Establishing a national network means an enormous amount of measurements and very high costs. Therefore it is crucial to use the most effective instrumentation and technique available. The Swedish LMV turned to Motorized Leveling.

It was clear in Sweden around 1970 that the national height networks did not meet the needs and requirements of users regarding quality or density, especially for the first order height network. To remedy this situation, the National Land Survey (LMV) started several investigations, studies and tests to develop an appropriate leveling technique. Up till then, the LMV was, like most other countries, using the leveling by foot (FL) technique for precise leveling, but it was clear that this technique was too slow and too costly. Therefore, other possible solutions were tested to increase the leveling speed: the use of bicycles or cars, ways to minimize the number of readings and the use of electronic field books.

Establishing a national network means an enormous amount of measurements and very high costs. Therefore it is crucial to use the most effective instrumentation and technique available or develop a technique of your own. To answer the problems associated with FL, the LMV turned to Motorized Leveling (ML).

ML was developed in the 1970s. The breakthrough for ML came when Carl Zeiss Jena (Oberkochen, Germany) introduced the leveling instrument Ni002 in 1972. This instrument was the first self-horizontaling pendulum instrument giving a quasi-absolute horizon (through two symmetrical pendulum positions). It also had a rotating eyepiece allowing measurements 360 degrees around and had the crosshair placed in the objective (no parallax error). Influenced by the promising results of the instrument, LMV decided to build a Swedish ML team to test the technique under field production conditions, verify and evaluate the results, and make improvements. J-M Becker introduced the technique to Sweden in 1973. After several years of test measurements, it was decided to use this technique for the third precise leveling of Sweden from its beginning in 1979 until its finish in approximately 2001.

Developing the Technique

ML uses three cars (one instrument car and two rod cars) and four people (two in the instrument car and one in each rod car) for all leveling operations. Communication between the observer and the driver in the instrument car, who also logs all the data, is done with microphones. All work is performed directly from the vehicles. The operators do not leave their cars except when connecting to benchmarks. This makes it possible to work during all weather conditions, increasing the production rates and the quality of the results.

From 1973 to 1985, many technical improvements were made. We developed a special tripod with long adjustable legs and special foot plates to reduce settlement effects and the influence from vibrations, wind and temperature variations. We constructed, together with Micronics (Sweden), the first field computer in 1981, which allowed us to store and perform automated field controls of all observation data. This was the first step in the digital production line. An automated rod comparator with laser interferometer was constructed in 1981. The comparator makes it possible to calibrate and calculate corrections for each graduation on the invar rod.

We changed from 3 to 3.5 m invar rods in 1982. This increased the production rate 15 percent by allowing us to increase the mean average observation sight length from 33 to 37 m. We also made it impossible to do readings on the lower 0.5 m in order to reduce refraction errors.

The rods are equipped with three bull's eye levels for permanent plumbing control. They also have electronic temperature sensors at 0.5 and 3 m over the ground to apply correction for temperature influence on the rod scale.

In 1985, two new motorized height determination techniques were developed: the Motorized Trigonometric Leveling technique (MTL), only for heights, and the Motorized 3-D technique (MXYZ). The instruments are centrally located in the modified ML car. Both use modern electronic total stations instead of the classic level Ni002. The achieved performances were astonishing both in quality and quantity. Precise leveling accuracy was performed. The purpose was to use the techniques in mountainous areas where classic ML has limited sight lengths. We also developed another height determination technique in 1988 called Trigonometric Leveling using motorcycles (TL). Four-wheel motorcycles were used for the transport of equipment and personnel mostly at swampy and hilly areas on the border to Norway. The measurements were made in the classic way with setup outside the vehicles.

Field operations have also been subject to some improvements over the years. For example, we are able to use nonequal sight lengths for back and forward observations at each setup. Differences of up to 10 percent at each setup are accepted. We do only four readings (two for each pendulum position I and II and only one height difference on each scale of the invar rod graduations) at each setup. We perform automatic registration and control check in accordance with the tolerance specifications at each setup. We want to have as many independent measurements as possible, which means we change observers between the forward and backward measurement of a section. We never carry out the forward and backward measurements of a section immediately after each other. We perform field operations under all weather conditions. To be sure that the equipment is not defective during a field season, instrument checks are made once a week. The invar rods are calibrated twice per year.

The Third Precise Leveling of Sweden-
The Net Configuration

To cover the whole territory of Sweden, we estimated a need for 50,000 benchmarks. It was suggested that all leveling should be performed in one unified network and not in a network of two different orders (i.e., a primary network with subsequent densification). The network is leveled in closed loops with a circumference of about 80 to 120 km along roads. The distance between two benchmarks is about 1 km, and the establishment of a benchmark is done with great care to secure long-term stability and good accessibility. The advantages are that the height network will be easier and less expensive to realize; the first results can be given to users in a shorter time (about three years); and a more homogeneous net of high quality is established.

Comparing the Results

The average hourly progression for precise leveling by ML is around 2.2 km with average sight lengths of about 35 m (maximum allowed 50 m). The total time used at each setup, including the moving time, varies between 1.6 and 2.4 minutes depending on the sight lengths. The average daily production is 12 km single run, including 7 percent releveling-about 11 km. Statistics show the effective measuring time by ML is about 5.5 hours per day. These production rates are the result of an optimal work distribution between all actors in the ML team to record all interesting data at each setup. The same statistics concerning other techniques show FL at about 3.5 km per single run (with classic levels like the Wild N3, (Leica Geosystems Ltd., Norcross, Ga.)) and 5 km using digital levels; for CL about 6 km; and for MTL around 10 km. Table 1 on page 42 shows production rates during the '90s in the third precise leveling. We have tested digital levels. All information and measurements have to be recorded and stored directly in the instrument by the observer. It has been shown that more time is needed at each setup and that the optimal work distribution within the team is disturbed. Our tests give about 15 percent decreased production rates, which is important for production costs.

The most significant advantage ML has over other techniques is its ability to perform well during the whole field season and the whole working day-independent of the weather and surrounding conditions. These results have been confirmed in other countries with very different working conditions like Zambia, Malaysia, France and the United States. From an economic point of view, the possibility to perform non-stop measurements under normal working conditions in Sweden has reduced the production price by more than 50 percent compared to FL or CL.

Working Conditions and Field Use

Performing classic leveling has always been considered a hard, monotonous task. Few surveyors are attracted or interested in working an entire field season with these techniques. The introduction of motorized techniques has improved field working conditions since nearly all operations are performed from the vehicles. This has also made it possible for more women to participate in field activities. Nearly half of the personnel today are women. This was never the case when Sweden used FL and CL techniques. The field working hours for ML are the same as office working hours.

ML also means increased safety for workers. Our experience shows that safety has strongly improved since operators now have better protection through their position inside the cars and because of the numerous warning signs on and around the vehicles.

We have 25 years of experience with motorized leveling techniques from Sweden and other countries. We believe that digital levels are the preferred type of leveling instrument for use in foot (FL) or bicycle (CL) leveling. Their use for high, precise leveling is subject to limitations as sight lengths (<35 m) and restrictions (influence of temperature variations and light) reduce their performance (quality and production).

Digital levels are not competitive for motorized leveling compared to the Zeiss Jena Ni002 for several reasons:

  • Decreased production rates of about 15 percent due to non-optimal distribution of work within the leveling team.
  • Too sensitive to nonequal sight lengths and temperature variations because of no quasi-horizon by double pendulum positions.
  • Their ocular is not rotating around the horizon, which complicates the observations.
  • Some digital levels need to see a big part of the invar staff, which limits the sight lengths.

More than 100,000 km of motorized leveling shows that the ML technique is still outstanding after 25 years and the most efficient technique for high, precise, large-scale leveling work. The results can be summarized as follows:

  • Accuracy in terms of mean standard error less than 1 mm/vkm even under unfavorable conditions.
  • Production under all weather conditions during normal daily working hours and throughout the normal field season everywhere.
  • Improved working environment makes it physically easier for women.
  • Increased production rates of more than 50 percent compared to FL and more than 15 percent to MTL.
  • Production cost per leveled km reduced by about 50 percent compared to other techniques. o Increased measuring capacity by 40 percent per year in Sweden.
  • Increased security for the operating team.

New technologies like GPS have not yet shown that they are fully able to take over the role of the more "classic" height determination techniques. Also in the future, traditional height determination techniques will be used and compete on the leveling market at least for high-precision leveling networks.