5 Ways to Deal with Unstable GNSS in Drone Surveying

Table of Contents

• What happens when GNSS becomes unstable during drone surveying?

• Main causes of GNSS instability in drone surveying

• Measure 1: Review takeoff and landing sites and overhead visibility

• Countermeasure 2: Isolate the effects of the radio environment and surrounding structures

• Countermeasure 3: Thoroughly perform initialization and confirm the fixation status before starting observations

• Countermeasure 4 Reconfigure flight plans and shooting conditions based on GNSS

• Countermeasure 5: Prepare supplementary measures to prevent re-measurement

• Decision criteria for field personnel at sites with unstable GNSS

• Summary

What Happens When GNSS Becomes Unstable in Drone Surveying?

In drone surveying sites, attention tends to focus on aircraft performance, camera performance, and flight planning, while what has a major impact on practical quality is the stability of positional information. In particular, when GNSS becomes unstable, the flight itself may appear to conclude without issue, yet doubts about the reliability of the results can arise in later processes. What makes this troublesome for field personnel is that GNSS instability does not necessarily manifest as an obvious anomaly on the spot. Although the capture may be completed, problems often emerge afterward in the form of shifted orthophotos after processing, poor alignment of point clouds, mismatches when compared with existing drawings or known control points, or large differences in repeated measurements.

The role of GNSS in drone surveying is not merely to know the aircraft’s approximate current location. It is involved in multiple processes, including flight stability, reproducibility of capture positions, georeferencing of captured data, and the reliability of initial values during post-processing. Therefore, if work proceeds while GNSS remains unstable, even if it appears that data were captured satisfactorily on site, unnecessary checks and re-surveys are likely to occur during the product-creation stage. In surveying practice, the occurrence of such re-surveys represents a significant loss, because invisible burdens—personnel, travel, waiting for weather, coordination with stakeholders, and on-site attendance—increase all at once.

Furthermore, GNSS instability is not necessarily a problem only in mountainous areas or urban areas. Even on open development sites or construction sites with many temporary installations, observation quality can fluctuate due to surrounding metal materials, slopes, embankments, temporary fencing, heavy equipment, power transmission equipment, communication conditions, and the choice of takeoff and landing locations. In other words, this is not a special issue that occurs only at particular difficult sites; it should be treated as a general risk that can occur at any site.

That is why it is important not to respond hastily and improvisationally when GNSS becomes unstable, but to be prepared to address issues before flight, during flight, and after flight while isolating the causes. This article organizes and explains five practical countermeasures that are easy to implement on site when GNSS is unstable during drone surveying. Rather than merely describing functions, it is summarized from the perspective of how field operators should make decisions, what to review, and how to reduce re-shooting and rework.

Main causes of GNSS instability in drone surveying

Before looking at countermeasures, it is important to first clarify why the GNSS is becoming unstable. If the cause remains unclear, it is easy to misjudge whether the issue on site is caused by the aircraft, the surrounding environment, or operational procedures. As a result, you may waste time repeatedly revising settings when simply changing the takeoff/landing location would have sufficed, or conversely keep moving location when what is actually needed is a review of the settings.

One common cause of GNSS instability is spatial conditions that make it difficult to receive signals from satellites. In locations where the sky overhead is not sufficiently open, the number of visible satellites tends to decrease and the azimuthal distribution can become severely biased. Even when the sky appears to be visible, reception conditions may actually be poor if there are tall buildings, slopes, trees, bridges, temporary structures, or similar obstructions nearby. In particular, when low-elevation directions are blocked, positioning stability is especially likely to be affected.

The second is the effect of reflections. Because GNSS signals are extremely weak, picking up signals reflected from nearby metal surfaces, water surfaces, glass surfaces, large construction machinery, steel materials, fences, and so on can readily increase errors. This can occur even when the sky above is open, so you cannot simply assume that open space is safe. In places with many metal objects, such as development sites and material storage yards, this is an easily overlooked factor.

The third factor is the communication environment and the status of acquiring correction information. In operations that perform high-precision positioning, it is assumed that reception of correction information and initialization status are stable. In situations where communication is weak, there are many obstructions, or connections to the base station or network are unstable, it may take longer to achieve a fixed solution, or the accuracy state may change partway through. What field personnel tend to overlook is becoming complacent because the communication is temporarily connected. A momentary connection and a sustained condition that can withstand surveying-quality requirements are not the same.

The fourth issue concerns the timing and procedures for starting observations. If you take off immediately upon arriving on site or begin operations before initialization is complete, you can carry instability through the entire subsequent flight. This tends to occur at sites that shorten the preparation phase in order to prioritize flight time. Because GNSS is not a visibly moving component, it is easy to skip checking it, but omitting this step can lead to irretrievable rework.

The fifth issue is that the flight plan itself does not match the on-site conditions. Long-distance flights, routes with significant terrain undulation, continuous low-altitude flights, flights close to obstacles, and operations spanning multiple batteries amplify quality risks even if GNSS becomes only slightly unstable. Ideally, the flight area, altitude, course selection, overlap rate, and the placement of ground control points should be adjusted according to the site's reception conditions, but flying with the same template as always makes the problem worse.

In this way, GNSS instability does not necessarily arise from a single cause. It is most often the result of a complex combination of factors—location, environment, communications, procedures, and planning. Therefore, countermeasures should not rely on a single configuration change; instead, you need an approach that systematically isolates and rectifies on-site conditions one by one.

Countermeasure 1 Review takeoff and landing sites and overhead visibility

The most basic and yet highly effective countermeasure is to reevaluate the takeoff and landing location. When GNSS is unstable, what should be suspected first at many sites is the location rather than the aircraft. Even within the same site, moving just tens of meters (tens of ft) can often improve reception. Nevertheless, there are cases where the spot that is easiest for bringing in equipment, parking vehicles, or gathering personnel is used as the takeoff and landing point as-is, and as a result work begins from a position with poor reception conditions.

What matters when choosing a takeoff and landing site is not just whether it is flat and safe enough to place the aircraft. You need to assess the openness of the sky, nearby obstructions, the presence of reflective sources, and how easily communications will pass. For example, areas near slope faces, beside trees, next to temporary housing, close to steel materials or heavy machinery, along guardrails, or in the downstream space under a bridge may appear easy to work in at first glance but can be unfavorable for GNSS. Positions close to metal objects or large structures are particularly susceptible to reflection effects—more so than to simple reception issues—so caution is required.

On site, it is important to begin with the mindset of having multiple candidate locations. If conditions at the first spot are not stable, you should be ready to move to another site immediately. Changing the takeoff and landing site may seem troublesome in the field. However, spending 5 to 10 minutes to reassess there is far more efficient than spending hours after the flight checking because you suspect a positional offset. In surveying, flying itself is not the goal; the goal is to obtain reliable results. Therefore, you should prioritize good starting conditions over the speed of takeoff.

Also, when assessing the overhead sky view, it is not enough to look only straight up. Check whether the sky is widely open down to low angles, whether any particular direction is significantly obstructed, and whether there are nearby surfaces that might cause reflections. Even an intuitive judgment from the spot where you are standing on the site can make a difference, but if you are not accustomed to it, it becomes easier to judge if you make a habit of walking around the area, looking up, and checking which directions are closed. This check is effective not only in urban and mountainous areas but also at civil engineering sites with many material yards and temporary structures.

When reviewing takeoff and landing sites, also check their relationship to the flight route. Even if reception conditions are good at the starting point, it's pointless if the flight immediately enters an area with strong signal obstruction. From the takeoff point through the first turn, climb, and the position where imaging begins, it's important to secure a flight path that allows the aircraft to stabilize easily. Especially when using autonomous flight, stable behavior immediately after the start affects the quality of the subsequent data.

When GNSS is unstable on site, people tend to focus on equipment settings or rebooting, but simply developing the habit of first questioning the location can greatly improve the initial response to problems. Choosing takeoff and landing sites may seem mundane, but it is the foundation that determines the quality of drone surveying.

Countermeasure 2: Separate the effects of the radio environment and surrounding structures

At sites where GNSS is unstable, it's important not just to look at satellite visibility but to isolate the impacts coming from the surrounding environment. By "isolate" we don't mean trying to avoid everything that simply seems problematic; rather, it means clarifying which factors are likely having what degree of effect and prioritizing corrective actions. Being able to do this speeds up on-site decision-making and reduces needless trial and error.

First, check for proximity to metal objects and large structures. Near heavy machinery, steel materials, temporary enclosures, fences, material containers, steel frames, bridges, and metal roofs, you are more likely to be affected by reflections. In such environments, even if the receiver status display hasn’t completely degraded, position dispersion can increase. As the site supervisor, it’s important not to be reassured by a seemingly normal display alone. If conditions look even slightly suspicious, shifting the takeoff and landing point a few meters to several tens of meters (a few ft to several tens of ft) away to increase distance from nearby reflective sources can often improve the situation.

Next, you should check for the presence of water surfaces and wet ground. Along rivers, around retention ponds, on developed land with many puddles, and on paved surfaces after rain, reflection conditions may change. The presence of a water surface does not necessarily mean conditions are immediately unusable, but the impact can be significant when it overlaps with surrounding structures. Careful judgment is especially required for low-altitude flight and for takeoff and landing locations close to bodies of water.

Furthermore, the communication status is also a point that should be checked simultaneously. GNSS instability and communication instability are separate issues, but in high-precision positioning operations they can appear as a single problem from the field perspective. In other words, satellite reception may be fine but acquisition of correction information may be unstable, resulting in an unstable accuracy state, or conversely communication may be available but reflections from the surroundings may prevent the position from stabilizing. Confusing these will lead to incorrect countermeasures. If communication is weak, move to a location with better communication conditions; if reflections are suspected, move away from the reflection source—actions should be separated according to the cause.

A simple approach to isolating problems on-site is to change location and observe the condition. If moving slightly leads to improvement, the surrounding environment may be having a strong influence. If waiting for a while leads to improvement, initialization or the status of acquiring correction information might be the cause. If it becomes unstable when flying in a particular direction, consider the effects of obstructions or structures along the flight route. Accumulating these observations will prevent you from simply dismissing the issue as a malfunction of the aircraft.

Also, sharing awareness among workers on site is important. Even if GNSS is described as unstable, the phenomena observed can differ between the pilot/operator, the surveyor, and the post-processing staff. The pilot may perceive it as the aircraft being unsettled, the surveyor as difficulty maintaining a fixed state, and the processor as difficulty with position alignment. If these are not treated as separate problems but can be shared along a single axis—the influence of the surrounding environment—the quality of on-site decision-making improves.

The effects of radio-wave environments and surrounding structures tend to be underestimated precisely because they are not easily visible. However, in drone surveying, these hard-to-see conditions can greatly affect the stability of results. Deciding not to choose a location based solely on apparent ease of work, and always questioning how the surrounding environment may affect GNSS, is the first step to preventing re-shooting.

Countermeasure 3 Thoroughly initialize and verify the fastening status before starting observations

What is easily overlooked when GNSS is unstable is the initialization and status check before starting observations. On site, people tend to want to fly as soon as preparations are complete. This tendency is especially strong on days when changing weather is a concern or at sites where stakeholders’ attendance time is limited. However, if you skimp on the start-up conditions to save a few minutes here, you will end up losing many times that amount of time later.

In drone surveying that assumes high-precision positioning, merely acquiring satellites is not the same as being in an observation state that can meet survey-quality requirements. You need to properly check before starting things such as the number of satellites being received, the positioning/solution status, the reception status of correction information, the stability of any fixed solution, and how convergence evolves over time. The important point here is not to judge by a quick glance at the display. Even if it looks good at the start, the condition may fluctuate after a while. It is important to verify whether it remains stable even for a short period.

In practice, rather than taking off immediately upon arrival on site, simply allowing a little time to observe the aircraft and receivers after setting them up can make a difference. This time is not idle waiting; it is time for quality checking. In surveying, more than whether the number of photos taken or the flight time match the plan, whether the initial conditions at the start were appropriate determines the quality of the results. Especially at sites where GNSS instability is suspected, it is effective to avoid rushing initialization.

Also, it's important not to be reassured just because a fix has been acquired once. Check whether it can be maintained stably, whether the condition changes before and after takeoff, whether reception of corrections is uninterrupted, and whether there are any issues at the flight plan's start position. If even a small concern remains here, you should postpone takeoff to isolate the cause. What you most want to avoid on a surveying site is going ahead despite recognizing a problem. In my experience it may sometimes work out, but from a quality control perspective it has low reproducibility and is not suitable for systematic operations.

Furthermore, it is effective to verbalize the checklist items for each site. For example, if personnel maintain a common procedure that covers the flow—checking sky visibility, checking nearby reflective sources, checking communication status, checking correction acquisition, confirming stability of the fix, and confirming the suitability of the takeoff/landing site—decisions become less prone to inconsistency. GNSS instability tends to recur more easily the more it is handled by feel, so it is important not to leave verification tasks to individual hunches.

Thorough checks before beginning observations make the flight itself more likely to remain stable. The pilot can avoid unnecessary anxiety, and the surveyor can be confident in the preconditions for the results. The few minutes before starting are the most valuable time, especially at sites where GNSS is unstable.

Countermeasure 4: Reconfigure Flight Plans and Shooting Conditions Based on GNSS

A common failure at sites with unstable GNSS is applying your usual successful patterns unchanged. Even if the same flight altitude, the same overlap rate, the same course settings, and the same area segmentation have always been fine, they may not work as-is at sites with poor GNSS conditions. The important thing is to adjust the flight plan itself according to the site conditions.

The first thing to reconsider is the area you attempt to cover at one time. If you try to capture a wide area all at once at a site where GNSS is unstable, any fluctuation in observation conditions partway through will have a large impact. The larger the area, the harder it becomes to pinpoint where the problem occurred. Therefore, when conditions are questionable, it is safer to divide the area and reliably acquire data in shorter segments. Even if this increases processing effort somewhat, the practical benefits are substantial if it reduces the risk of having to re-survey.

Next, review how you plan the flight route. If there are many obstructions nearby, poor reception in certain directions, or large terrain undulations, you need to take measures such as avoiding directions prone to problems, changing the starting position, or choosing a route with extra margin. It is important not only to efficiently cover the survey target but also to create flight lines that help the aircraft remain stable. Especially in automated flights, the initial course selection can strongly affect later quality, so designing to prioritize stability over simply choosing the shortest path is effective.

In terms of capture conditions, attention must also be paid to image overlap. When GNSS is unstable, ensuring sufficient connectivity between images is effective for improving consistency in post-processing. However, simply increasing overlap indiscriminately isn't the answer; you must balance it against flight time, battery consumption, and processing load. The key idea is that the worse the field conditions, the more you should allow room to rely on downstream processing. If GNSS conditions are not ideal, you need a capture strategy that supports quality through image connectivity and supplementary information.

Also, flying at too low an altitude or flying close to obstacles is easily affected even by slight GNSS instability. Planning to get too close to the target object may at first glance seem to yield highly detailed results, but in environments where stable positioning information is hard to obtain it can have the opposite effect. It is important to adopt reasonable conditions while balancing the required accuracy, safety, and stability.

On sites with large terrain variations, you should reconsider how you handle elevation differences. In areas where ridges, valleys, slopes, embankments, and cuttings coexist, actual ground conditions can differ greatly even with the same altitude setting. If GNSS is unstable in such locations, it increases variability in capture quality and makes post-processing more difficult. Therefore, rather than treating the entire site uniformly, it is effective to separate the plan by terrain type.

Reviewing a flight plan is not merely an efficiency adjustment. It is quality planning to protect the quality of deliverables in environments where GNSS is unstable. The level of risk that can be tolerated depends on the surveying objective—whether creating plan views, verifying earthwork volumes, or capturing as‑built conditions. That is why, when site conditions are poor, you should not fly as usual but reorganize the plan into one that is necessary and sufficient for the objective.

Countermeasure 5: Prepare complementary measures to prevent re-measurement

At sites where GNSS is unstable, one of the most important principles is not to entrust everything to GNSS alone. Of course, positional information is at the core of survey quality, but in practice it is important to have complementary measures so that the entire operation does not collapse when any single condition fails. What truly causes trouble on site is realizing a problem after the flight and finding that there is nothing left there to support the quality.

As a complementary measure, the first thing to consider is to have reference information available on-site. For example, simply handling known points and references that can be checked in the field carefully makes it easier to verify positional validity in subsequent steps. Even if GNSS is not completely stable, having a reference for comparison allows you to evaluate the results. Conversely, without a reference for comparison you cannot judge how much you can trust the data if a problem arises.

Also, do not overlook keeping site photos and records. If you record the condition of the takeoff and landing points, nearby obstructions, locations where communication was poor, and any behaviors during flight that seemed concerning, it will be easier for the post-processing personnel to infer the situation later. Problems during GNSS instability often involve information that only those who were on site know. If you hand over only the data without leaving records, it will be treated in later stages as an unexplained positional shift, increasing unnecessary checks.

Furthermore, it is important to consider introducing redundancy into on-site acquisition methods where necessary. For example, measures effective at avoiding a full remeasurement include arranging to be able to retake only critical locations under different conditions, adding observations for verification, and securing materials that are easy to cross-check in later processes. You do not need to duplicate everything perfectly, but having at least a fallback for quality checks for important targets and boundary areas can be invaluable if something goes wrong.

The idea of having complementary measures does not mean distrusting GNSS. Rather, it means creating a state in which high-precision positioning information can be reliably used on site because it is supported by other sources of information. In surveying practice, the stronger the dependence on a single condition, the greater the loss when problems occur. Conversely, if multiple means of verification are available, it is easier to maintain overall quality even with some instability.

Especially at construction sites, conditions change daily. Even locations that had no problems yesterday may have different conditions today due to the placement of heavy machinery or the addition of temporary structures. Therefore, it is important not to rely solely on past successes but to prepare each time with complementary measures in mind. A truly resilient operation at a site with unstable GNSS is not one that eliminates problems entirely, but one that can recover without re-surveying when issues occur.

Decision-Making Criteria for Field Personnel at Sites with Unstable GNSS

So far we have presented five countermeasures, but on-site there are always time constraints, so it is not always possible to examine everything thoroughly. What becomes important for practitioners, therefore, is to have criteria for deciding whether to continue, change the conditions, or suspend operations. It is not as simple as stopping immediately just because GNSS is unstable; you need to separate tasks where instability can be tolerated from those where it absolutely cannot be tolerated.

First, it is important to consider the intended use of the deliverables as the criterion. The conditions required differ between tasks whose primary purpose is to capture relative changes roughly—such as gaining an overall picture or checking progress—and surveying results that demand high positional reliability. You must distinguish between operations that can proceed even when GNSS is somewhat unreliable and operations that should be reconsidered if the initial conditions are even slightly uncertain. If this is left ambiguous, it will later lead to a misalignment of understanding among stakeholders.

Next, check whether there is a trend of improvement. If improvement can be confirmed — for example, it stabilized after changing the location, the condition settled after waiting, or issues became less likely after separating the flight area — there may be a possibility to proceed conditionally. On the other hand, if the condition fluctuates no matter what you do, the cause cannot be narrowed down, or behavior differs greatly from flight to flight, you may need to decide not to force progress. On site, continuing in the expectation that “it’s probably fine” because things got a little better can lead to wholesale doubts in later stages. It is important to regard the presence or absence of improvement as a change, not as a feeling.

Whether it can be explained on the spot is also a criterion. For survey results, it is important to be able to explain later why they were carried out under those conditions. If the reasons for changing the takeoff and landing points, dividing the flight area, or adding verification information are clear, the operation is coherent. Conversely, a decision to proceed simply because it seemed likely to work is weak from a quality-control standpoint. In practice, not only whether it succeeded but also whether the decisions were reasonable will be scrutinized.

Furthermore, the cost of revisiting is also a practical factor to consider. At sites where a revisit is extremely difficult, you should act to secure as much supplementary information on the spot as possible. Conversely, at sites where the likelihood of a revisit is relatively high, it may be better to wait for more favorable conditions rather than forcing the collection of low-quality data. The important point here is that trying to finish everything in a single visit is not necessarily the best approach. Surveying is evaluated by the reliability of its outcomes, not by the small number of visits.

The most dangerous thing when judging GNSS instability is carrying the field’s unease into later processes. If the on-site person feels even slightly unsure, it is important to record that concern, change the conditions if necessary, and, in some cases, have the courage to stop. Drone surveying is an efficient method, but if you prioritize efficiency too much and become lax in quality judgment, you will lose its intended benefits.

Summary

When GNSS is unstable during drone surveying, it is important not just to suspect the aircraft or settings, but to comprehensively review factors including the takeoff and landing site, the surrounding environment, initialization procedures, the flight plan, and supplementary measures. Many GNSS problems encountered in the field arise from a combination of environmental and operational factors rather than a single cause. Therefore, instead of trying to resolve the issue with a single setting change, it is important to methodically isolate and address causes in sequence.

The five countermeasures introduced here are not limited to special sites. Reviewing takeoff and landing locations, considering the effects of reflections and obstructions, never skipping pre-flight checks, tailoring the flight plan to site conditions, and having backup measures are fundamental practices that help prevent the need for re-shooting at any site. In particular, in practical operations decisions should be based not on whether a flight is possible but on whether it will produce reliable results.

At sites where GNSS is unstable, the surveying team's planning ability and judgment directly affect the results. Even if flights appear to proceed the same, differences in the initial conditions and on-site decisions manifest as quality differences in later stages. That is why it is important to stop when you sense something is off on site, sort out the cause, and readjust the conditions. The accumulation of that habit leads to stable operations with fewer re-surveys.

Also, at sites where you want to make daily positioning checks and on-site location awareness more reliable, measures to make high-precision positioning information easier to handle on the ground are also effective. For example, if you want to streamline site reference checks, auxiliary positioning, and position management before and after data capture, using an iPhone-mounted GNSS high-precision positioning device like LRTK can make it easier to increase the information available for on-site decision-making. To stabilize the quality of drone surveying, it is important not only to acquire data from the air but also to consider how to support position management on the ground. The more prone a site is to GNSS instability, the more an approach that considers positioning from both the air and the ground leads to operations with less rework.