CHAPTER 2

Theory of Errors and Basic Adjustment Principles

Introduction
 Measurement
- Observation carried out to determine the values of quantities :
- distances,
- angles,
- temperature …
 The process of taking measurements involves physical operations like:
- setting up,
- calibrating,
- pointing,
- matching,
- comparing etc of the instrument.
 The fundamental principle of measurement of surveying is that no measurement is exact .
 No matter how a sophisticated instrument used and all necessary cares are taken the result of a survey measurement
will contain some error.
 This is due to:
- Lack of perfection by the surveyor in his senses of seeing touching, hearing.
- Imperfection by the instruments and methods (in construction and adjustments and
environmental factors in their operation and approximations etc) 1
 … … … ... … CONTINUED

 Therefore, it is theoretically impossible to get the ”exact” value or ”true” value of any
measured quantity.
 WHY?
- All of our measurements will contain some error and
- The “ exact” value should contain infinite significant digits
(Which is practically impossible).
 why so measuring, as we cannot get the “exact “value of a measured quantity?
- In surveying and generally in any scientific measurements what is important is
not the “exact’’ value rather the possibility of carrying out the measurement to
the degree of accuracy sufficient to the desired purpose.
 Hence, as long as the desired accuracy is achieved our measurements can be used as an
equivalent to the “true “value.

2
 MEASURES OF QUALITY

 Accuracy:
- indicate the closeness of measured value to the “true’’ or ‘’exact” value of a quantity.
- It indicates the degree of perfection obtained in measurements.
- The further a measured value from the “true” value the less accurate it is.
True error = (Measured value – True value)
Relative error = (Measured value – True value) / Measured value.
 Precision :
- Closeness of measured values to one another regardless of their closeness to true value.
 It is possible to obtain both accuracy and precision by exercising care, patience and using good
instruments and procedures.
 In measuring distance, precision is
defined as the ration of the error
of the measurement to the
distance measured.
 A measurement can be:
- accurate but not precise;
- precise but not accurate;
- precise and accurate; and
- neither precise nor accurate
3
 ERROR

 Deviation of the measured value from the “exact” value of a quantity.

 The study of errors helps to:
- Understand the sources of error and
- Exercise the necessary care and
- Apply corrections.
Error sources
- Classified as:
 Personal:

- Occurs due to lack of perfection in the surveyor’s sense of sight, touch, hearing.
- Mistakes due to carelessness or fatigue of the surveyor.
- Can be minimized with care and vigilance.
 Instrumental:

- Occurs due to imperfection of the instruments :

- In manufacture and
- During adjustments.
- Mistakes due to failure or damage of the instrument.
- Can be minimized with:
- Careful handling,
- Maintenance and adjustment of instruments, and
- Applying corrections.
 Natural:

- Errors due to effect of temp, pressure, humidity, magnetic variation etc.

- Can be minimized :
- By applying correction and
- By carrying out the survey when their effect is minimal. 4
 TYPES OF ERROR

 Classified in to three;
- Mistakes,
- Systematic errors and
- Random errors.
Mistakes
- Usually are gross in magnitude compared to the other two types.
- Blunder made by surveyor or his equipment.
- Can occur at any stage of the survey :
# During - reading,
- recording
- computing and
- plotting.
- Source: May be due to one of the following
- Carelessness or fatigue by surveyor
- Failure of equipment
- Examples :
- Reading wrong scale
- Transposing figure in recording ex 56 instead of 65 or reading 6 instead of 9.
- Omitting digits during recording ex 200 instead of 2000
- Sighting towards wrong target etc.
- Generally, these class of errors can be easily detected and eliminated.
5
 … … … ... … CONTINUED
 Some of measurements to be taken:
- Taking multiple independent readings and checking consistency.
- Careful checking of both sighting on targets and recording.
- Using simple and techniques for verification, applying logic and common sense.
- Checking and verifying performance of equipment.
- Repeating the measurement with slightly different techniques or datum.
- Applying geometric or algebraic checks to detect mistakes.
- Simply noting that mistakes have large magnitude so can easily be detected.
 Note: an observation with a mistake must be discarded unless the mistake is removed, otherwise, that observation
Systematic errors
- Relatively small magnitude compared to mistakes.
- Are result of some systems whose effect can be expressed in mathematical relations.
- Their magnitude and sign can be estimated (determined).
- Sources: can be : - Personal,
- Instrumental and
- Environmental conditions or
- May be result of choice of geometric or mathematical model used.
- For constant conditions they remain the same in sign and magnitude:
- Repetition will not help in detection and elimination.
- As their values can be determined correction can be applied.
- Calibration and adjustment of instruments contribute to minimizing their effect. 6
 … … … ... … CONTINUED
Random Error
- Errors that remain after elimination of mistakes and systematic errors.
- The magnitude of this type can be estimated.(standard deviation)
- Unlike that of systematic errors their sign cannot be determined.
- Hence correction cannot be applied.
- Thought to have normal probability distribution.
- Their effect can be greatly decreased by exercising care and vigilance
and by using high precision instruments.
Significant figures
- Those digits in a number that have meaning.
- Can be any one of the digits 1, 2, 3… 9.
- 0 is a significant figure except when used to fix a decimal point.
- EXAMPLE:
- 0.00456 has three significant figures and
- 45.601 has five significant figures.
- 0.0240 has three significant figures.
 Rounding off a number is the process of dropping one or more digits so that the answer contains only
significant digits. EXAMPLES: - 78.3749  78.37
- 78.375  78.38
- 78.385  78.38 7
 ADJUSTMENTS

 The given values of the observations must be altered or adjusted.

 To make them consistent with the model, leading to the uniqueness of
the estimated unknowns.
 Several techniques can be used to adjust redundant measurements. The
most rigorous and commonly used is the method of least squares.
 Other approximate adjustment techniques can also be employed.
 If it is assumed that measurements are of equal quality, or weight, the
residual could be divided equally among the observations.
 If these were not true and the different weights of the observations
could be estimated, then each observation would be assigned a residual
that is somewhat different from the others.

8
 MOST PROBABLE VALUE
 Different conditions, under which the measurements are made, cause variations in measurements.
 A fixed value of a quantity may be conceived as its true value.

 The difference between the measured quantity and its true value  is known as error, , i.e.,

= x - 
 Sincethe true value of a measured quantity cannot be determined, the exact value of  can never be found.
 However, if a best estimate xˆ which is known as the most probable value of , can be determined.

 Most probable value can be used as a reference to express the variations in x.

 If we define  as residual then,

 Theresiduals express the variations or deviations in the measurements.

 Standard Deviation
 Measure of spread of a distribution and for population, it can be expressed as

- μ cannot be determined from

a sample of observations.
- Instead, the arithmetic mean is accepted as
the most probable value.

 The standard deviation given by the above expression is also called the standard error.
9
 … … … ... … CONTINUED
 The population standard deviation is estimated as

Variance

- Variance of a quantity is expressed as:

- used as a measure of dispersion or spread of a distribution. 10

 STANDARD DEVIATION OF THE MEAN

 Standard error indicates the limits of the error bound within which the ‘true’ value of the
mean lies, with a 68.3% certainty of being correct.
 Given by, m

11
 WEIGHT

 - is a parameter indicating the relative reliability of recorded values .

- The assignment of weight to a measurement is a matter of judgment.
- For measurements carried under similar conditions the weight is assigned
proportional to the number of observations.

- Also if variance of the measurement is known it is assigned proportional to 1/variance (the

inverse of variance).

Vvvvvvvvvvvvvvvvvvvvvvvvvvvvvvvvvvvvvvvvvvvvvvvvvvvvvvvvvv Relative
precision
12
 EXAMPLES

1. The length of a base line was measured using two different EDM instruments A
and B under identical conditions. The following results were recorded.
Determine the relative precision of the two instruments and the most probable
length of the base line.

13
SOLUTION

14
… … … … CONTINUED

15
… … … … CONTINUED

16
 … … … … CONTINUED
2. An angle was measured with different weights as follows, determine
a. The most probable value of the angle,
b. The standard deviation of an observation of unit weight,
c. The standard deviation of an observation of weight 3, and
d. The standard error of the weighted mean

17
 SOLUTION

Can get,

18
 … … … … CONTINUED
b)

19
… … … … CONTINUED

20
 EXERCISE

 Three angles α, β, and γ are measured in a plane triangle.

The respective weights estimated for the angles are ωα,
ωβ, and ωγ respectively.
What would be the adjusted angles, if:
 A. α = 90, β = 43, γ=44, and ωα = ωβ = ωγ = 1
 B. α = 90, β = 43, γ=44, and ωα 1 =1, ωβ = ωγ = 2

21
 CHAPTER 3

Basic Survey Measurements

22