You are on page 1of 1

HOW THINGS WORK

Sponsored by
Geomatics Industry Association of America

DIN 18723 Specification


for Theodolite Accuracy
S urveyors have seen the expression
“per DIN 18723” associated with
statements of accuracy of theodo-
lites since the introduction of electronic
theodolites. DIN stands for Deutsches In-
. . . One needs to apply
the rules for random
error propagation, for the
ing DIN 18723 is actually one that meas-
ures precision. As with surveying, by ac-
counting as well as possible for system-
atic errors, it is possible to arrive at an es-
timate for accuracy. Thus if the angles are
stitut für Normung which loosely trans- particular way you use not measured in face I and face II, if col-
lates into the German Institute for Stan- limation has not been checked and ad-
your instrument.
dards (English language website located at justed, if the instrument has not been
http://www2.din.de/index.php?lang=en). At leveled properly, if the so-called “height
the time the standard was first quoted by and a face II direction.” Understanding of standards” adjustment is incorrect, and
manufacturers (mostly from continental this is key to using the DIN specification so on, then the “fundamental accuracy”
Europe) as an accuracy specification, it value. If you take a theodolite or total of the theodolite as given by the DIN
was actually a draft (~1983). However it station with an angle measuring “accura- spec may not be true at all!
has become widely accepted by manufac- cy according to DIN” of 5 seconds, for
turers worldwide for specifying the accu- example, and observe the horizontal cir- Environmental Influences
racy of theodolites. It should be noted cle reading to a precise target in face I In addition to the above-mentioned
here that DIN accuracy cannot be inferred (telescope in the direct position) and sources of error resulting from practices
from the least count of a theodolite, which then invert the telescope and take anoth- and instrumentation, there are the influ-
is the finest measurement or count that an er horizontal circle reading to the same ences of circumstances, mostly environ-
instrument is able to make. In fact, with target in face II (reversed position), the mental (such as atmospheric distur-
the advent of electronic instruments, re- specification tells you that the standard bances) and the practices and adjustment
liance on the least count for anything but deviation (that is, confidence level of (or lack thereof) of the accessories that
an estimate of precision achievable (not ~68%) of the mean (or average) of the will affect the accuracy that is achieved.
accuracy) is highly inadvisable. two directions is ±5 seconds. This how- Examples of these are: optical plum-
ever is not an angle as surveyors are used met(s) adjustment, whether the targets
DIN Spec vs. Accuracy to thinking about. To measure a single are prisms or precise traverse targets,
The standard is equally applicable to angle, we are required to measure two di- whether these targets are mounted on
optical theodolites, and in fact has occa- rections (first a backsight, then a fore- tripods or on range poles, or whether the
sionally been used for that purpose. But sight). Using the equations for propaga- targets are subject to the problem of
whether the theodolite is optical or elec- tion of random error, the standard devia- “phase” error, particularly if a pole or mi-
tronic, surveyors have tended to assume tion of the angle, if measured in face I ni-pole is used as the target.
that a “5-second” theodolite measures an- and face II and then averaged is 7 sec- In summary, DIN accuracy values in-
gles with an accuracy of 5 seconds. This onds! (Take the square root of the sum of dicated for instruments are not the values
is rarely the case (except by coinci- the squares of the errors in each direc- to be assumed that can be obtained
dence). The specification is useful for tion, which in this case simplifies to when measuring angles. Depending on

comparing theodolites however, in that √2 •5.) how the instrument is used (ignoring ac-
all theodolites classified as having a DIN Thus the first lesson to be learned cessories and conditions), the angle ac-
accuracy of 3 seconds, for example, will from this discussion is that one needs to curacy may be higher or lower than the
be roughly equal in terms of angle meas- apply the rules for random error propa- specified value. Once the correct value is
urement performance. For purposes of gation, for the particular way you use computed, however, it must then be
survey design, analyzing data after a sur- your instrument. For example, if you combined with knowledge of practices,
vey, especially when trying to apply a measure the angle twice in face I and instrumentation and conditions, includ-
weight to the accuracy of angle measure- twice in face II, the standard deviation of ing knowledge of the level of adjustment
ment, the DIN specification value must the angle will be 5 seconds. If, however, of all the component parts of the survey-
be interpreted in the light of how the in- you only measure the angles in face I, ing system and accuracies.
strument was used. you can expect the angles to have an un-
If you read the fine print, you are like- certainty of 10 seconds. Note: Information for this article was
ly to find text in manufacturer’s specifi- The second lesson, which may be the compiled by the technical staffs of the
cations that reads something like “stan- more important, is that the method of de- GIAA members that manufacture theodo-
dard deviation of the mean of a face I termining the accuracy of theodolites us- lites and total stations.

DISPLAYED WITH PERMISSION • PROFESSIONAL SURVEYOR MAGAZINE • November 2002 • WWW.PROFSURV.COM • ALL RIGHTS RESERVED

You might also like