Surveying & Measurement

Accuracy and Precision

 Accuracy and Precision
• What is accuracy?
 Accuracy is the nearness of a measurements
to the true value.
• What is precision?
 Precision is the amount by which a
measurement deviates from its mean.
 Accuracy and Precision
• Suppose that you measure the same line
five times.
• The first party reports the following
measurements: 736.80, 736.70, 736.75,
736.85, and 736.65 m. [more accurate]
• The second party reports the following
measurements: 736.42, 736.40, 736.40,
736.42, and 736.41 m. [more precise]
• The true length of the line is 736.72 m
 More on Errors
• Gross Errors and Mistakes - These are serious
mistakes made by surveyor, such as reading
15.45 instead of 5.45 or writing 9.64 instead of
6.94.
• Systematic Errors - These are due to
instruments or operations, and have cumulative
effect.
 As an example, you have a meter to measure
distance with a production error. It shows 1 meter but
in reality the distance is 99.5 cm. Thus, the
measurement always longer than the real distance.
• Accidental Errors - These are unavoidable
errors arising from weather condition, change in
temperature, humidity, mood of observer, etc
 Rules
• There are several rules to ensure that
measurements are made and calculated using
the correct degree of precision:
• Rule 1: Recording Record a measurement with
a number of figures appropriate to the
measuring and computing precision, with one
extra digit to avoid future round-off errors.
• Rule 2: Adding & Subtracting In adding or
subtracting, the number of significant figures in
the sum or difference is determined by the
fewest decimal places in the numbers involved.
 Rules
• Rule 3: Multiplying & Dividing. In
multiplication or division, the number of
significant figures in the product or
quotient is the same as the fewest number
of significant figures in the measured
values used, conversion factors or
constants having no influence on the
significant figures in results.
 Rules
• Rule 4: Avoiding Round-Off Errors To
avoid computational round-off errors:
 Record values correctly as given in Rule 1.
 Use at least one digit more in constants and
conversion factors than is in the measured
value with the fewest significant figures.
 Carry one extra digit in computed quantities to
avoid round-off errors in computing other
quantities.
 Rules
• Round values according to this standard
procedure:
 When dropping a digit lower than 5, write the number
without the digit. So 87.373 becomes 87.37.
 When dropping a 5, use the nearest even number for
the preceding digit. (this helps balance results from
series computations) So, 87.375 becomes 87.38 and
87.385 also becomes 87.38.
 When dropping a digit greater than five, increase the
preceding digit by one. So, 87.376 becomes 87.38.
 Four Components of
A Measurement
• An estimate of the size of the measured quantity
“The distance from my home to the market is 4.1”
• The units for the measured quantity
“The distance from my home to the market is 4.1 km.”
• An estimate of the range of error, or uncertainty
“The distance from my home to the market is 4.1 (±0.2)
km.”
• A level of certainty or level of confidence
concerning the range of error.
“I am 95% confident that the distance from my home to
the market is 4.1 (±0.2) km.”
 Expressing Uncertainty
• Absolute Uncertainty – an uncertainty
stated with the same units as the value.
E.g.: “23.3 (±0.2) kilometres”
• Fractional Uncertainty – the ratio of the
absolute uncertainty to the value. This is
generally used in uncertainty calculations,
but the final results are generally given in
absolute (or sometimes percentage)
uncertainties.
 Expressing Uncertainty
• Percentage Uncertainty – the fractional
uncertainty expressed as a percentage.
(must be rounded to one significant digit
because they are uncertainties.)
• E.g.: “23.3 km (±0.9%) kilometres”
 The Statistical Nature of Errors
• Real-life decisions come with a degree of
uncertainty.
• In surveying much of uncertainty comes
from the random errors left in
measurements after the blunders and
mistakes have been removed and the
systematic errors have been compensated
for.
 Mean
• Mean – the arithmetic average of a series
of n measurements
 Standard Deviation
• Standard Deviation – a measure of the
spread of values in the normal distribution.
(a measure of precision – a distribution
with a smaller standard distribution has
less spread, and represents more precise
measurements)
 Standard Deviation
• The standard deviation is used to measure
the precision of the method.
• A larger standard deviation indicates
larger random errors and lower precision.
• If the normal distribution probability density
function is plotted, the area under the
curve between plus and minus “sigma” is
68.27 percent of the total area.
Normal Distribution
 Laws of Probability
• Laws of probability

Increased precision Decreased precision

 Percentage Likelihood
• The percentage likelihood of any error’s
probability can be determined in the same
way.
• Ep = Cp ×σ
Where:
Ep is a certain percentage p error, and
Cp is the corresponding numerical factor taken
from studies of the normal distribution.
Probable Error
 Probable Error
• Probable Error – (E50) The error
establishing limits within which the
measurements should fall 50% of the time.
• Maximum Probable Error – (approximately
E99.5) The error establishing limits within
which all the measurements should fall.
(Any measurements outside this range
should be examined for mistakes in
method, recording or other atypical error
sources).
 Error Spread
in Adding and Subtracting
• Examples:
 A surveying team might use a 50 metre tape
to determine a long distance and add several
readings together in order to calculate the
total distance
 Several land areas may be added together in
order to determine a total land quantity and
value
 A surveying team may calculate several
volumes of cut or fill and add them together to
calculate a total site cut or fill
 Error Spread
in Adding and Subtracting
• Simply adding together the magnitudes of
all the errors in a sequence of
measurements is likely to overestimate the
total error.
 Error Spread
in Adding and Subtracting
• The best way to estimate random error
propagation is by a Pythagorean-type
approach, explicitly recognizing the
independence of the errors in each of the
added components.
 Error Spread
in Adding and Subtracting
 Error Spread
in Adding and Subtracting
• For a series of n measurements all with
error E
 Error Spread
in Adding and Subtracting
• Example: A surveying team tapes four
lengths each with an error of ± 0.005
metres. The four lengths are added
together to determine a total length. What
is the total error in the measurement?
 Error Spread
in Adding and Subtracting
• Example: A distance of 1500 metres is to
be taped with an error of no more than ±
0.100 metres, using a 100 metre tape.
How accurately must each 100 metre
length be measured?
 Error of The Mean
• The most likely value of a group of
repeated measurements is the mean.
• if we divide the error in the sum by the
number of measurements we will know the
error in the mean:
 Error of The Mean
• Example: A surveying team makes six
measurements of an angle using a hand-
held compass. Each measurement has an
error of ±0.5 degrees. The mean of the
measurements is 56.3 degrees. What is
the error in the calculated mean?
 Error of The Mean
• Determine the error in the mean at any
percentage probability, the appropriate
value of Epercentage from the normal
distribution:
 Error of The Mean
• Example: What is the standard error of
the mean and the maximum probable error
of the mean for a set of 12 measurements
with σ = 0.09 metres?
 Weighted Measurements
• Situation: have greater confidence in some
values
 How to determine weight?
 Commonly use weight derived by inverse
proportion

1 1
p= or p =
factor factor 2
 Weighted Measurements
Example
Group 1 Group 2 Group 3
2.4 2.8 4.5
3.0 3.0 3.2
3.1 3.2 5.1
Sum 8.5 9.0 12.8
Mean 2.8 3.0 4.3
S.D. 0.38 0.20 0.97
(2.83) + (3.00) + (4.27)
Regular Mean = M = = 3.37
3
 Weighted Measurements
Example
Group 1 Group 2 Group 3 1
pi =
2.4 2.8 4.5 (s.d.)2
3.0 3.0 3.2
3.1 3.2 5.1
Sum 8.5 9.0 12.8
Mean 2.8 3.0 4.3
S.D. 0.38 0.20 0.97
Weight 6.96 25.0 1.06

(6.96) × (2.83) + (25.00)(3.00) + (1.06)(4.27)

Mw = = 3.00
(6.96) + (25.00) + (1.06)
 Units of Measurement
• Linear distance units
 Historic: Gunter’s chain: 1 chain = 66 feet =
4 rods = 100 links
 English: international foot, US survey foot,
yard, mile
 Metric: meter
 Units of Measurement
• Angular distance
 Sexagesimal: Degrees/minutes/seconds: 360°
= circle
 Decimal degrees
 Grads: centesimal minutes and seconds:
400g = circle
 Mils: 6400 mils = circles
 Radians
 Units of Measurement
• Area units
 English: Acre: 1 acre = 1 chain x 10 chains =
43560 ft2
 Metric: Hectare: 1 ha = 100m x 100m =
10,000 m2
• Volume units
 English: Cubic yards, Acre-feet
 Metric: Cubic meters
• The U.S. and the metric system