all computations for the most accurate artillery firing can be
         simply yet accurately performed and especially one in which the
         integrity of angles is preserved. A mathematically exact graticule,
         such as that presented by the meridians and parallels, requires
         the use of geodetic functions to solve the spherical triangles
         involved, and entails a long, time-consuming complicated computation.
         Moreover, due to the convergence of the meridians, the arc of the
         latitude increases. Complicated geodetic formulas would be
         necessary in the computation of any distance except one along a
         meridian. Complex fire control instruments would be needed,
         manned by personnel highly trained in a branch of advanced
         mathematics. Neither the personnel, the instruments, nor the time are
         normally available. As a consequence, the system adopted should
         be one in which plane trigonometry can be employed in the solution
         of triangles.  In such a plane system for general application to
         large areas, the simplest and quickest computations can be secured
         through use of a grid network of equally spaced parallel and
         mutually perpendicular lines.

   4.03. Grid Accuracy. A high degree of accuracy is, of course, desirable.
         However, grid accuracies which are greatly in excess of the
         accuracies for the most precise weapon using the grid appear to
         be neither necessary nor practicable. By reference to paragraph
         3.03 above, it will be noted that the probable errors of artillery
         weapons are much greater in range than are their probable errors
         in deflection. The minimum probable error of permanently emplaced
         guns rarely is less than 1/555 in range and 1/5,000 in deflection.
         Consequently, a suitable military grid should be one designed to
         conform to these minimum probable errors.

   4.04. Adaptability to Various Projections. The grid system selected
         should be adaptable for use on native maps without complicated
         recomputation or redrafting of that map. There are many map
         projections used in the  making of large scale maps throughout the
         world.  It is desirable to be able to overprint the adopted grid
         system on any or all of these projections without the introduction
         of errors in range and azimuth beyond that probable in good
         artillery practice.

   4.05. Unit of Measure. Three general systems of linear measure are
         commonly encountered on maps and in grids: the metric system, the
         so-called English system, the nautical system. Mixtures of these
         systems unfortunately are prevalent. This matter is further
         complicated by the fact that three differing elements are involved -
         map quantities, grid quantities, and the quantities employed by
         the using arms and weapons.


         a. Map quantities include azimuths, horizontal. distances, contour
            intervals, and underwater depths. To the map user, the unit
            of measure in which horizontal distances on a map are expressed
            is not particularly important, as the  conversion from map
            distances to ground distances is frequently done graphically
            against either an appropriately graduated bar scale on the
            map or range scale. However, the unit of linear measure used
            in the basic survey of the map may  complicate the computation
            and compilation trig lists for fire control. This latter
            operation is already quite complex due to the differing origins
            of  longitude, datum planes, spheroids and schemes of projection,
            and other variations encountered in the native surveys of the
            world. Thus, the conversion from one unit of linear measure
            to another incommensurable. unit adds an operation subject to
            mistakes and affecting final accuracies.

         b. Contour intervals and spot elevations should, but may not, be
            in the same unit of measure as the horizontal distances, in
            order to provide for the ready calculation of true slant
            ranges, defilade, mask, profile, etc.

         c. Underwater depths are generally expressed either in meters or
            in fathoms, although shoal water depths may be also expressed
            in feet. It is highly desirable that these units of measure
            be the same as the horizontal unit in order to be readily
            useful in the computation fo underwater profiles and beach
            gradients.

         d. The military grid, while essentially concerned with angles and
            horizontal distances, must be precisely related to the
            computed geographic positions.  The necessary correlation between
            the vertical unit of measure and the horizontal unit if measure
            on the grid as indicated in b above, is essential for the
            quick solution of problems involving defliade.,mask and true
            gun target distance.

         e. The using arms and weapons are not entirely coordinated in the
            units of measure employed in the laying of the piece. The
            Coast Artillery measures azimuths in degrees and hundredths
            of a degree from grid south as the origin. The Field Artillery
            measures angles in mils, with field orientation of base circles.
            The Coast Artillery measures its range in yards ,while the Field
            Artillery may measure it in yards or meters. Due to the
            tangent relationship of the mil, Field Artillery can readily
            transpose from angular measure to linear distance in either
            meters or yards. Each artillery weapon is served and laid by
            employing a multiplicity of tabular information, plotting
            tools, and gunner's instruments. All these things must be
            related to the unit of measure selected for map and grid
            quantities. At the present moment, due to the recent
            tremendous concentration of field artillery weapons in the
            European campaigns and to the use of the metric system
            throughout in that area, our Field Artillery is well equipped and
            trained in the use of the metric system. The Seacoast
            Artillery of the United States, including the Panama Canal and
            Oahu, are not so equipped or trained. They use the yard-
            hundredths of a degree system, except in the coast defenses
            of Sen Diego where coast Defense grid graduated in
            feet rather than in yards.

         f. Existing Conditions. The majority of the large scale maps of
            the world are made on the metric system.Exceptions to this
            rule are the United States, Canada, Australia, United Kingdom,
            Union of South Africa, India, Melanesia and Middle East. The
            following table shows teh unit of measure of native maps and
            military grids employed in operational areas of the recent
            campaigns.


                               MAP UNITS                  GRID UNITS
            -------------------------------------------   -------------------
AREA        HORIZONTAL UNIT    VERTICAL    DEPTH UNIT     GRID   GRID TYPE
            (Map Bar Scale)    UNIT (Map   (Bathymetric   UNIT
                               Contour)    Contour)
-----------------------------------------------------------------------------
1. France        Meter         Meter       Meter          Meter  Lambert
2. Germany       Meter         Meter       Meter          Meter  Trans. Merc.
3. Italy         Meter         Meter       Meter          -----  -----
4. Tunisia       Meter         Meter       Meter          -----  -----
5. Libya         Meter         Meter       Meter          Meter  Trans. Merc.
6. Egypt         Meter         Meter       Fathon         Meter  Trans. Merc.
7. Okinawa       Meter,Cho     Meter       Meter          Meter  Unknown
8. Burma         Miles         Foot        Fathom         Yard   Lambert
9. China         Li,meter      Meter       Fathom         Yard   Lambert
10. Russia       Meter         Meter       Foot           Meter  Trans. Merc.
11. Hawaii       Mile          Foot        Fathom         Yard   Polyconic
12. Philipines   Mile          Foot        Fathom         Yard   Polyconic
13. Poland       Meter         Meter       Meter          Meter  Stereographic
14. Holland      Meter         Meter       Meter          Meter  Stereographic
15. Belgium      Meter         Meter       Meter          Meter  Bonne
16. Japan        Meter,Cho     Meter       Meter          Meter  Trans. Merc.


   4.06. Width of Zone. An inherent fault of any system of plane coordinates
         applied to the surface of a spheroid is the fact that inaccuracies
         increase as the zone is extended east and west or in other
         cases, as the belt is extended north and south. It is also
         desirable, although not entirely necessary, to keep at a minimum
         the deviation of grid north from true north. This deviation
         likewise increases materially as the distance from the central
         meridian increases. These inaccuracies can be kept within reasonable
         limits by the adoption of narrow grid zones. It has been
         stated only semi-facetiously that there are three military
         engineering axioms:

         a. It always rains in war.
         b. It is always too cloudy to get aerial mapping photographs.
         c. Battles are always fought on grid junctions.

         This last axiom is spoken from the depths of bitter experience,
         rendering it obvious that a grid zone should include as much
         area as possible so as to obviate too frequent junctures between
         grid zones on the battlefield.However,this desirable criterion
         cannot be widely applied without including intolerable
         inaccuracies of the grid. Elimination of the undesirable features
         consequent upon fighting a battle on a grid juncture can be
         partially accomplished by providing for overlap between grid zones,
         the 9 degree width of the military grid system presently prescribed
         for use in the United States introduces appreciable inaccuracies
         near the edges. A reduction in width to 6 degrees betters this
         situation materially. The computation of ranges and azimuths
         where the gun position is located in one grid zone and target
         in another can be provided for by a half  degree overlap between
         grid zones, enabling the coordinates of the gun zone to extended
         a half a degree into the grid zone in which the target is located.

5. Comparison of Grids

   5.01. Polyconic Grid. For military Purposes, a grid may be regarded as
         a set of perfect squares ruled on a plane map, scale 1:1, and then
         transferred to the earth's surface. Evidently after being transferred
         to the earth's surface the squares will no longer be perfect;
         and distortions they will have received in being put on
         the surface of the earth will reflect the distortions of the
         projection used for the map.



         a. The polyconic projections is defined as one which the central
            meridian and all parallels are mapped to scale and with true
            curvature.  All other lines are stretched, the amount increasing
            as the square of the distance from the central meridian, and
            being greatest for north-south lines. Angular errors also
            appear, increasing with distance from the central meridian.
            Of course it is possible to commute these errors, at least
            roughly, and to allow for them, and this is regularly done by
            engineer survey troops. But the corrections are generally
            considered beyond what is to be expected of artillery units
            in the field, and for that reason all mention of them is omitted
            in artillery technical manuals, even when survey procedures are
            discussed.  It is proper, therefore, to compare the errors of the
            projection as shown in the following table, directly with the
            errors of the guns (Section 3.03).


                      Errors of U.S.Military Grid at 4°30'
                      from Central Meridian, and 30° Latitude
-------------------------------------------------------------------------------------------------
         |                            |                             |             True
         |       True  Azimuth 0°     |      True Azimuth 45°       |          Azimuth 90°
True     |----------------------------|-----------------------------|----------------------------
Range    | Range  Rela- | Defl. Rela- | Range  Rela- | Defl.  Rela- | Range  Rela- | Defl.  Rela-
(yds)    | Error  tive  | Error tive  | Error Defl.  | Error  tive  | Error  tive  | Error  tive
         | (yds)  Error | (yds) Error | (yds) Error  | (yds)  Error | (yds)  Error | (yds)  Error
---------|--------------|-------------|--------------|--------------|--------------|-------------
10,000   |  23    1:435 |  0      0   |  12   1:836  |  12    1:836 |  0       0   |  0       0
30,000   |  70    1:428 |  4   1:7500 |  35   1:859  |  35    1:859 |  0       0   |  0       0
40,000   |  70    1:428 |  4   1:5000 |  47   1:850  |  47    1:850 |  0       0   |  0       0

            It is evident that at 40,000 yards, and an azimuth of 45°, the error
            of the grid in deflection is almost five times the probable error
            of a 16-inch gun, Errors of the above amount are characteristic
            of the so-called non conformal projections, i.e. those in which
            the shapes, as well as the scale of small areas is distorted.
            Such non-conformal grids were widely used prior to World War I;
            but most of them have been abandoned in recent years, chiefly,
            no doubt, because of the breakdown of the old French Bonne grid
            during the War.  In addition to its lack of conformality, the
            polyconic projection possesses the disadvantages that it has not
            been studied thoroughly from a mathematical standpoint. Hence
            the small corrections needed for precise surveys are not known;
            the transformation from other grids to the polyconic is not
            known; not even the transformation from one polyconic belt to
            another has been studied.

         b. So far as grid junctions are concerned,the polyconic is theoretically
            excellent; it can be extended indefinitely both north and
            south, so that the world can be divided up into meridional strips.
            In practice, the present World Polyconic has two latitudinal
            junctions, one near 24°, due to failure to put the origin of the
            old U. S. grid sufficiently far south; the other is near 49°,
            and is due to inaccuracy in the old tables which amounts to 1.1
            meters.

         c. The polyconic grid is not well suited for foreign maps due to
            its lack of conformality.  By the older laborious hand methods
            of grid plotting, this introduced no difficulties other than
            laborious calculations; but the use of the coordinatograph, for
            rapid plotting of grids and projections, requires a conformal
            grid.

   5.02. Cassini-Soldner Grid. This is a non-conformal grid,very similar to
         the polyconic, and open to the same objections.  It differs only in
         that the grid east-west lines, rather than the parallels are represented
         to true scale and with true curvature.

   5.03. The Bonne Grid. Both meridians and parallels are represented to true
         scale on the Bonne projection; the error shows up on lines which run
         NE to SW, or NW to SE,  It is just as bad as in the case of the other
         non-conformal projections; and this projection has been generally
         abandoned.

   5.04. The Stereographic Grid. The stereographic grid may be briefly defined
         as a conformal grid (that is, one having zero angular distortions
         for small distances, and very small angular distortions for any
         distance) in which the scale error is zero on a standard circle.
         The fact that this grid is conformal insures that within about 200
         miles of the center-point, the error in deflection will be less than
         the error of the most accurate guns. The range error is considerably
         larger than the deflection error, but is considerably less than the
         range error of permanently emplaced artillery. Unfortunately, this
         grid cannot be extended more than 300 miles from its center point
         and grid zones are circular in shape. It is quite suitable for small
         roughly circular countries such as Poland, Holland, or Romania, which
         use it; but on a world-wide basis it would lead to great multiplicity
         of grid junctions, and points where three or more grids meet could
         not be avoided.

   5.05. Lambert Grid. Like the Stereographic, the Lambert is a conformal
         projection. It may not defined as a conformal projection in
         which the scale errors are zero along two parallels. It is well
         suited to the mapping of moderately large areas, and has been
         extensively used by the British and French, especially in
         recent years. The numerical values of the errors are similar
         to those for the Transverse Mercator given in the next paragraph.
         The errors are very small in deflection, as is necessary
         for an artillery grid. It is readily adaptable to gridding
         maps on other projections. In general, a slight change
         of scale only would be required. On the other hand, the
         rapidly increasing grid declination makes it impracticable
         to extend the grid more than about 15° from the central
         meridian (except along the Equator, where grid declination
         is always zero). For this reason, the British and French were
         forced to introduce numerous zones, and to permit junctions
         of three grids at the same point. These numerous grid zones
         necessary, both north-south and east-west, make the Lambert
         undesirable for extensive world coverage.

   5.06. Transverse Mercator Grid. The Transverse Mercator grid may
         be defined as a conformal grid in which the central meridian
         is represented by straight line at the true scale. It is well
         suited to large areas, and is being used by the Germans,
         Russians, British, and Japanese. The errors at a given point
         vary little in all azimuths end average values for different
         ranges are given in the following table for 30° north latitude
         and for 3 1/2° from the central meridian.

         -------------------------------------------------------------------
         |        | Transverse                  |   Transverse             |
         |  True  | Mercator Grid               |   Mercator Grid          |
         |  Range |     Range        Relative   |   Deflection    Relative |
         |  (yds) |     Error          Error    |   Error (yds)     Error  |
         |        |     (yds)                   |                          |
         |--------|-----------------------------|--------------------------|
         | 10,000 |      10           1:1000    |       0             0    |
         | 30,000 |      30           1:1000    |       3          1:10000 |
         | 40,000 |      40           1:1000    |       5.8        1:6900  |
         |--------|-----------------------------|--------------------------|

         Obviously,t he grid is well suited to artillery purposes since the
         inevitable errors are thrown into range rather than deflection.
         The grid can be extended indefinitely in latitude like the polyconic.
         Hence it is never necessary to have a grid junction
         involving more than two grids (except, of course, near the
         Poles). Transformation of coordinates from one belt to another
         can be done by a formula already worked out. The formula is
         always the same, and is very simple in character.  The grid
         declination will remain moderate throughout the belt.  The grid
         can be readily adapted to use on other projections. Much
         theoretical work has already been done on this subject by a
         large group of mathematicians, including especially Professor
         W. K. Hristow. Extensive computations, especially for the
         Balkan Countries, were done by the German High Command (O.
         K. H.)during World War II which could, in an emergency, be
         promptly utilized if the proposed projection is adopted.
         In addition, much geodetic data of foreign areas on file at
         Army Map Service are on this system. The projection is well
         suited for converting data on various spheroids to a common
         basis. Traverse and lower orders of triangulation may be
         computed and adjusted directly on the grid due to its
         conformality. This feature, which is a large saving in field
         and office, is not practicable where a non-conformal
         projection such as the present Polyconic is used.



   5.07. Tabular Comparison of Grids.

   --------------------------------------------------------------------------------------
   GRID SYSTEM           APPLICABILITY          GRID           MAXIMUM         MAXIMUM
                         TO FOREIGN MAPS**   JUNCTIONS         RELATIVE        RELATIVE
                                                             RANGE ERRORS*   DEFLECTION
                                                              (45,000 yds)      ERRORS*
                                                                             (45,000 yds)
   --------------------------------------------------------------------------------------
   Polyconic             Very Poor          Few & simple       1/1228          1/2454
   Cassini-Soldner       Very Poor          Few & simple       1/1228          1/2454
   Bonne                 Very poor          Many & complex     1/2454          1/615
   Stereographic         Good               Many & complex     1/2454          1/15,360
   Lambert               Good               Many & complex     1/2416          1/7,663
   Transverse Mercator   Excellent;
                         much work
                         already done       Few & simple       1/2416          1/7,663
   ---------------------------------------------------------------------------------------
   * Range and deflection errors are maximum values within 160 miles from the
     center of the projection, whether the center is a line, as in the Transverse
     Mercator, Lambert, Cassini-Soldner, and Polyconic, or a point as with the
     other two. The figures are based on GSGS "Survey Computation". 160 miles
     is the approximate distance (at 40° of latitude) from the center of the
     proposed Transverse Mercator Zones to the junctions, about 3° of longitude.
     The actual maximum errors of the present world Polyconic grid are
     considerably larger, since the grid zones are 9° in width.

6. Conclusions as to System to be Adopted

   6.01. The Lambert Orthomorphic projection is conformal but is not
         suitable as it requires grid zone junctures both north and south
         and east and west. The ployconic grid  system now prescribed for
         use as military grid on all maps of U. S. is inaccurate in both
         azimuth and distance.  The greater inaccuracy is in azimuth and
         is more than the probable error in deflection of permanently
         emplaced guns.  The transverse mercator grid is conformal and is
-----------------
**A grid system is considered applicable to a foreign map if it can be put on most maps without changing map or grid except is scale.

         immediately applicable without plottable error, to the majority
         of the map projections commonly encountered on the native
         maps of the world. The transverse mercator grid reduces
         inaccurecies to a point where they are compatible with the
         accuracies required by all modern artillery weapons. This
         grid is sufficiently accurate to eliminate the necessity for
         a special Coast Artillery grid in the vicinity of coast
         defense locations.


                                  APPENDIX 2

1. FIRE CONTROL REFERENCING
   Existing regulations of the Department of the Army (See FM 6-40, Part
   Four, Chapter 2) designate the following method for reading a gird
   reference.

   a. Designatiuon of sheet, parenthesis, X coordinate, decimal, location
      to nearest yard, a dash, Y coordinate, decimal, location to nearest
      yard, parenthesis.

      Example: Annapolis (804.729-1286.684)

   b. When the map is definitely understood, its designation may be omitted.

      Example: (804.729-1286.684)

   c. If the location to the nearest 10 or nearest 100 yards only
      is desired, or if the measurements can not be made with greater accuracy,
      the digits indicating units or tens may be omitted.

      Examples: (1) (804.72.1286.68) to nearest 10 yards.
                (2) (804.7.1286.7) to nearest 100 yards.

   d. For expediency it is permissible to include only two digits to
      the left of decimal point (10,000 and 1,000 digits), omitting
      any preceding digits.

      Examples: (1) (04.729-86.684)
                (2) (04.72-86.68)
                (3) (04.7-86.6)

   e. If the point is fixed within an area 10,000 yards square, only one
      digit need be given before the decimal point of each coordinate.

      Examples: (1) (4.729-6.684)
                (2)(4.72-6.68)
                (3)(4.7-6.7)

   f. If a large number of points are being designated by the
      abbreviated coordinates shown in example (3), the decimals
      and dashes may be omitted and the reference given as (4767).

2. PACIFIC WORLD WAR II PRACTICE

   During the war, the Pacific and Southwest Pacific Commands found it
   feasible to use a system for reading general grid references
   similar to that used with British Girds. Apparently, a broad
   interpretation of existing regulations was made to find authority for the
   change. The name of the map was not mentioned (authority: see 1 b
   above) digits to the left of the 10,000 and 1,000 unit digits were
   omitted (authority see 1 d above) decimals and dashes were omitted
   (authority see 1 f above).

   a. A reference on a map employing a 1,000 unit  interval was determined
      by reading the 10,000 and 1,000 digits of the easting line to
      left of point, estimated tenths(100 units) eastward to point,
      the 10,000 and 1,000 digits of the northing line south of the point,
      estimated tenths (100 units), northward to the point.  This was
      written as a 6 digit continuous number.

      Example: 143286

   b. A similar procedure was follows in reading a reference on a
      map using a 10,000 unit interval, except that the digits for the
      gird lines were for the 100,000 and 10,000 units (the last four
      digits being omitted) and the estimated tenths represented 1,000
      units. Thus, a reference for the same point cited in (1) above,
      was as: 214128

   c. There were two major faults with the pacific system: danger of
      confusion between a reference taken from a map bearing a 1,000
      unit interval and from one of a 10,000 unit interval, inasmuch
      as both were six digit numbers; and lack of connection between
      references for a common point taken from a 1,000 unit gird and
      from a 10,000 unit gird, inasmuch as in reading a reference from
      a 1,000 unit gird the principal digits were the 10,000 and 1,000
      ones and for a reference from a 10,000 unit grid the principle
      digits were the 100,000 and 10,000 ones. Thus, references for
      a common point might read: 047866 (from a 1,000 unit gird) and
      804286 (from a 10,000 unit gird).


3. BRITISH GRID REFERENCING

   References for british grids are read according to the following methods:

   a. Maps bearing a 10,000 unit interval (1:100,000 to 1:500,000):
      letter of 500,000 unit source (written as a small capital
      letter), letter of 100,000 unit square (written as a large
      capital letter), 10,000 digit of easting line to the left of
      the point, estimated tenths (1,000 units) eastward to point,
      10,000 digit of northing line south of the point, estimated
      tenths (1,000 units) northward to the point.

      Example: cA1428

      This locates point to nearest 1,000 units.

   b. Maps bearing a 1,000 unit interval (1:5,000 to 1:100,000 inclusive):
      Letter of 100,000 unit square, 10,000 and 1,000 digits of easting
      line to left of point, estimated tenths (100 units) eastward to
      point, 10,000 and 1,000 digits for northing line south of the
      point, estimated tenths (100 units) northward to the point.

      Example:A143286

      The locates point to nearest 100 units.