Traversing Project Report

Geographical and Earth Sciences

GUID: 0208837
University of Glasgow

Executive Summary
This report presents an overview of the work carried out at the University of Glasgow
facility for the period of two weeks commencing on Tuesday 11th November 2014. The
principle aim of this project was to map the west side of the main building using the
traversing method. The reference points used for the traverse method were attained using
GNSS and a check was performed to validate the final coordinates of the deduced control
point with the GNSS data. The results were downloaded onto a desktop and imported
into two software packages, named, Starnet and N4CE for post-processing respectively. A
detailed outline is stated implying any and all complications experienced during the project
and any solutions implemented to achieve a satisfactory conclusion. These conclusions
derived were critically evaluated for potential future improvements for any subsequent
work carried out.

Keywords: GNSS, traversing, levelling, total station, N4CE, Starnet

Introduction

There were three members, Manpreet Puri, Marcus Muller, and Mike Jaradeh participated in
this project and the tasks were completed as a group to mitigate any single user error and to
perform checks for every step taken. The main tasks completed within this report are stated
as the consequent separate sections, viz. Specification, Reconnaissance, Acquisition of ROs via
GNSS, Traversing Methodology, Post-Processing of Data and Conclusions. The main purpose of
this project is to provide secondary control from which the outer limits of the Main Building can
be surveyed with a total station.
In the present study, focus will be given to the complications arisen during the observation
stages, validating the final deduced results and comparing the results between both postprocessing packages. These control value results will then be assessed on their merit of being
implemented successfully into a 1:1000 scale detailed survey.
It should also be noted that the west side of the building was traversed by two separate groups.
This proved as a secondary point of attestation to confirm if any discrepancies or inaccuracies
that may arise from the project fieldwork carried out. Both groups used the same ROs and CPs,
displayed in figure 1 to carry out the field work but the total station and prism set up will be
different for the traversing fieldwork.

Figure 1: A Google Map view showcasing the approximate established Control Point locations

Specification

The specification stated was to provide satisfactory foundation values for a detailed survey of
the west side of the University of Glasgow building at a scale of 1:1000. The smallest point that
can be marked on the map is 0.2mm, which is 20cm. Therefore, the specification has to be twice
as accurate as this, implying that the smallest point that can be marked on this map to be 10cm.
While control points have to be a further accurate by a factor of 2, stating that control points can
be mapped with an accuracy of 5cm, or 50mm.
These values can be calculated from the allowable linear fractional misclosure and the allowable angular misclosure.

Allowable Fractional Misclosure

At 50mm accuracy, the allowable linear fractional misclosure can be stated as:
Fractional Misclosure =

1
accuracy/length o f traverse
1
= 2544.64 mm
50/127232

Hence, the allowable precision for this project is

1
.
2544.64

Allowable Angular Misclosure

The least count of the Leica TCR805 total station was 5 seconds, therefore, the allowable angular
misclosure can be calculated as:
2

Angular Misclosure = 3 least count

= 33.54102 seconds

(1)

where N is the number of legs within the traverse path and -33.54 is the allowable angular
misclosure.

Reconnaissance

This section briefly investigates the established location of the six control points (CPs) and two
reference object points (ROs) used within this project. One main criteria for this task was to
ascertain the CPs and ROs to be inter-visible. The ROs have to be selected in areas with no
over-cover i.e. tree branches, lamp posts, building extrusions and to be located faraway from the
connected first CPs but still maintaining a clear line of sight to it. The intermediate CPs had to be
chosen around the west side of the main building from where potential detailing can be carried
out from these CPs. Therefore, one key requirement during this assessment was to verify that
every corner of the west side of the main building was visible from each traversing CP selected.
As shown in figure 1 (image taken from
Google Maps), we can clearly see that the control points Z, B, C, D, E and A and the two
reference object points RO1 and RO2 are intervisible. The closed traversing path initiated
from RO2 and finished on RO1 Control points
Z, A, B, C, D and E will be used to capture the
outline of the building using Detailing, hence,
each control point for detailing must be able to
see another control point to act as a reference
object for radiation. These control points were
marked with wooden pegs and the position
was calculated using slope distance derivation
from the total stations and prisms used.

RO1 in different View-Points

A general description for each ROs and CPs

was mapped out in the field book for an approximate point selection and this was validated by camera photography. Most of the selected GNSS points had a number of associated
issues with them that could lead to potential inaccuracies in the recorded data required for
closed traversing to proceed.
Figure 2: RO1 from different View-Points

3.1

Reference Point RO1 and CP Z

Due to the site limitations on the south front the University of Glasgow main building, RO1 was
chosen on the pavement, beside a sign board as showcased in figure 2 for point verification and
CP Z was chosen near a pedestrian gate, north of the university car park. These selected points
had a number of issues associated with them, namely:
1 Steep hill facing the south side of the building, giving difficulty in the selection of
RO1
3

2 Points located near a busy private road where open access was necessary at all times
3 A number of parked vehicles which moved constantly throughout the day

3.2

Reference Point RO2 and CP A

These points were chosen at the north face of the University building. Due to the complex
topography of the University of Glasgow a number of complications arose from this, specifically:
1 The main gate for the campus was located at the North-West front that had constant
access throughout the day
2 A very busy main road which was the sole access to the University from outside
campus
3 Pedestrian crossing with sole direct access to the University library & other student
services from the main building

3.3

Final Location Proposition

Even though these complications can be severe at times, a number of key scenarios was analysed
to reach our desired locations. Warning traffic cones were attained from the University of
Glasgow security guards to make sure the tripod set-up was visible to oncoming traffic and
pedestrians. Locations were selected based on least disruption to surrounding architecture and
each equipment was used was manned for the duration of the fieldwork carried out. Another
factor that was taken into consideration was the weather forecast for the coming weeks. Control
point locations were selected on solid ground where water flooding would not be an issue.
From figures 2 and 3 it can be seen that the reconnaissance was carried out with the secondary
group on the 18th November 2014. This was a secondary attempt at reconnaissance as the first
attempt had failed due to the initial reconnaissance carried out not meeting the set requirements
of the project. A limit of six control points was given to each team due to the limitations of the
software packages post-processing this data but initially eight control points had been selected
for traversing. The extra two points arose from affirming that every intricate detailed corner
aspect of the building side was able to be mapped.

Acquisition of ROs and two CPs via GNSS

From the figures 2 and 3, the reference object points were marked with a yellow wax marker
for future verification and validation. The set locations were selected on the attributes of
the surrounding environment, such as area markers i.e. map board, gate hinge etc. for easy
identification. In order to achieve an accurate baseline, measurements were downloaded via
GNSS to the GNSS receiver every ten seconds over a twenty minute time-frame. In essence,
the GNSS receiver was set up over the pre-determined location and left to take measurements,
once twenty minutes had passed, the values were recorded for precision location determination.
These key steps for GNSS location measurement were also carried out for Control Points Z and
A. These two control points along with the two ROs acted as the baseline for the traverse to
begin and finish on respectively.

RO1 in different View-Points

Figure 3: R02 from different View-Points

One complication arose from the GNSS measurement reading at CP Z whereby the measurements downloaded from GNSS were not accurately saved onto the receiver. An error stating the
incremental steps had not been saved, this was the first time receiving this error and the group was
unsure on how to proceed. In order to make sure, accurate measurements had been taken and
recorded, the task was carried out three times by changing the input attributes on the receiver
for a period of twenty minutes each. From post-processing the data, it was deduced that the
all three separate time-scale measurements had been recorded and the error emerged from not
inputting a number after the control point value and therefore a standardized increment was
automatically set up by the GNSS receiver for subsequent measurements.

Traversing Methodology

In this section, a closed traverse was performed starting at CP Z and finishing on A. The closed
traverse was performed one two separate days as the measurements taken on the first day did
not meet the set requirements for the project. Characteristics defining both days are defined in
further detail below:

5.1

First Attempt

The first attempt at the closed traverse was performed morning/afternoon on Wednesday 19th
November 2014. As set in the traverse requirements, the initial point for the Leica TCR805 total
station was set up on CP Z with the backsight facing RO2 where a prism had been set up, sixteen
measurements with consecutive face left followed by face right of the total station were taken
to the backsight. Then the total station was manoeuvred to the foresight, named B (while not
moving the tripod set-up), and a repeat of sixteen measurements were taken to the forward sight.
once these measurements were satisfactory, the total station was moved to the prism tripod
set-up at B and the prism was set-up at the initial CP Z as the new backsight, followed by a new
5

From Control Point C facing D

and E

From Control Point C facing B

Figure 4: A collage showcasing the different intermediate CPs

prism tripod set-up at CP C as the foresight. These measurements were repeated at each control
point moving the total station to each point periodically until reaching the final CP A with
the foresight facing RO1. The main complication with these measurements was the structure
in which the readings were measured. Due to N4CE software limitations, the measurements
should have followed the structure: backsight face left, foresight face left, foresight face right
and then backsight face left, instead of using sixteen consecutive measurements facing one
direction. The reasoning behind this structure is the actual fact of measuring relative angles and
not point locations. In order to measure relative angles, we need consecutive measurements of
two different directions for the software to subtract the measured angles to derive the absolute
angles between each CP.

5.2

Second Attempt

Even though it was postulated that the measured values in the closed traversing attempt one
could have been re-aligned and re-ordering the measured data within the N4CE software. It was
concluded that this may take up valuable time during the post-processing phase of the project
as no member of the project team was vastly experienced in using N4CE. From this deduction,
the decision as made to repeat the traverse fieldwork on the weekend to mitigate any external
sources of error arising, such as large density of individuals moving around the set-up area,
university being closed, and favourable weather conditions predicted. The traverse was carried
out on Sunday 23rd November 2014 in the morning/afternoon and all data used within the
post-processing was collected on this day.
In order to perform checks for each measurements, height readings were taken for the target
locations, backsight and foresight along with height measurements for the instrument as well.
During the review process of these height measurements, it was deduced there was a discrepancy
between instrument height and target height at various CPs during the traverse. These height
measurements should have been the same for instrument and target as the height to the trunian

axis of the target prisms and total stations were the same with a negligible difference. In order
to see any inaccuracies within the measured data, it was concluded that further investigations
in regards to this anomaly would be completed during the post-processing data of the project.
The final angular misclosure determined was 12 seconds with a error in height of 1.6cm.
During the post-processing section of this report, we will discuss these results in detail but a
field explanation of why these errors may have arisen is stated below:
1 The level was not accurately focussed during initial set up, which can lead to parallax
within the observations.
2 The observer has used different cross-hairs to take the relevant measurements within
the same traversing path.
3 The total station was not focussed accurately on the prism due to light refraction
from rainfall on the day of observations.
4 The concrete slab at CP Z was loose and moved acutely with pressure on the left
hand corner, in return, transferring the acute instability to one of the tripod legs.
The explanations listed above should be taken as postulations and not relative reasons for the
final error reading.

Post-Processing of Data

A detailed closed traverse was carried out on the west side of the University of Glasgow main
building and all measurements taken were imported into N4CE and StarNet software packages.
The RO2 and CP Z acted as the initial baseline for bearing determination for any subsequent
measurements. The coordinates at each intermediate CP were derived from these baseline
values and the calculated bearing at the end of the closed traverse was compared with the
bearing determined from the GNSS coordinates. This section of the report is divided into two
subsections, namely, Post-Processing of N4CE software and Post-Processing of StarNet software.

6.1

Post-Processing of N4CE software

The data was imported into N4CE software for post-processing, relevant attributes were assigned
to the data to distinguish it accordingly. A traverse was plotted displaying the misclosures
before and after Bowditch adjustment and it was apparent from visual inspection that the errors
attained were too large to be satisfactory. A angular misclosure of 12 seconds was deduced
which is within satisfactory limits but a height error of -493cm was encountered.
A large discrepancy in the height measurement was expected from the difference in prism
and instrument heights noted in the field journal during the traversing phase. The instrument
heights at the initial CP was altered to the prism measured height rather than the total station
measured height. This adjustment lowered the height error to one third of its initial value but
the new derived value was still too large for post-processing. This complication resulted in
a detailed investigation into the equipment used and any human error that may have arose
during the fieldwork. After inspecting the traversing kit, it was noted that the total station kit
had two different tape measures with one damaged and starting at 0.38cm. It was also noted by
one of the project members that this tape measurer was used on the last CP A instrument/prism
measurement. When the value of 0.38cm was added to the height measurement of CP A,
7

the height error within N4CE changed to 16mm which is within acceptable range for postprocessing.
Therefore, a secondary traverse was computed for the altered measurements showcasing the
misclosures before and after a Bowditch adjustment. List of these misclosures can be seen in table
1 and it is apparent that the angular misclosure has been resolved using the Bowditch adjustment
with a slight change in the Easting and Northing values. From this table, it can be stated that
the angular misclosure is within allowable range while the linear fractional misclosure is out
with the satisfactory value. This could be from the reasons mentioned above or due to the GNSS
derived coordinates for our baselines having an accuracy of 2-3cm.
Misclosures
Before
After

Height (mm)
16.0
16.0

Angle
-12.0"
0

LF Error
1600
1662

Easting Control (mm)

59.5
56.2

Northing Control (mm)

-52.7
-51.9

Table 1: Table defining N4CE misclosures before and after Bowditch Adjustment
The N4CE report generated the Easting and Northing coordinates along with height for each
measurement taken in table 2. It should be noted that within this table the adjustment from the
calculated values for each CP vary drastically in a random manner.
Name
RO2
CP Z
CP B

CP C

CP D

CP E
CP A
RO1

Easting (m)

Northing (m)

Height (m)

256899.1700
256899.1746
256854.3000
256854.2870
256822.5117
256822.5114
256822.5244
256815.7968
256815.7945
256815.7979
256808.8689
256808.8671
256808.8691
256808.4293
256808.4272
256808.4302
256800.2000
256806.2401
256800.2000
256800.1932

666773.9100
666773.9180
666696.0300
666696.0276
666690.1040
666690.1030
666690.1435
666668.1405
666669.1276
666668.1435
666648.8592
666648.8530
666648.8675
666628.2243
666628.2161
666628.2358
666597.5200
666597.5084
666578.8800
666578.8883

42.5100
43.0689
38.6900
38.8465
38.8841
38.5608
39.0487
38.8841
39.1920
39.5146
39.5006
39.5108
39.4966
39.5600
39.5611
39.5638
39.7100
39.5638
38.7800
38.6503

Status
Known
Calc. BS
Known
Calc. BS
Adjusted
Calc. BS
Calc. FS
Adjusted
Calc. BS
Calc. FS
Adjusted
Calc. BS
Calc. FS
Known
Calc. BS
Calc. FS
Known
Calc. FS
Known
Calc. FS

Table 2: Report generated by N4CE Software expressing all measurements observed

6.2

Post-Processing of StarNet software

The unadjusted traverse measurement values were imported into StarNet software to apply
a misclosure correction using the Least Squares Adjustment method. It should be noted that

within StarNet only the intermediate CPs were adjusted while the derived GNSS coordinates of
CPs Z and A were fixed at all times along with both ROs.
The initial and adjusted values for the angular misclosure can be seen in tables 3 and 4
respectively. It can be seen from table 4 that the standard residuals increase as you increase
along the traverse path. If one were to draw a scatter plot for these residual errors along the
traverse path, it would depict the errors increasing in a linear path approximately.
Station Names
At From To
Z
RO2
B
B
Z
C
C
B
D
D
C
E
E
D
A

Non-Adjusted
Angle
229-27-50.00
117-33-21.00
182-45-46.00
161-27-43.00
182-51-24.00

Standard Error
16.18
26.28
35.62
37.16
31.41

Table 3: Table depicting Angular Misclosure derived by StarNet

Station Names
At From
To
Z
RO2
B
B
Z
C
C
B
D
D
C
E
E
D
A
A
E
RO1

Adjusted
Angle
229-27-47.46
117-33-36.92
182-47-31.67
161-30-51.87
182-54-31.61
193-58-50.24

Residual
-0-00-02.54
0-00-15.92
0-01-45.67
0-03-08.87
0-03-07.61
0-04-43.24

Standard Error
16.18
26.28
35.62
37.16
31.41
32.36

Standard Residual
0.2
0.6
3.0
5.1*
6.0*
8.8*

Table 4: Table depicting Adjusted Angular Misclosure derived by StarNet

The initial and adjusted values of the distance observations attained from post-processing in
StarNet can be seen in tables 5 and 6 respectively. From the non-adjusted and adjusted distance
observations, it can be implied that the standard errors and residual errors are sporadic and
have no direction correlation with the traverse path. However, this flawed result fits well with
our current deductions of possible errors associated with height measurements taken.
Station Names
From To Distance
Z
B
32.3240
B
C
22.9730
C
D
20.4929
D
E
20.6479
E
A
30.7938

Non-Adjusted
Standard Error
0.0156
0.0155
0.0155
0.0155
0.0155

Height of Instrument
1.965
1.645
1.543
1.673
1.569

Table 5: Table depicting the Distance Observations derived by StarNet

The traverse path was mapped in StarNet to give ellipsoidal errors at each control point while
this feature was not available in N4CE. It should be noted that figure 5 has been produced after
Least Squares Adjustment had occurred at each relative control point with a ellipse exaggeration
value of 20. From figure 5, it can be deduced that the ellipsoidal errors are elongated along the
observers view (northing) and not along the perpendicular field of view (easting). This can also
imply that there is a stronger error in distance rather than angular misclosure which sits well
within our previously stated deductions. The largest ellipsoidal error appears to occur within
9

Station Names
From To Distance
Z
B
32.2058
B
C
22.9757
C
D
20.4860
D
E
20.6714
E
A
30.8086

Adjusted
Residual
-0.1182
0.0027
-0.0069
0.0235
0.0148

Standard Error
0.0156
0.0155
0.0155
0.0155
0.0155

Standard Residuals
7.6*
0.2
0.4
1.5
0.9

Table 6: Table depicting the Adjusted Distance Observations derived by StarNet

Figure 5: Figure representing the Traverse path taken with the Ellipsoidal Errors
the intermediate control points along the traverse path. These control points had n validation
checks in place and therefore, no on-site checks could have been taken to deduce these errors
during the fieldwork excursion. It can also be deduced that partial correlation occurs towards
the mid to end of the traverse path where the ellipsoidal errors seem to deviate to a 45 degree
angle.

Conclusion

It should be noted that for greater the accuracy in measurement, the smaller will be the ultimate
misclosure to be eliminated; therefore, different methods in the adjustment will produce smaller
and smaller variations in the corrections, until the point where all error is eliminated. This can be
observed from the results presented in N4CE and StarNet, representing the differences between
the adjusted Bowditch and Least Squares method. Therefore, due to the variational errors
deduced from this report, clear comparison can not be analysed in depth and generalisation
statements would not fit well with the scope of this report.
It is apparent that this practical exercise was very helpful in understanding the classroom

10

theories applied in actual exercise. Team work was carried out well within the group and
relevant tasks were delegated successfully to be time efficient. A number of errors have been
listed in the preceding sections and it is apparent that not all objectives of the assignment were
met as if this project was to be further developed to a produce a detailed map at a 1:1000 scale,
the measured distance observations would fail to the project specifications set. These errors can
also be referred to the following reasons:
1 Poor planning of the execution of relative tasks leading to confusion and repeat
excursions for reconnaissance and traversing fieldwork.
2 Flawed note taking of instrument/target height measurements due to inconsistent
equipment usage.
Even though a number of errors had arisen, it was understood where the inaccuracies had
occurred and solutions can be derived accordingly to improve future fieldwork excursions. It
should be noted that although the Leica instrumentation shows it was a 1 second interval error,
the error given in the manual states it is 5 seconds.

Appendix

8.1

Equipment provided

In this section we will list all the materials, equipment, hardware and software provided prior,
during and post practical assessment.
8.1.1

Material & Information provided

The location for the base control station was provided and coordinate measurements were
downloaded from the Ordinance Survey database. Accurate GNSS reading were taken and
calibrated with these downloaded data sets. A full description of the task was provided along
with a map of the project area where potential locations of control and reference points could be
marked out. A task list stating objectives was also provided, outlining the requirements needed
to match each task assessment.
8.1.2

Equipment provided

The following equipment was provided for the practical assignment:

- Leica TCR805 total station
- Leica GST 20-9 tripod * 3
- Prisms * 3
- Wooden peg * 3 and wooden mallet
- Tribriach * 3
- Yellow wax crayon
- Measuring tape
- Field book
- Leica system 1200 static GNSS base station and receiver
11