System 1200 Newsletter – No.

28
Resections
total station. The location of the total station
RESECTIONS WITH TPS1200 could be anywhere on a circle which has its cen-
tre at the measured target point and radius of the
The last newsletter gave some background in- distance measured to the first target point.
formation into all the setup methods which are
available with TPS1200. After measuring the second target point the loca-
tion of the total station is ambiguous – it could be
This newsletter focuses on the Resection at one of 2 locations where 2 circles intersect.
method and will cover the basics - the next But knowing where you are in relation to the 2
newsletter will go into a much deeper level – so target points, you can choose the correct loca-
ensure to read and understand this newsletter tion – and you have now determined the coordi-
first! nates of the total station.
But now measure the third target point – what
RESECTION – THE BASICS does this mean? We now have “redundancy” –
or put another way, we have more measure-
A Resection (also sometimes called a Free- ments than we need. This can be visualised as
Station) allows the coordinates (easting, nor- the intersection of 3 circles as shown below. But
thing and height) and orientation of the TPS sta- because the measurements will not be perfect
tion to be determined by measuring to a number (and probably, the coordinates of the target
of known (target) points. points are also not perfect) then circles will not
So this method is perfect to use where you need perfectly intersect.
to have the freedom to set up the total station
where it is most useful – the only restriction be-
ing that you must be able to see a number of
target points. The coordinates of the target
points can be known in position and height, posi-
tion only or even height only (more on this later).
The coordinates of the total station are computed
using a least squares method. Next week’s
newsletter will also introduce a method known as
the robust method, but for now, only the least
squares method will be discussed.
So how can we now decide where is the “cor-
INTRODUCTION TO LEAST SQUARES rect” or even the “best” location of the total sta-
tion? Is it exactly in the centre of the “curved”
Before looking at the resection application itself triangle? Is at one of the intersection points of 2
on the TPS1200 it is worthwhile to get a basic of the circles? The “correct” answer can never be
understanding of least squares using a simple known, but we can make a “best guess” using
example. the method known as least squares.
But what does it mean to “use the method known
LEAST SQUARES – SIMPLEST CASE as least squares”?
Imagine your total station can only measure dis- Imagine that we guess that the most likely loca-
tances – it cannot measure horizontal or vertical tion of the total station is somewhere near the
angles (it could be a DISTO on a tripod!). You centre of the triangle and compute the coordi-
need to determine the easting and northing co- nates of the centre of the triangle. It is now pos-
ordinates of the location where you set up the
sible to compute the distance from this “guess”
total station. to each of the target points (the “computed”
You set up this total station in a location where distance). This distance will not be the same as
you can measure a distance to 3 known target the distances we originally measured (the “ob-
points. You now measure a distance to each of served” distance).
the 3 target points individually. Knowing the “computed“ distance and the “ob-
After measuring a distance to the first target served” distance, it is possible to compute the
point, you still do not know the coordinates of the difference between these 2 distances – the “re-
sidual”. This can be repeated for each of the

BU Surveying
Surveying & Engineering Division
 System 1200 Newsletter – No. 28
Resections
target points. We then know the residuals for Or maybe the coordinates of the target point are
each of the target points. We must now square wrong (or maybe the easting and northing is
each of these residuals and then add them to- correct and the height is wrong). To handle this,
gether which gives us the sum of the squares of it must be possible to use only the 2D position or
the residuals. height only coordinate of the target point – this is
also possible using the TPS1200 resection
Putting it in the simplest terms, the least
method.
squares methods states that the most likely
solution is the one where the sum of the
squares of the residuals is minimised. SOLVING FOR SCALE
So we could now repeatedly choose different Imagine you need to complete a survey in an
locations for our total station, calculate the “com- area which is surrounded by target points for
puted” distances, compute the residuals and which the coordinates are known. You are sure
compute the sum of the squares of the residuals. that the points are relatively homogeneous (that
Eventually, we would find the location where the is, they fit together well), but you have no idea
sum of the squares of the residuals is minimised about how the points were established. Were
and this would be the most likely solution (this is they measured using GPS? Were they meas-
why this method is called the method of least ured using TPS – if yes then what ppm value
squares). In reality this process is automated, was used? Were they measured “on the ground”
but fundamentally remains the same - the aim is or “on the grid”?
to find the solution where the sum of the squares For your survey, it actually does not matter how
of the residuals is minimised. the target points were measured – it is only nec-
This process is used in the resection method. essary that you can compute the coordinates of
the total station which “fits” or “ties into” these
points.
LEAST SQUARES – EXTEND THE BASIC IDEA
In order to do this, it is necessary to also com-
Now that the basic concept of least squares is
pute the scale factor. This is basically another
understood, it is possible to extend this idea.
“unknown” – the coordinates and orientation of
Imagine the more realistic problem of computing the total station are unknowns (it was already
the easting, northing and height coordinates of described how to compute these unknowns), the
the total station and the orientation using the scale factor is simply an additional unknown.
resection method. Additionally, this time, the total Once the scale factor is computed, this scale
station method has also measured horizontal factor can also then be used when all subse-
and vertical angles, has measured distances to quent measurements are made – thus the com-
only some points and only the easting and nor- puted coordinates of the total station and all
thing coordinates of some target points are measurements “fit” with the control points.
known – for others, only a height is known.
This problem is much more complex – there are THE RESECTION METHOD ON TPS1200
many more measurements (both angular and
distance) and there are more unknowns to solve This section will now describe step by step how
(easting, northing, height and orientation) but the to use the resection method on the TPS1200.
fundamental way of solving the problem is the
same as before. CHOOSING THE METHOD

BAD OBSERVATIONS, BAD DATA? In the SETUP Station Setup panel, choose Re-
section at the Method prompt.
Imagine now that one of the measurements to
the target points is wrong (maybe the wrong It is also necessary to choose the Fixpoint job
reflector type was chosen). To handle this, it (this is where the target points are stored), to
must therefore be possible to identify and then enter the Station ID (what point ID should be
remove this measurement from the computation given to station being computed) and the In-
so that the result (the coordinates of the total strument Ht.
station) is not influenced by the “wrong” meas- The F5(PPM) button allows you to quickly ac-
urement. With the TPS1200 resection method, it cess the CONFIGURE TPS Corrections panel if
is possible to disable individual measurements you need to check the current atmospheric or
and re-compute the resection. geometric corrections.

BU Surveying
Surveying & Engineering Division
 System 1200 Newsletter – No. 28
Resections
SETUP Find Target panel – which is similar to
the normal stakeout panel. Alternatively you
could press shift F4(POSIT) to turn the instru-
ment to the point (this may be useful if you did
not configure the instrument to automatically turn
to the point).

If you need to make changes to any configura-

tion settings then press shift F2(CONF) to ac-
cess the SETUP Configuration panel (the set-
tings on this panel were described in last weeks
newsletter). Press F1(CONT) to access the
SETUP Measure Target 1 panel.
Once the instrument is aimed at target point 3,
MEASURING THE TARGET POINTS then it is possible to already get an indication of
Select the target point which is to be measured, the quality of the resection by looking at the
aim the instrument at the point and measure the ∆Azimuth, ∆Horiz Dist and ∆Height values.
point. Store the measurement and press F5(CALC) to
access the SETUP Results panel. Note, you
may now see some warning panels if the accu-
racy tolerances are exceeded (there are different
messages for the positional calculation, height,
orientation and scale calculations). If no solution
is possible (you have probably measured to the
wrong points) then another message would be
shown.

THE RESULTS OF THE RESECTION

This panel contains 4 page views.

Once the point is stored, the SETUP Measure

Target 2 panel will appear (it is possible to
measure up to 10 points).
Again, measure to this target point – once this
point is stored it is possible to compute the coor-
dinates of the station and hence the message
line Press CALC to display results is shown.
In our example, we will measure one more target
point. Since the coordinates of the station can
now be computed, the instrument (if configured)
will turn automatically to the next point.
The Stn Coords page is straightforward and
To help you find the next target point more eas-
shows information about the computed station,
ily, you can press shift F2(FIND) to access the
how many points were measured etc. Here it is

BU Surveying
Surveying & Engineering Division
 System 1200 Newsletter – No. 28
Resections
possible to decide what should be “set” at the Since we also chose to compute a scale, the
Set prompt. result of the computed scales is shown (remem-
ber, we computed this in order to “fit” to the exist-
The F3(ROBST) button allows a “robust” solution
ing control points). The Set as prompt allows this
to be computed – this will be explained in more
computed ppm value to be “used” as the current
detail in the next newsletter.
Geometric ppm– this would make sense in
The F4(INFO) button accesses the SETUP Ad- most cases since we do want our survey to fit to
ditional Information panel which allows the the control points.
residuals relating to the individual measurements
to be viewed – remember, the residuals are the
computed values minus the observed values.
The next newsletter will show that sometime the
observation with the biggest residuals does not
always mean there is a problem with that obser-
vation!

The Stn Code page allows a point code to be

attached to the station and the Plot page gives a
graphical plot of the results. The points used in
the resection will be shown in black, all other
points are grey.

It is also possible here to decide if each observa-

tion should be used as a 3D, 2D or 1D observa-
tion or if the observation should be excluded
from the calculation. More on this panel next
week – but for now, press F1(RECLC) to recom-
puted the resection and return to the SETUP
Results panel.
The F5(SURVY) button returns to the panels
where you can measure the target points – you
can either survey additional target points, or re-
measure previously measured points (in this
case, the measurement to the previously meas-
ured points is “over-written”). The F5(SURVY)
button is the same as pressing ESC. Remember to use shift F3(FIT) to fit the plot to
all points, use shift F3(FIT R) to fit the plot to the
The Sigma page gives an indication to the qual- result of the resection.
ity of the computed solutions (the solutions in
this case are the Easting, Northing, Height and Finally, F1(SET) stores the resection result to the
Orientation). The sigma values are computed database and sets the new coordinates and ori-
as part of the least squares method – the bigger entation of the total station.
the residuals of the observations (or the worse
the accuracy of the coordinates of the control SUMMARY
points) the higher the sigma values of the com-
puted coordinates and orientation - more on this The next newsletter will go even more deeply
in the next newsletter. into resections – be prepared!

BU Surveying
Surveying & Engineering Division