What Causes GNSS Multipath and Survey Error?

A GNSS rover can show a healthy correction link, plenty of satellites and a fixed RTK solution, yet still produce a position that does not agree with a check point. In many of these cases, what causes GNSS multipath is not a problem in the correction service or receiver settings. It is the site itself: a satellite signal has reached the antenna by more than one route.

For surveyors and site engineers working close to buildings, plant, retaining walls or steelwork, multipath is one of the most persistent sources of avoidable GNSS error. Understanding where it comes from makes it far easier to decide when a GNSS observation is suitable, when to change position, and when a total station is the better tool.

What causes GNSS multipath?

GNSS multipath occurs when an antenna receives both the direct signal from a satellite and one or more delayed copies reflected from nearby surfaces. The receiver is designed to calculate range from the time taken for the satellite signal to arrive. A reflected signal has travelled further, so it arrives slightly later than the direct signal.

The receiver may not be able to separate those signals completely. Instead, the reflected energy can distort the measured code or carrier phase, creating a biased range measurement and, ultimately, an error in the calculated position.

The effect is not constant. It changes as satellites move through the sky, as the rover changes position, and as the surrounding environment changes. That is why a point may appear acceptable at one moment but drift or fail a repeat observation a few minutes later.

Reflective surfaces are the main culprit

Hard, smooth and electrically conductive surfaces are particularly effective at reflecting GNSS signals. On construction and infrastructure sites, common examples include glazed façades, metal cladding, steel frames, containers, hoarding, bridges, concrete walls, parked vehicles and excavators.

Water can also be troublesome. A canal, flooded excavation, wet roof or large puddle can create reflections, especially when satellites are low in the sky. The problem is often more noticeable close to the reflecting surface because the reflected signal can remain relatively strong.

A single nearby surface can cause an issue, but multipath becomes more likely in enclosed or built-up areas where signals can reflect several times. These are often described as urban canyons: locations between tall buildings, alongside a façade or beneath overhanging structures where the antenna has a restricted view of the sky.

Obstructions make direct signals weaker

Multipath is more damaging when the direct satellite signal is partly blocked or weakened. A receiver operating in open ground has a strong direct signal and is generally better placed to reject weaker reflected signals. Place the same antenna beside a wall or beneath tree cover, and the direct signal may be attenuated while reflections become more influential.

Trees do not act like a clean mirror in the way glass or steel does, but foliage can block, scatter and weaken signals. Dense, wet leaves are especially challenging. In woodland, the result is often a mixture of poor satellite visibility, signal attenuation and multipath, rather than one isolated cause.

Low-elevation satellites deserve particular attention. Their signals pass through more atmosphere and are more likely to encounter nearby objects before reaching the antenna. A receiver may still track them, but their measurements can be less reliable in a difficult environment.

Why multipath matters to RTK surveying

RTK corrections remove many common GNSS errors by comparing observations from a reference station with those at the rover. They are highly effective for satellite orbit, clock and atmospheric errors across an appropriate baseline. They cannot fully remove local reflections around the rover antenna.

That distinction matters on site. The reference station may have an excellent open-sky location, while the rover is beside a warehouse wall, under a crane jib or close to a lorry. The reflected signal path is unique to the rover, so it remains in the measurement after corrections are applied.

Carrier-phase RTK is very precise, but it is not immune to multipath. A fixed solution confirms that the receiver has resolved integer ambiguities consistently; it does not guarantee that every observation is free from local bias. If the result is critical, such as setting out structural steel, verifying grid lines or establishing control, an independent check is still essential.

Multipath may show up as coordinates that wander, unexpected residuals, repeat shots that disagree, prolonged ambiguity resolution or a position that looks credible but fails against known control. The last case is the most dangerous because it can encourage false confidence.

Site conditions most likely to create multipath

The highest-risk locations are predictable. They include work beside tall glazed or metal-clad buildings, narrow streets, bridge decks, beneath canopies, around gantries, within compounds lined with containers, and close to large items of plant. Working on a reflective roof or near standing water can have a similar effect.

Temporary conditions matter as much as permanent structures. A rover position that worked during an early site visit may become unreliable once steelwork, hoarding, welfare units or parked plant are in place. Similarly, a control point installed too close to a wall can become a recurring source of doubt throughout a project.

The antenna itself also needs space. Holding a rover pole against fencing, leaning it close to a vehicle or standing beside the antenna can introduce local interference and reflections. Good field technique is not a substitute for the right equipment, but it remains central to consistent results.

How to reduce GNSS multipath in the field

The most effective control is usually to move. Even a small change in position can alter the reflected signal geometry enough to improve the observation. If a point is close to a façade or other reflector, where the task permits, stand clear of it and use an offset measurement or a conventional instrument to transfer the point.

Give the antenna the clearest practical view of the sky. Avoid setting up immediately beside walls, metal fencing, vehicles and plant. Keep the pole vertical, hold the antenna above nearby obstructions where it is safe to do so, and avoid observations directly beneath overhangs.

For critical work, observe the point more than once. Reoccupy it after a short interval, ideally when satellite geometry has changed, and compare the result with independent control. Averaging can help reduce random noise in favourable conditions, but it should not be used to disguise a repeatable bias from a nearby reflector.

Receiver configuration can also help. A sensible elevation mask reduces reliance on low satellites, although setting it too high can reduce satellite availability and weaken geometry. Multi-constellation, multi-frequency receivers provide more measurements and can improve resilience, but they do not make a poor site suitable for GNSS. The trade-off is straightforward: more capable hardware improves the chance of a reliable fix, while sound observation locations remain the first line of defence.

Professional GNSS receivers use advanced signal processing and carefully designed antennas to suppress multipath. Ground planes, antenna calibration and, on specialist static applications, choke-ring designs can reduce sensitivity to reflected signals. These measures are valuable, particularly on control networks, but no antenna can remove the effect of a severely obstructed location.

Choose the instrument to suit the environment

GNSS is exceptionally productive for topographic survey, earthworks, machine control, asset capture and set-out in open or moderately obstructed areas. It can save significant time by removing the need for line of sight between stations. The limitation is environmental: GNSS needs a usable view of the sky and enough separation from strong reflectors.

Where that is not available, a robotic total station can provide more dependable local measurement. It is often the practical choice for internal courtyards, streets enclosed by tall buildings, beneath structures and detailed work around façades. On complex projects, the efficient approach is commonly a combined workflow: GNSS for open-area control and rapid capture, then total station observations where multipath risk is high.

A site assessment before work begins helps avoid costly rework. Identify likely reflective surfaces, establish control in open locations, define check points, and agree the accuracy required for each task. A general earthworks pickup and a structural set-out operation should not be judged by the same tolerance or verification process.

A practical response when results do not agree

If a GNSS position looks doubtful, do not simply keep measuring in the hope that it settles. Step away from nearby reflectors, allow the solution to reconverge if necessary, and reobserve against a known point. Check that the pole height, antenna type, coordinate system and correction source are correct before assuming multipath is the cause.

If the discrepancy remains location-specific, treat the environment as the likely issue and change the measurement method. This is where access to technical advice, equipment hire and instrument support can make a real difference. Survey Tech can help teams select GNSS and total station equipment appropriate to the site conditions, rather than forcing one method into every task.

The useful habit is to regard GNSS quality indicators as evidence, not a guarantee. When the sky view is poor or reflective surfaces are close, verify the work, move the observation where possible, and use the instrument that gives the most defensible result.