Introduction & methodology
The task of this exercise was to carry out a looped levelling survey using the rise and fall booking method around the CSM building. Our group consisted of Christopher Crabb, Alex Levantis and Robin Watson. The weather conditions were overcast with a fairly pervasive mist throughout.
We chose the corner of a drain cover closest to the campus buildings to act as our first Backsight(BS) and last Foresight (FS), the level we determined for this point witha handheld GPS was 124m AOD. From here we proceeded to take one Intermediate Sight (IS) for each setup of the level, before taking a FS onto our first Change Point (CP) where the staff remained whilst the instrument was moved to its next setup position. The difference between each sight and the previous IS or BS was then worked out and applied to the RL level of the last point as either a rise or fall.This was repeated for 11 different set-ups.
As can be seen on Fig 1., where practicable we kept our sights to the magnitude of 20-25m to reduce collimation error, with an IS on a memorable location between these two CPs. Each of us took turns to read the level, book the results and hold the staff.
Results
Fig 1 showing the location of our survey stations, change points and intermediate sights.
Fig 2.Booking sheet with arithmetic checks.
On completing our survey we originally calculated our final error to be 0.051m, on examining the booking sheet, it was noticed that one fall (point 3) had been treated as a rise and this was responsible for 0.047m of theerror. The booking sheet shown at Fig 2.is the revised version. Our final error was then -0.004m; well within the allowable misclosure where m=5mm
m√11=0.0165…
Likely sources oferror include wind buffeting the staff,especially at its fullest extension.The mist certainly made sighting onto the staff difficult at certain points. Changing the position of the staff on the CP is another possible source of error.
Use of sights of unequal lengthsalong steep slopes may have exacerbated any collimation error present in the instrument,other calibration issues may also have affected our results.
Reading errors, or inconsistencies in the method of estimating to the nearest 1mm between the three people taking those readings may also contribute; booking errors would certainly have caused an error outside the allowable misclosure (0.051m) had this not been identified during our arithmetic checks (Uren & Price, 2010: 43, 54-55).
Contouring exercise
The next part of the assignment involved the recording of a series of levels in a rough grid across a slop. The position of these spot heights was then to be worked out using the Stadia Tachometry method. For each point a bearing was recorded, along with the upper and lower stadia readings, and a reading from the central crosshair. The difference between these points multiplied by 100 gave a horizontal distance to the point being levelled.
A=100d,
where d = the difference between the upper and lower stadia. These coordinates and their heights are shown in Fig 3.
Our coordinates were then calculated with the sin/cosine rule and plotted in fig 3.
ΔN =cosθ°A, ΔE = sinθ°A
where θ=the bearing of the point and A=the distance to the point calculated by the stadia tachometry method above;
Results
Fig 3.Location of spot heights with Point 19’s adjusted position in red.
Fig 4. Stadia Tachometry booking sheet with outlying result highlighted. Note conversion to radians for use in Excel.
An obvious outlier in the results appears to be due to a reading/booking error, as the IS does not lie in the middle of the upper and lower stadia readings as it should do. The IS for point 19 should be equidistant between the upper and lower stadia; 2.9425 rather than 2.938 as booked.
Lower Stadia = 2.892 IS = 2.938 Upper Stadia = 2.993
Lower Stadia – IS = -0.044, and IS – Upper Stadia = -0.055
We would expect these two differences to be equal. If we adjust the upper stadia by the difference between these two differences and recalculate the coordinates, then we can now mark the point at 9m from the level, 8.998629256 -0.157071658.
2.993 – 0.011 = 2.982
so 100(2.982 – 2.892) = 9m
so cos91x9 = -0.157071658 and sin91x9 = 8.998629256
As shown in Fig 3.this point is far closer to the actual progression of points across the slope. Of course the exact booking error cannot be known, but it seems reasonable that an extra 0.010m was added, probably by reading the wrong 0.010m block on the E type staff used.
Alternative levelling methods
An electronic laser level would have eliminated our booking error of 0.047m and any other potential errors from booking or reading the staff, laser levels can also be used accurately over greater distances (up to 100m). Poor light or weather would affect the ability of the electronic level to sight on the barcoded staff but these conditions affect all methods of levelling. Perhaps the only significant disadvantage of the electronic level is the increase in price compared to the advantages it affords; competent surveyors with a systematic approach to reading and booking can complete a survey in a comparable period of time with an automatic level (Uren and Price, 2010: 41).
Levels can be calculated trigonometrically using a theodolite. Trigonometric levelling can be prone to error as gravity is not used as a control element; sighting and reading of vertical angles must be absolutely precise or errors in calculation become exacerbated as the slope distance increases (Uren and Price 2010, p175). Another limitation is that the horizontal distance between the survey station and the point being levelled must be known or calculated, again increasing the possibility for errors due to inaccurate sighting. A survey of this kind involves far more calculation than direct levelling with an automatic level, with the corresponding time required and increased chances of booking error. However, the trigonometric method is useful (and indeed may be the only possible method) in circumstances where the points to be levelled are inaccessible, such as mountaintops, or overheard power cables.
Levelling with an EDM or total station is entails using the same calculations involved with trigonometric levelling, only carried out automatically by the instrument itself. The EDM allows this method to be used where the distance to a point is not known by accurately measuring the slope distance. The same errors resulting from inaccurate sighting still remain however, the instrument must be sighted exactly in the centre of the prism, and it is advised that numerous readings be taken and any anomalies discarded, thus adding to the length of the survey. Total stations are far more sensitive than automatic levels and atmospheric conditions require that the instrument be given a chance to acclimatise to the ambient temperature, and low and high levels of light or ground shimmer can require mathematical adjustments to be made to the readings . Levelling with a total station is an effective way of obtaining heights for a topographical or walkover survey (Uren Price 2010: 174).
Alternative methods for a contouring survey could involve carrying out a detail survey with other methods (i.e., total station, GPS or even chain survey), then simply adding spot heights with an automatic level and the height of collimation method. Indeed this is the most common method within the archaeological profession and one I have some experience of.
As can be seen there are advantages to using survey equipment other than an automatic dumpy level, but these all have the potential for errors and a greater complexity than the direct levelling method. An automatic level is economical, simple and robust where other methods may be complicated, delicate or expensive.
Bibliography
Uren, J and Price, B. 2010, Surveying for Engineers, Fifth Edition, Palgrave Macmillan, Basingstoke.