PRECISE LEVELLING

Precise leveling may be required in certain instances in construction such as in deformation

monitoring, the provision of precise height control for large engineering projects such as long-span
bridges, dams and hydroelectric schemes and in mining subsidence measurements. For example,
a dam that has been in place for many years is unlikely to be moving. However, should the dam
fail the results would be catastrophic for those on the downstream side. Being under the pressure
of water when full, the dam may be liable to distortion. The behaviour of the dam must therefore
be monitored. One way of monitoring any vertical movement along the dam is by leveling. Since
early warning of small movement is required, and since conclusions about movement must be
made with statistical confidence, the leveling must be very precise.
There is more to precise leveling than precise levels. High quality equipment is very important,
but so is the method by which it is used. Indeed the two components of precise levelling are
precise equipment and precise procedures. Precise leveling uses the same principles as ordinary
leveling but with:
a. Higher quality instruments and more accurate staves
b. More rigorous observing techniques
c. Restricted climatic and environmental conditions
d. Refined booking and reduction
e. Least squares adjustment for a leveling net

Functions of precise levelling

i. To provide the level control frame work the country eg provide bench marks for
future use
ii. Provide information for geodetic research including the tectonic movement and
mean sea level studies
iii. Provide accurate data for major engineering works
iv. To enable the surveying distances to be reduced to the spheroid (ellipsoid) of
reference
v. To provide heights for gravity stations

Precise invar staff

The precise leveling staff has its graduations precisely marked (and checked by laser
interferometer) on invar strips, which are attached to wooden or aluminum frames. The strip is
rigidly fixed to the base of the staff and held in position by a spring-loaded tensioning device at
the top. This arrangement provides support for the invar strip without restraining it in any way.
Usually there are two scales on each staff, offset from each other by a fixed amount f.igure 1
The staff is placed upon a change plate at intermediate stations. A conventional leveling change
plate is small and light and is designed to give a firm platform for the staff on soft ground. Precise
leveling should only ever take place on firm ground and the precise levelling change plate is
designed to be unmoving on a hard surface. It is therefore heavy. The feet are rounded so that they
do not slowly sink or heave when placed on tarmac. The top is smooth, round and polished. The
change plate in Figure 3.34 is made from a solid piece of steel and weighs about 10 kilograms.

Prepared by Shabola
 Figure 1: Segment of a precise leveling staff

Figure2: Precise leveling change plate

Prepared by Shabola
For the most precise work, two staffs are used; in which case they should be carefully matched in
every detail. A circular bubble built into the staff is essential to ensure verticality during
observation. The staff should be supported by means of steadying poles or handles.
(1) The staff should have its circular bubble tested at frequent intervals using a plumb-bob.
(2) Warping of the staff can be detected by stretching a fine wire from end to end.
(3) Graduation and zero error can be counteracted by regular calibration.
(4) For the highest accuracy a field thermometer should measure the temperature of the strip in
order to apply scale corrections.

Instruments

The instruments used should be precise levels of the highest accuracy. They should provide high-
quality resolution with high magnification (×40) and be capable of being adjusted to remove any
significant collimation. This may be achieved with a highly sensitive tubular bubble with a large
radius of curvature that gives a greater horizontal bubble movement per angle of tilt. In the case of
the automatic level a highly refined compensator would be necessary. In either case a parallel plate
micrometer, fitted in front of the object lens, would be used to obtain submillimetre resolution on
the staff. The instrument’s cross-hairs may be as shown in Figure 3a. The distance that the staff is
away from the instrument will affect which side of the cross-hairs is to be used. If the staff is far
away use the normal horizontal hair, right-hand side of diagram. If the staff is close, the mark on
the staff will appear too large to be bisected accurately. By comparing the two white wedges
formed by the sloping cross-hairs and the mark on the staff, see the arrows in Figure 3(b) where
the mark on the staff is not correctly aligned and Figure 3. (c) Where it is, a more precise setting
of the micrometer can be made.

Figure3: Precise level cross-hairs

Parallel plate micrometer

For precise leveling, the estimation of 1 mm is not sufficiently accurate .A parallel plate glass
micrometer in front of the object lens enables readings to be made direct to 0.1 mm, and
estimated to 0.01 mm. The parallel plate micrometer works by refracting the image of a staff
graduation to make it coincident with the cross-hair. There is, therefore, no estimation of the
position of the cross-hair with respect to the graduation. The principle of the attachment is seen
3

Prepared by Shabola
from Figure 4 Had the parallel plate been vertical the line of sight would have passed through
without deviation and the reading would have been 1.026 m, the final figure being estimated.
However, by manipulating the micrometer the parallel plate is tilted until the line of sight is
displaced to the nearest division marked on the staff, which is 1.02 m. The rotation of the
micrometer drum is proportional to the displacement of the image of the staff. The amount of
displacement s is measured on the micrometer and added to the exact reading to give 1.02647 m,
only the last decimal place is estimated. It can be seen from Figure 5 that the plate could have
been moved in the opposite direction, displacing the line of sight up. Since the parallel plate
micrometer run is normally equal to the gap between two successive divisions on the staff it will
not be possible to gain coincidence on more than one division.
The displacement is related to the rotation of the parallel plate as follows. In Figure 6 the plate
pivots about A. The displacement is BC and the rotation is equal to the angle of incidence i. The
thickness of the plate is t and the ray of light from the staff is refracted by an angle r. μ is the
refractive index of the glass of the plate.
Displacement = BC = AB sin (i –r)

Figure 4: Parallel plate micrometer

Figure 5: Parallel plate

displacement

Prepared by Shabola
Since t and μ are fixed properties of the plate then displacement is directly proportional to
rotation. Parallel plate micrometers are also manufactured for use with 5 mm graduations.

Field procedure

At the beginning and end of each leveling run a stable and precise benchmark is required.
Intermediate points are not observed. To avoid accidental damage or vandalism wall mounted
benchmarks can be removed from the wall leaving the barrel, which has been fixed with epoxy
resin, capped for protection (Figure 7. The size of the leveling team depends upon the observing
conditions and the equipment available. In ordinary leveling an observer and staff holder are
required. In precise leveling there are two staves and therefore two staff holders are required. If a
programmed data logger is available then the observer can also do the booking. If the
observations are to be recorded on paper a booker should also be employed. The booker’s task,
other than booking, is to do a series of quality control checks at the end of each set of
observations, before moving to the next leveling bay. Finally, in sunny weather, an umbrella
holder is required because it is necessary to shield the instrument and tripod from the heating
effects of the sun’s rays.

Prepared by Shabola
(Figure 7. Wall mounted benchmark

Just as with ordinary leveling, a two-peg test is required to confirm that the instrumental
collimation is acceptable. Precise leveling procedures are designed to minimize the effect of
collimation, but even so, only a well-adjusted instrument should be used.
Precise level lines should follow communication routes where possible because they generally
avoid steep gradients; they are accessible and have hard surfaces. However, there may be
vibration caused by traffic, especially if using an automatic level. The following procedures
should be adhered to when carrying out precise leveling

Precise leveling can be manpower intensive, and therefore expensive to undertake. It is

important to carry out a full reconnaissance of the proposed levelling route prior to observations
being taken to ensure that the best possible route has been chosen.
1. End and intermediate benchmarks should be constructed well before levelling starts to
prevent settling during leveling operations.
2. Steep slopes are to be avoided because of the unequal and uncertain refraction effects on
the tops and bottoms of staves.
3. Long lines should be split into workable sections, usually each section will not be more
than about 3 km, because that is about as much as a team can do in one day. There must
be a benchmark at each end of the line to open and close on. The length of each line will
depend upon terrain, transport, accommodation and other logistical considerations.
4. Each section is to be treated as a separate line of leveling and is checked by forward and
backward leveling. This will isolate errors and reduce the amount of re-leveling required
in the case of an unacceptable misclosure
5. On each section, if the forward leveling takes place in the morning of day 1, then the
backward levelling should take place in the afternoon or evening of day 2. This will
ensure that increasing refraction on one part of the line in one direction will be replaced
by decreasing refraction when working in the other direction. This will help to
compensate for errors due to changing refraction effects.
6. On bright or sunny days an observing umbrella should be held over the instrument and
tripod to avoid differential heating of the level and of the tripod legs.
7. Take the greatest care with the base plate of the staff. Keep it clean. Place it carefully
onto the change plate and do not drop the staff. This will avoid any change in zero error
of the staff. When the staff is not being used, it should be rested upon the staff-man’s
clean boot
8. The distances of foresight and backsight must be as nearly equal as possible so as to limit
the effect of the Earth’s curvature, refraction and bad instrumental collimation. This will
also avoid the need to re-focus the level between sightings.
9. Take care when leveling along roads or railways. Stop leveling when traffic or vibrations
are heavy. When the staff is not being used, it should be rested upon the staff-man’s clean
boot. Vibration may damage the staff base plate and so change its zero error.
10. On tarmac and soft ground the instrument or staff may rise after it has been set up. This
may be apparent to the observer but not by the staff person.
11. In gusty or windy conditions stop levelling because there will be uncertainty in the
readings. In variable weather conditions consider levelling at night.

Prepared by Shabola
 12. The bottom 0.5 m of the staff should not to be used because of unknown and variable
refraction effects near the ground.
13. If a precise automatic level is to be used, it should be lightly tapped and rotated before
each reading to ensure that the compensator is freely operative. This will reduce errors by
ensuring that the compensator always comes from the same direction. Some automatic
levels have a press button for this purpose.
14. The rounded centre on the change plate should be kept polished and smooth to ensure
that the same staff position is taken up each time it is used.
15. The change plate must be firmly placed and not knocked or kicked between foresight and
backsight readings. Remember there is no check on the movement of a change plate
between these observations. The staff holder should stand clear between observations.
16. The observation to the back staff must be followed immediately by an observation to the
forward staff, both on one scale. This is to ensure that refraction remains constant during
the forward and back observations of one bay. Then, an observation to the forward staff
is followed immediately by an observation to the back staff on the other scale. This
procedure helps to compensate for unknown changes in refraction, by balancing the
errors. Using two double scale rods the sequence of observation would be:
a) BS left-hand scale on staff A
b) FS left-hand scale on staff B
c) FS right-hand scale on staff B
d) BS right-hand scale on staff A
Then (1)−(2) = _H1 and (4)−(3) = _H2; if these differences agree within the tolerances
specified, the mean is accepted. Staff A is now leapfrogged to the next position and the
above procedure repeated starting with staff A again (Figure
e) If the back staff is observed first at one set-up, then the forward staff is observed first at
the next set-up. This ensures that changing refraction will affect each successive bay in
an equal and Fig.

Fig
Staff

leapfrog

opposite manner. The order of observations this time will be:

a) FS left-hand scale on staff A

b) BS left-hand scale on staff B
c) BS right-hand scale on staff B
d) FS right-hand scale on staff A

Prepared by Shabola
Note that in each case the first observation of a bay is to the same staff, which is alternately the
back and then the forward staff

19. The same staff that was used for the opening backsight must also be used for the closing
foresight. This will eliminate the effect of different zero errors on the two staves. This means that
there must always be an even number of set-ups on any line.
20. Levelling should always be carried out in both directions, forward and back. If, on the
forward levelling, the A staff was used to open and close the line, then the B staff should be used
to open and close the line on the backward levelling. This will equalize the number of readings
on each staff.
21. Lines of sight should not exceed 50 m, especially in haze, or on sloping ground. This will
minimize the effects of refraction, curvature of the Earth and difficulty of reading the staff. A
good average length of sight is 35 m. (22) Use the procedure already outlined for levelling the
circular bubble on automatic levels. This will happen as a matter of course if the telescope is
aimed at staff A each time when centring the circular bubble.

Booking and computing

Figure shows a sample of precise leveling observations. Note the order in which the
observations are made and that it agrees with paragraphs (17) and (18) a bove. Once the
observations are complete

Precise leveling observations

Prepared by Shabola
 9

Prepared by Shabola