(Campus Network 2001 on MGA94/AHD71)
Vertical Control by EDM-Heighting
The computations for the first UNSW CAMPUS NETWORK were completed in 1980. In 1994, the network was revised by final year students (CAMPUS NETWORK 1994 by R. Da-Col and A. Quade). Because of the major construction work on campus since then, a large number of marks has been lost and a number of new marks has been placed. (It seems that more marks were lost between 1994 and 2001 than between 1980 and 1994). In addition, the geodetic datum of Australia and the map projection in New South Wales changed on the first of January 2000. The combination of these factors gave reason for the CAMPUS NETWORK to be re-measured.
The requirements were to design a new height network for the main Campus of UNSW. The design precision was for +5 mm (standard deviation) in a free net datum. I designed the field measuring procedures as far as the vertical control was concerned. The Australian Height Datum 1971 (AHD71) elevations of the survey marks in the new network were established by a least squares adjustment. The final elevations provide a best fit to the official AHD71 elevations of the neighbouring Permanent Marks (PM's) and State Survey Marks (SSM's) that are part of the network.
The design of the network was based on an investigation of the existing marks. New marks were placed where required. The network was designed to follow the natural layout of the campus, following corridors of visibility and aimed at long traverse lines with interconnecting parallel lines. The final network incorporates 56 stations with 25 of those being new. This shows the high 'mortality' rate of the old marks around the campus due to construction. 31 lightning conductor are included in the network. However, due to time constraints, the elevations of these marks will be calculated at a later date.
The diagram below (courtesy of J. M. Rüeger) shows the height traversing network (with some later additions). The northern boundary of the UNSW Kensington Campus is about 1060 m long. The solid squares represent State Survey Marks and Permanent Marks and the solid circles (UNSW) survey marks that were either existing or placed as part of this project. The hollow circles are elevated targets (mainly lightning conductors) that have been determined by trigonometric heighting.
Before the actual field work could begin a number of calibrations and tests were performed on the instruments to be used. These were:
Problems were encountered during the test of the first instrument for the precision of the zenith angle measurements after DIN18723. It was noted that the vertical circle index correction increases with the zenith angle. As a result, another instrument was selected (WILD TC1610 S/N 365555). The results from this second instrument were much more stable. The precision of a zenith angle measured in two faces (after DIN18723) was obtained as +/- 1.1 seconds of arc. Due to time constraints, the EDM calibration of this instrument could not be repeated; older results were used for the instrument correction.
Because the instrument correction of an electronic tacheometer is determined for a particular instrument-reflector combination, I determined the corrections for the other reflectors (relative reflector constants). The relative reflector constants of the WILD GPR-1 reflectors varied between 0.0 mm and -0.85 mm. It was decided not to apply these corrections as they would of made little impact on the height network.
The technique for the calibration of the height scale on the centring rods of the KERN tripods (to the trunnion axis of the WILD TC1610 electronic tacheometer) was devised by myself with the help of the supervisor, and incorporated some aspects of reciprocal EDM-Heighting techniques. WILD GDF21K tribrachs were used. These fit onto the KERN tripods and ensure an almost constant height of the instrument or reflector above the tripod head.
The observations in the network with the WILD TC1610 electronic tacheometer were recorded digitally on WILD GRM10 REC modules to provide quicker access to data processing and to ensure that no data was omitted. For quality control, procedures were developed for the observation routines and a field form designed for instrument and reflector set-ups. These forms proved extremely useful during the processing stages as they helped in finding the source of errors and made it possible to fix most of them.
The observations were performed by Mr. G. Collyer and M. London over a period of fifteen days, with an extra four days needed for re-observations due to bad results. Mr. I. Chami assisted on some occasions. Each line was observed two to three times in two faces. Zenith angles and slope distances were recorded. All lines between traverse stations were measured both ways. The heights of instrument and reflector were obtained from readings on the two scales of the centring rods of the Kern tripods, with the calibration values added. The centimetre scale was read to the nearest millimetre.
The observations were delayed, and even halted at times, due to car and pedestrian traffic being in the way of the line of sight. The frustration experienced was endless, when a car would park in the way when one was not looking, but had everything set up. Often a day's work was cut short due to this.
The weather also played a role in delaying the observations. On 1 August 2001, gale force winds knocked over two tripods. As a result, two KERN tripods and one tribrach were damaged.
After the downloading of the data from the recording modules, the following steps were involved in getting the final product:
The least squares adjustment was done for, both, the separate forward and backward height differences and the mean of the forward and backward height differences. In the adjustment with the separate forward and backward height differences there were 609 height differences and 61 unknown elevations. For the mean forward and backward adjustment there were 296 height differences and 61 unknown elevations.
The block shift was computed using the published heights for marks PM21478, PM21477 and SSM47120. The differences between the heights published and those computed after the block shift adjustment were -1.8 mm, +5.3 mm and -3.5 mm, respectively. The following table shows the precision of the computed elevations and their uncertainties at 95% confidence level, both for the freenet solution.
From the results listed above we can see that the design precision of the network was met as all listed precision values are less than the prescribed +5 mm. The elevations of the lightning conductors and trigonometrical stations will be calculated at a later date. Due to time constraints, the precision of the network was not tested in the design stage. If this testing had been done, the network could have been improved by identifying areas that needed strengthening.
Overall the project was a success and, hopefully, will benefit future students.