Geographical

Information
Science
GEOG8700
Errors Using GPS

Oluwadamilare Olawale
Ajayi
2141139
Introduction

GPS (Global Positioning Systems) is an innovation that uses satellites to provide for one its position on the
Earth with the help of a GPS gadget. GPS might be fused into GIS by utilizing a GPS unit to gather points,
lines, or polygons, which could be imported into a GIS application for future investigation and analysis.
The GPS is made up of a series of 24 satellites that are in a 12-hour orbit of the earth, at 20,000km altitude
(Longley et al. 2011, p. 141). These satellites transmit a radio signal at accurate time intervals. The signals
are triangulated by a GPS receiver (GPSr) to calculate its approximate position. As the number of the visible
satellites increases, so too does the accuracy of the calculated position (Chapman, Thornes & Bradley 2002,
pp. 615-621). The motivation behind this activity is to learn the essential ArcMap functions by gathering
few global positioning system (GPS) locations and conduct some basic GIS analysis. In other words, the
main objective of this practical is to collect and input Global Positioning System (GPS) point data into
ArcMap and overlay these points on a coloured aerial imagery of the Flinders University campus and then
liken the GPS location with the corresponding areas depicted on the digital image.

Methods

Before proceeding into the field to collect the GPS points, locations were carefully selected by finding all
those areas on the image. Most of the features selected are well-defined features that can be easily be
identified both in the image and on the field which will can be called the pre-survey phase. These are the
things done in the laboratory before going into the field. During the next phase, we proceeded to the field
with a Garmin brand of GPS receiver to measure and collect the point positions of the locations selected
earlier by measuring the geographic co-ordinates represented using three coordinates: X,Y, Z: latitude,
longitude, and Heights. The Coordinates, Height, EPE (estimate probable error) and the description was
recorded from the GPS receiver on the paper for each point position measured. The final phase is what was
done when we got back into the laboratory after the 10 GPS points positions have been collected and
recorded on paper. These were then transferred into an excel sheet before been added on the ArcMap and
imagery to determine the measured error XY and measured error Z (height).

The measured error XY was determined by comparing the GPS location with the corresponding areas
depicted on the image data on the ArcMap and then recorded (refer to the table below). For instance, point 2
which is the right hand corner of the anchor with easting value of 278417, northing 6121218 and EPE of 6m
was calculated with a measured error 4m away from the original location on the image data. The highest
EPE we had was 5.8m for point 8(corner of service car park at bottom of service road) and the lowest was
0.8m for point 3(corner of green house).

Measured error Z (height) was calculated in the ArcMap choosing the nearest approximate contour line
value to the GPS point location on the image and then subtracting this value from the GPSr’s elevation. For
instance, for point 8(corner of service car park), the nearest contour line to this point was 135m but the
approximate value that was chosen for that point is 134m (143 – 134 = 9m) and in some cases like the point
3(corner of green house) where the point location is in the middle of the both contour lines, the average of
the contour lines were taken and then subtracted from the GPSr’s elevation to get the approximate contour

1
value used (110+115=225/2) =119-112.5=6.5m). In other words, points location measured error Z were
estimated based on either the average of the two contour lines or by choosing the closest approximate
contour value for that point based on the location of the point on the digital image and then subtracted from
the GPSr’s elevation.

Result

Point Easting Northing Height EPE Measured Error XY Measured Error Z(Height) Description
1 278356 6121213 124 6 4.1m 11m Grate in middle of rd outside anchor CRT park 7
2 278417 6121218 129 6 4m 15m Right hand corner of anchor
3 278341 6121160 119 6 0.8m 6.5m Corner of green house
4 278425 6121264 119 6 4.1m 10m Magazine stand facing stairs
5 278457 6121317 115 5 2.7m 9m Grate in middle of walkway in front lake
6 278435 6121041 134 4 0.9m 7m Silc bottom of ramp closest to carpark
7 278546 6121109 130 6 2.1m 6m Lamp post opposite earth sciences
8 278610 6121041 143 5 5.8m 9m Corner of service car parks at bottom of service rd
9 278632 6120902 152 5 4.6m 14m In front road sign opposite service rd engineering building
10 278520 6121144 135 8 5.3m 16m Fire hydrant earth sciences corner on lakeside walk

Table 1

Discussion

There are various elements or reasons that could decrease the accuracy and thus expand the potential error.
As said earlier before, the number of satellites the GPSr can get straightforwardly associates to its
correctness. Obstruction of satellites is the main cause of GPSr error (Chapman, Thornes & Bradley 2002,
pp. 615-621). The source of these impediments normally originates from close-by structures or geology, for
example, trees or buildings or valley; anything that decreases the GPSr’s clear view of the sky. The accuracy
also depends on the receivers; an inexpensive store bought receiver may have an accuracy of around 10m,
while a top end receiver may be accurate to within centimetres (Longley et al. 2011, p. 141). Differential
GPSr’s make use of signals received from the base stations of radio or telephone networks. As these base
stations have a fixed position, these signals can be used to offset GPS errors, improving overall accuracy.
This means that the differential receiver may work in places where a traditional receiver would not, such as
inside a building or under a bridge (Longley et al. 2011, p. 141). Other reasons for the variations in errors or
inaccuracies could be due to the following; Human error – coordinates been writing/typed wrongly, the GPS
timing (GPS clock) errors, image data/location and the on-screen measurement error, signal multipath and
bad PDOP and relief displacement.
Our GPS measurements were not resistant to blunders, with inaccuracies ranging from under 1m to
above 5m. We recorded the Estimate Probable Error (EPE) from the GPSr as both coordinates were taken
and then used ArcMap to determine the measured error (refer to the attached table for results). The results
showed that the EPE and measured error were quite different, with the EPE value being an overestimate of
the actual error in every instance. For example, point 5 had an EPE of 5m, while the measured error from
ArcMap was only 2.7m. Despite the variations in error estimation, there is still an acceptable correlation, in
most cases, between the amount of sky visible to the GPSr and the accuracy of the recorded point. For

2
instance, point 2 was recorded with a tall building near it and also nearby trees with a measured error of 4m
while point 6 was recorded in an area away from buildings and trees, with a measured error of 0.9m. Other
example of poor accuracy caused by obstruction includes point 8 which is the corner of car park at bottom of
service road next to a high vertical wall. The exception to this was point 3 corner of greenhouse which was
located down with reasonably dense tree cover and a near tall building. The EPE was 6m while the
measured error was only 0.8m making this the most accurate measured result. The reason for this could be
put down to chance that the GPS receiver was able to correctly estimate its position despite the high EPE.
Other sources of errors that could affect our readings are;
Atmospheric conditions - when the ionosphere and troposphere both refract the GPS indicators, this causes
the velocity or speed of the GPS motion in them to be different from the rate of the GPS motion in space.
Signal Multipath and bad PDOP - a multipath error occurs when receiving GPS signals reflected from
surfaces of ground or other objects in the vicinity of a receiver, and leads to a positional error resulting from
the computation of the range between the satellite and receiver (Cetin Mekik & Ozer Oman, 2010, pp 2186-
2200). It is when a GPS signal bounces off a reflective surface when close to an obstruction (e.g buildings or
trees) before arriving at the GPS receiver antenna and poor distribution of the available satellites.
Human error – these are mistakes made by writing or typing the wrong coordinates of the point location.
Relief displacement – these are the effects landscape/undulating terrains can have on image distortion, this
could have a great effect on image precision which is apparent in this imagery.
The GPS timing (GPS clock) errors, image data/location and the on-screen measurement error – these are
errors emanating from the GPS, image data and the on-screen measurement used in the ArcMap to compare
the distance between the GPS point location and the exact point on the imagery

Conclusion

The Global positioning system is a framework which gives the exact location based on information
transmitted from a heavenly body of 24 satellites. To calculate the point position X, Y, Z (latitude, longitude
and elevation) using a GPS receiver, four or more satellites must be visible. With some consideration for
error, the GPS can give any point on earth with a special location (its exact area). From the analysis of the
GPS observations in the table above, it shows that errors exist in our readings and the sources of these errors
could include human errors, signal multipath, atmospheric error, relief displacement and errors associated
with the GPS.
In conclusion, GPS lets us know that we are at point X,Y, Z while GIS lets us know that X,Y,Z is a feature
like a building or a tree. In other words, GPS tells us the "where" while GIS on the other tells us the "what".
GPS and GIS work hand in hand; the GPS has the ability to find features which adds such a great amount to
the usefulness of the GIS system. On the other hand, without feature data, a coordinate location is of little
significance.

.References

 Cetin Mekik & Ozer Can, 2010, An Investigation On Multipath Errors In Real Time Kinematic Gps
Method’, vol. 5(16), pp. 2186-2200
 Chapman, L, Thornes, J & Bradley, 2002, ‘Sky View Factor Approximation Using Gps Recievers’,
International Journal Of Climatology, vol. 22, January, pp. 615-621
 Longley, P, Goodchild, M Maguire, D & Rhind, 2011, geographic Information Systems and Science,
3rd edition, Wiley, USA, p.141

3
 Errors Using GPS at Flinders Univesity

"
) 90
"
)9
"95 0
)
"
)
75

10 ) " "
)
85
"
)

"
)
80

100
5

"
)
"
)85

90

Grate in middle
"
)
95

of walkway in
front lake
"
)
100 !

"
)
105 Magazine

"
) 11
0
stand facing
stairs
!
Grate in middle
of rd outside
anchor CRT!park 7
!
Corner of
Right hand
corner of
"
)11
5
"
)
95

green anchor Fire hydrant earth

house sciences corner
!
on lakeside walk
120

! Lamp post
)
"

opposite earth
sciences
!

Silc bottom of "

)
130 Corner of service
ramp closest car parks at bottom

)5
"
12 !
to carpark
!
of service rd

)5
"
13
"140
)
145
)
"

150
"
)

In front road sign

opposite service rd
engineering building
!
"
)0 10

"
)95 " 1)
)
110
90
"
)

"05 95 1"
"85
) ) )00
" Aerometrex

! GPS points
Contours 5m

Source: Aerometrex 2012

± Ajayi Oluwadamilare(2141139)
Date: 18/08/2014

Coordinate System: GDA 1994 MGA Zone 54

0 50

Metres
100

Page 4