3A1 Lecture 3

Topic 1: Linear Measurement
measurement errors
Aims of this topic

-To understand the different types of errors that can occur during surveying and
how they can be eliminated or minimised.
-To explain the differences between precision and accuracy.
-To calculate mean values and standard errors for repeated measurements in a
survey.
-To understand the propagation of errors
Types of Error
The types of measurement fundamental to surveying are distances, heights and
angles. Measuring all of these quantities either manually or electronically
involves some errors. The different types of error that can occur in surveying
and other measurements are:
Gross Errors mistakes or blunders which are usually much larger than the
other category of errors. E.g. reading a levelling staff incorrectly
Systematic Errors are those errors which follow some mathematical formula
and have the same sign and magnitude in a series of measurements that are
repeated under the same conditions. E.g. errors due to temperature, tension,
slope etc.
Random Errors a series of repeated measurements of the same quantity will
be subject to some random error beyond the control of the observer, but
methods can be adopted to ensure they are kept within acceptable limits.
Accuracy and Precision

Precision: represents the repeatability of a measurement and is concerned
only with random errors. Good precision is obtained from a set of observations
that are closely grouped together with small deviations from the mean of the
observations. A set of observations that are spread out widely have poor
precision.
Accuracy: is considered to be an overall estimate of the errors present in
measurements, including systematic effects. A set of observations are
considered accurate if the mean of the observations is close to that of the true
value.
Standard deviation and standard error

The Standard Deviation is a measure of the spread or dispersion of a set of
measurements. If the standard deviation of a set of measurements is small,
the measurements will have good precision.
The standard deviation, x, of a variable is computed according to:
in=1 ( x xi ) 2
x =
n
Where,
Xi is an individual measurement i
is the mean of the set of measurements
n is the number of measurements in the set
For statistical reason calculations of the standard deviation should be based

on large numbers of data, however most surveys involve a small number of
measurements and the standard deviations obtained from them may be
biased. For this reason the standard error, Sx, is a better measure of precision
in surveying. This is obtained by replacing n with (n-1).
in=1 ( x xi ) 2
sx =
(n 1)
(n-1) is know as the number of degrees of freedom or the redundancy and
represents the number of extra measurements taken to determine a quantity.
If a distance in measured 10 times it has (10-1)= 9 degrees of freedom and
there are 9 redundant observations, since only one is required to give the
distance.
Redundant observations allow us to quantify errors in our measurements.
Example:
An angle, a, is measured ten times with the same equipment by the same
observer and the following results are obtained:
47o5638
47o5640
47o5632
47o5640
47o5634
47o5637
47o5635
47o5642
47o5633
47o5639
Calculate the mean, standard deviation and standard error.

Solution:
Mean = 47o56 + (38+40+35+33+40+34+42+39+37) /10

= 47o5637
Propagation of Errors
Surveying measurements such as angles and distances are often used to
derive other quantities using mathematical relationships. E.g. heights are
obtained in levelling by subtracting staff readings, horizontal distances are
obtained from slope distances by calculations involving vertical angles, and
co-ordinates are obtained from a combination of horizontal angles and
distances.
In each of these cases, the original measurements will have errors associated
with them and it follows that any quantity derived from them will also have
errors.
The special law of propagation of variance (standard error) for a quantity U
which is a function of independent measurements x1, x2, .xn, where U = f(x1,
x2, .xn,) is given by:
Where,
U 2 U
s x1 +
sU2 =
x1
x 2
sU = standard error of U
are the standard errors of x1, x2, .xn
U
2
s x2 + ... +
x n
2
s xn

3A1 Lecture 3

Uploaded by

Copyright:

Available Formats

3A1 Lecture 3

Uploaded by

Document Information

Original Description:

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

3A1 Lecture 3

Uploaded by

Copyright:

Available Formats

Topic 1: Linear Measurement

Aims of this topic

Accuracy and Precision

Standard deviation and standard error

For statistical reason calculations of the standard deviation should be based

Calculate the mean, standard deviation and standard error.

Mean = 47o56 + (38+40+35+33+40+34+42+39+37) /10

You might also like