5.2.5 GPS Survey Planning

LOGISTICAL CONSIDERATIONS

The logistical considerations of a GPS project increase enormously with:

• an increase in the number of stations to be surveyed,
• an increase in the number of receivers deployed,
• an increase in the number of common stations occupied between sessions,
• an increase in the number of fixed stations to be occupied, and
• an increase in the number of sessions per day.

They are also complicated by such factors as:

• Terrain and the nature of the transport links? 4WD vehicles? helicopter?
• Requirements for contingency plans, in the event of mishaps, etc.
• Requirement for quality control checking and partial processing in the field office each day, hence the need for data from all receivers to be transferred to the field office.
• The amount of reconnaissance already carried out.
• Economic pressures, for example, to complete the project in the minimum time.
• Available accommodation to minimise travel time, etc., is it convenient to the project?
• Instrumental factors: do receivers need special equipment for data download?

The basic principle is to KEEP IT SIMPLE, hence it is important to keep baseline short (and hence minimise travel times), use say one hour observation sessions, and employ a suitable receiver deployment strategy.

There are a number of possible receiver deployment schemes that can be used. Each has its advantages and disadvantages with respect to logistical considerations such as cost, time, manpower, etc. Generally, some station or stations are occupied for all or some of the campaign, and the other receivers move between sessions. A combination of the "base station" (or "radiation") mode, and the "leap-frog" (or "traverse") mode, as shown in the Figure below, is usually used.

Figure 1. Receiver deployment modes.

Logistical Design Principles

With experience, the organisational design of a session-by-session observation schedule is a straightforward matter based on a few simple "rules-of-thumb":

• Receiver deployment schemes based on:
  • leap-frog mode,
  • base station operation,
  • or some combination of these.
• The "shape" of a GPS session, or final network (multi-session), plays little part in the final accuracy, unlike the situation with conventional surveys. Length of lines in a session do influence the final accuracy, both through the "ppm" relationship, as well as a change in the "ppm" value at distances above which ambiguity resolution can be carried out.
• Adjacent stations should be connected directly -- that is, keep baselines short.
• Stations on the perimeter of project area should be directly connected.
• Connect all stations in network through "pivot" or "link" sites common to two or more sessions, so that a minimally constrained GPS network is established. The percentage of multiple occupancies is directly related to the accuracy classification of the survey, and is usually defined in the prescribed "standards & specifications" (section 10.2.8).
• Maximise geometrical redundancy through having multiple occupancy of sites (though only within reason). There are two types of redundancy:
  • Reoccupation of several sites at the same time to define repeat baselines (to allow for checks on the internal consistency of GPS surveys).
  • Form loops of stations occupied during different sessions, permitting checks on loop closure statistics.
• Where possible sites should be revisited by different field parties, hence ensuring independent setups, in order to minimise the chance of misidentifying the station mark.
• Avoid "no check" baselines (Figure 2 below), where one station of the baseline has only been visited once.

Figure 2. Factoring in redundancy to prevent "no check" baselines.

There is a danger in the overuse of the "base station" or "radiation" mode of survey (be it with GPS, or conventional EDM/theodolite) as, unless all the sites are visited more than once, all radiated baselines are in reality "no check" baselines. It is tempting to consider a two-base station configuration as having a natural builtin redundancy (there are two baselines coming into each site, one from each of the base stations -- one being the primary base station, and the other the secondary base station). However, although the site is no longer fixed by a single "no check" baseline radiation, the additional baseline from the secondary base station is not independent! Hence, observation scheduling must take into account not only the requirement for multiple occupations to ensure redundancy, but also the degree to which these additional baselines must be truly independent.

The reader is referred to SNAY (1986) and UNGUENDOLI (1990) for a discussion of organisational design as it applies to surveying a given network with a certain number of GPS receivers. However, it is unlikely that such "academic" planning exercises are ultimately useful given the extraordinary variety of GPS networks that surveyors may be called upon to observe. It is preferable that the surveyor develop a "sixth sense" when it comes to planning surveys, such that due consideration is always given to ensuring there are builtin redundancy and quality control measures, rather than incorporating them into a design as an afterthought.

The issue of GPS network design is discussed again in section 5.5.6, in relation to the modern GPS surveying techniques such as "rapid static" and "stop & go".

Factors influencing GPS accuracy: Baseline results The "shape" of the GPS network is irrelevant as far as the baseline solution quality is concerned - orientation of baseline does not greatly influence its quality. The following factors do influence baseline quality: Length of baseline. Single or dual-frequency instrumentation. Length of observation session. Number of observed satellites. Observable being processed. Processing software. Quality of ancillary information (orbits, fixed sites, etc.). Typical horizontal baseline accuracy is expressed as: where L baseline length in km, a = 0.2 - 1 cm (centring error), b = 1 - 5 ppm Height is the weakest component (2-3x worse)

Factors influencing GPS accuracy: Overall network quality Baseline quality. Homogeneity of GPS survey (single and dual-frequency results, etc.). Number and distribution of independent baselines. Other geodetic observations (distance, angle, etc.). The overall network accuracy is a function of the number and distribution of repeat baselines and redundant station occupations.

Back To Chapter 5 Contents / Next Topic / Previous Topic

© Chris Rizos, SNAP-UNSW, 1999