This Master’s Thesis has examined the parameters that affect the precision and reliability
of a geometric network. It has been discovered that the precision and reliability of a
geometric network rely on the precision of the observations and the geometry of the system.
Hence thirteen parameters have been identified, and five of these have been examined
further. Those are:
1a. Placement of control points and setups.
2a. Types of observations.
1b. Precision of the total station.
7b. Lengths of sight from the station point to measured points.
1c. Precision of levelling instrument and rod.
The research has taken place through computations that forecast the precision and reliability
of a certain network design. With six initiatives in mind, different network designs
have been shaped and specific instruments have been chosen. The computations have been
completed with the program GeoADJUST since it provides every concept about precision
and reliability that has been described, based on studies of previous literature. Comparisons
between the results of these network designs disclose how the listed parameters affect
the precision and reliability of a geometric network. On the basis of the computations and
the percentage increases that the comparisons have led to, guidelines have been established.
It has been revealed that the parameters corresponding to the precision of the observations
have the greatest effect on the precision and reliability of a geometric network.
The geometry of the system comes in second. It requires no additional time in the field if
the surveyor makes sure to utilize a total station that provides observations with a higher
precision. Perhaps fewer observations then will be needed, and the extern reliability will
be improved.
Geometric levelling has proven to uprate the precision and reliability of the level, H, to a
great extent, even if the surveyor uses the best total station at hand, but it might not be
necessary for the establishment for a geometric network to complete geometric levelling
between the points. Therefore, the requirements for a certain geometric network has to be
kept in mind, when most decisions are being made.
Apart from the precision of the observations, it was determined that adding extra observations
improves the precision and reliability of a geometric network the most. When
it comes to the other parameters corresponding to the geometry of the system, having the
setups located closer to the center of the system gave better results, since the lengths of
sight were shorter and more equal to each other so the observations are weighted more
even. Using free setups instead of setups in known points has only shown slight improvements
in comparison to the other parameters.
Applying the guidelines on a real case where a superior geometric network had to be
established, unveiled some shortcomings in the guidelines and computations. It occurred
due to the dimension of the area that was about three times greater than in the network
designs have been shaped after. It would not be realistic to maintain that all control points
could be measured from the same setups. Hence a new parameter arose; examine how the
common control points that connect the setups to each other can be distributed the best.
The first network design assigned to solving the case turned out to be inadequate. Adding
an extra setup at the right place solved the defects of the first network design, and it improved
the magnitude of the precision and reliability of the geometric network a lot compared
to the small change. This seems to show that the new parameter that came about has
great importance and further studies are needed to clarify its effect on the precision and
reliability of a geometric network.