Why RTK Won’t Reach the Expected Accuracy: 6 Causes of Error Explained

You may have expected RTK to always provide high accuracy, but in actual field work the numbers can sometimes vary more than you thought. Many practitioners struggle because a point can be stable in one spot but suddenly lose accuracy after moving a little, or because measurements taken on a different day don’t match previous results and the cause is unclear.

RTK is a convenient positioning method, but simply turning on the receiver and measuring does not guarantee constant high accuracy. The true accuracy is realized only when multiple conditions are met: the satellite reception environment, the quality of correction information, communication status, initialization procedures, coordinate settings, and field operation habits. Conversely, if you understand where errors tend to enter, you can prevent many accuracy problems in advance.

This article organizes and explains six points to suspect when RTK accuracy is not achieved, aimed at practitioners searching for “RTK error cause.” Rather than mere theory, it focuses on common field mistakes and practical checks that are easy to review on site. If you want to reduce re-measurements and rework, please read to the end.

Contents

• First understand the mechanisms that affect RTK accuracy

• Cause 1: Poor sky view and multipath (reflected wave) effects

• Cause 2: Unstable correction information that cannot maintain fixed solution

• Cause 3: Observing during times with poor satellite count or geometry

• Cause 4: Inconsistencies in equipment setup or initialization procedures

• Cause 5: Coordinate system or height settings that don’t match field conditions

• Cause 6: Variation in observation methods and field operations that widen errors

• Field practices to enforce to reduce RTK errors

• Summary

First understand the mechanisms that affect RTK accuracy

To correctly identify RTK error causes, it is important to first understand why RTK achieves high accuracy. RTK does not perform standalone positioning using only signals from satellites; it combines reference station information with rover observations to correct positions and thereby achieve high precision. In other words, positioning accuracy is influenced not only by the receiver’s hardware performance but also by satellites, reference information, communications, the observation environment, and setup procedures.

A common field reaction when numbers are off is to assume “the machine is malfunctioning.” In reality, environmental and operational conditions often affect accuracy more than the receiver itself. Especially near buildings, below slopes, close to trees, on earthwork sites with moving heavy machinery, or in mountainous areas with unstable radio signals, observations taken using the same procedures can easily become erratic.

Also, even if the RTK screen shows that positioning is completed, you must verify whether that state is truly stable. Displaying coordinates is not the same as observing with trustworthy accuracy. You need to check multiple states together: whether a fixed solution has been achieved, whether there are enough satellites, whether correction information is being continuously received, and whether solution switching is occurring. If you do not confirm these, measurements that look valid may in fact contain errors.

Therefore, RTK accuracy problems are often not due to a single cause. In one field the sky view may be the primary factor, while in another the main issue could be communications or settings. What matters is not mindlessly re-observing when errors occur, but having an order for isolating causes. Simply suspecting the six items introduced below in order will make on-site decisions considerably easier.

Cause 1: Poor sky view and multipath (reflected wave) effects

One of the most frequent causes of RTK error is poor sky view and multipath effects. Satellite positioning works by receiving signals from the sky to compute position, so the more open the sky, the better. Conversely, when buildings, retaining walls, slopes, trees, guardrails, metal fences, heavy machinery, or material yards are nearby, satellite signals can be blocked or reach the receiver after reflecting off surroundings, making correct calculations difficult.

Pay special attention to multipath. Multipath is hard to notice visually and can occur even when the field staff feel “the sky is sufficiently open.” For example, near building walls or large vehicles, direct and reflected waves can mix, causing position values to drift gradually. This type of error often appears not only as extreme meter-level shifts, but also as intermediate differences of several centimeters to several tens of centimeters, which are particularly troublesome in practice. In tasks like as-built verification or stakeout where small differences matter, such deviations cannot be ignored.

On construction and civil engineering sites, changing site conditions are also problematic. A location that was open in the morning may have heavy machinery or materials placed there by noon, causing observation conditions to worsen in the afternoon. A point that posed no problem yesterday may show increased error today because the surrounding environment changed. In other words, past successful measurements alone cannot guarantee the same quality at the same location.

Signs to suspect this cause include: accuracy dramatically improves by moving a short distance, values vary with orientation or stance at the same point, it takes a long time to obtain a fixed solution, or values remain unstable even after a fixed solution is obtained. In such cases, the basic first step is to look around and check for nearby reflectors or obstructions.

The key countermeasure is to choose the most open sky possible. Avoid walls and directly under trees, and prioritize locations with fewer surrounding structures. If obstructions are unavoidable, moving a short distance to a safer location often improves results. Also, do not finalize a measurement immediately after values appear—confirm their stability over a short period. Many environment-induced errors can be prevented by on-site judgment, so develop the habit of first suspecting the sky-view environment.

Cause 2: Unstable correction information that cannot maintain fixed solution

RTK accuracy greatly depends on whether correction information can be received stably. RTK is not a standalone positioning method; it corrects errors while receiving reference information. Therefore, if correction reception is interrupted or its quality degrades, the expected accuracy cannot be achieved even if satellites are visible.

A commonly overlooked situation in the field is that positioning appears to continue on the screen while correction status has become unstable. In weak communication areas, correction reception can be intermittent, causing the solution to switch between fixed and float. As a result, the displayed numbers may look plausible but may not actually represent stable high accuracy. Judging by positioning results alone can miss this state.

This issue is more likely in areas where radio signals are hard to reach: mountainous regions, low areas on earthworks, near underground structures, or places surrounded by buildings. In operations that observe while moving, even slight communication quality changes can destabilize the solution state. Also, the base-side information itself may be unstable, or the state may not be settled immediately after reception starts, so corrections may appear to be applied even when they have not reached sufficient quality.

In practice, do not be reassured by the fixed-solution display alone. Immediately after switching to a fixed solution, values may still be settling, and the state can momentarily collapse and re-fix during observation. These small state changes can accumulate and later manifest as positional discrepancies when compared. Especially when measuring multiple points in succession, skipping state checks for each point risks producing results contaminated with errors.

Countermeasures begin with ensuring a communication environment capable of stably receiving corrections. Moving slightly can improve reception in weak spots. Also, after obtaining a fixed solution, do not finalize immediately—observe whether the state is maintained for a short time. If fixed and float repeat, question observations at that point and review the environment or position. Corrections are critical to RTK. If corrections remain unstable, no amount of careful work will stabilize accuracy.

Cause 3: Observing during times with poor satellite count or geometry

In RTK, seeing many satellites does not automatically guarantee high accuracy. What matters is not only the number of satellites but also their geometric distribution. If satellites are concentrated in one part of the sky, the conditions for position computation worsen and small observation errors can more easily amplify into coordinate deviations. Conversely, if satellites are distributed evenly across the sky, a more stable solution is likely.

In the field, the same location can be easier or harder to measure depending on the time of day because satellite geometry changes over time. Phenomena such as being stable in the morning but harder to fix in the afternoon, or good one day but poor another day, can be related to differences in satellite geometry. When results are unstable despite similar sky environments, suspect the time-of-day influence.

Also, even if the display shows a sufficient satellite count, relying primarily on low-elevation satellites makes observations more susceptible to terrain and atmospheric effects. Don’t be reassured by the numeric display alone—assess whether the observation conditions are genuinely good. A common practitioner mistake is applying yesterday’s successful settings without change today. Satellite conditions are not constant, so assume there will be days that are easier or harder to observe.

To detect this cause, compare measurements at the same point at different times. If the time of day affects time to fix, value stability, or reproducibility, satellite geometry likely plays a strong role. For tasks requiring strict accuracy, select observation times that are more likely to yield stable results rather than choosing times solely for convenience.

Countermeasures: if you have flexibility, be mindful of time-of-day conditions before observing and avoid times that look unfavorable. Also, if a point is hard to fix, do not force the observation—shifting the time slightly can improve the state and prevent rework. RTK results depend not only on equipment performance but also on observing timing. When accuracy is lacking, do not overlook time-of-day effects.

Cause 4: Inconsistencies in equipment setup or initialization procedures

RTK accuracy problems can be caused not only by field conditions but also by human procedures. A common issue is observing without sufficiently careful equipment setup or initialization handling. No matter how good the satellite or correction conditions are, if the setup assumptions are off, the obtained coordinates will be shifted.

Typical cases include a pole that is not plumb, ambiguous handling of setup height, immediately finalizing after re-setting the instrument, or moving to the next point before initialization has stabilized. Field work often prioritizes efficiency, which can lead to procedural shortcuts, but skipping small steps in RTK can cause centimeter-level differences. Height errors tend to be more apparent than horizontal errors, so pole tilt or height input mistakes easily appear in results.

Differences in procedures among observers are also problematic. One person may wait for sufficient stability after fixing before finalizing, while another finalizes the instant the display changes. That difference alone can create reproducibility gaps across the site. When multiple people work on the same site, standardizing operational rules is more important than the equipment itself.

Initialization issues should not be overlooked. Right after moving, immediately after communication recovery, or after reception restarts, the display may return but the state may not be fully stable. Rushing to finalize under these circumstances leads to point-to-point variance that shows up later. Being able to distinguish between “positioning is available” and “observation conditions are settled” is critical; if this distinction is vague, unexplained errors remain.

Countermeasures: First, enforce the basics of setup. Do not skip confirming pole plumbness, consistently handle observation height, check stability after re-setting, and avoid finalizing immediately after fixing. Also, define on-site criteria for “what state to confirm before finalizing” so that anyone observing will make the same decision, which improves reproducibility. RTK is a high-functioning method, but the final accuracy depends on the quality of field procedures. If basic actions are neglected, even the most sophisticated equipment will not deliver expected results.

Cause 5: Coordinate system or height settings that don’t match field conditions

Some RTK errors arise not from satellites or communications but from mismatched settings. Especially common are cases where the coordinate system or height reference does not match the field’s standard. These errors often appear only when overlaying results with existing drawings or other survey outputs, despite measurements appearing clean on site, and they can cause significant rework downstream.

For example, even if the horizontal coordinate system seems correct, if you are comparing to a different reference or the height standard is inconsistent, the site may appear shifted on plans despite having measured correctly. This is troublesome because the values are stable during observation, leading the operator to suspect the equipment rather than the drawing. In many cases the cause is simply a mismatch in interpretation conditions rather than positioning performance.

Also be careful when multiple people or multiple days are involved and settings are not unified. If one day you observe with the correct reference but another day collect data with different settings, inconsistent datasets will mix within the same site. When problems arise during earthwork volume comparisons, as-built checks, drawing overlays, or point cloud usage, identifying the cause can take time. Often only a vague feeling of “this is different from last time” appears first, delaying detection of the settings discrepancy.

Height standards deserve special attention. Even if horizontal coordinates look consistent, differences in vertical reference can cause only elevation to be offset. This can produce unexpectedly large discrepancies in fill and excavation volumes or final surface checks. If you assume this is an observation accuracy issue and repeatedly re-measure, you won’t solve it—because the cause is a reference mismatch, not observational noise.

Countermeasures: before observing, clarify the reference standards to be used in the field and align stakeholder understanding for both horizontal and vertical references. Identify which existing drawings, design data, past results, or field checks will be used as comparison targets and ensure observation settings match those comparison conditions. If observing the same site over multiple days, record settings so you can continue under identical conditions. Many problems that look like RTK accuracy issues are actually setting inconsistencies. If accuracy feels off, first re-examine the measurement assumptions.

Cause 6: Variation in observation methods and field operations that widen errors

RTK errors are expanded not only by single observation conditions but also by variation in overall field operations. Even with the same equipment, site, and day, a lack of standardized observation methods causes point-to-point quality differences. This is a very common field cause, but because each observer often believes they are doing the right thing, the problem is slow to surface.

For example, one operator inspects surroundings before observing and records only after the fixed state stabilizes; another prioritizes speed and observes many points in quick succession. One person re-observes suspicious points while another finalizes on a single result. Such differences across the site will produce a mix of good and poor quality points when results are compiled, reducing overall reliability.

Moreover, on sites without verification points, errors can go unnoticed. RTK’s convenience tends to make operators assume every point is correct. In practice, however, you need representative points, known points, or re-observation points to check quality during work. Without these, everything may look fine on site, but issues will surface only when overlaying drawings or point clouds later—often after leaving the site and unable to recheck.

Another problem is the inability to stop work when signs of poor accuracy appear. When schedules are packed, teams may push on even if points are hard to fix, values fluctuate, or surroundings look poor. But pursuing quantity under bad conditions creates problems later. Small per-point differences accumulate into large overall deviations. The more points collected, the more critical operational rules become for quality.

Countermeasures: do not make field operations depend on individuals. Predefine when to finalize, under what conditions to re-observe, how many points between verifications, and who decides when in doubt. Such rules stabilize quality. Also, provide a system for simple on-site result checks. Immediately checking observed coordinates and verifying connectivity with adjacent points or consistency with known points makes it easier to detect outliers early. RTK accuracy depends not merely on equipment performance but on the design quality of field operations.

Field practices to enforce to reduce RTK errors

Having reviewed six causes, the most important RTK error countermeasure is to eliminate high-risk situations in advance, rather than searching for causes after problems occur. What is truly effective on site is not memorizing complex theory, but having an ordered checklist for pre- and post-observation checks.

Before observing, first check sky openness and nearby reflectors. Next, confirm that corrections and communications are stable. If satellite conditions look poor, suspect time-of-day effects. During observation, carefully confirm setup and initialization. Before finalizing, verify coordinate system and height reference. Finally, use verification points or re-observations to confirm quality. Simply formalizing this flow as a rule will greatly reduce field accuracy problems.

The key is not to regard RTK as infallible. RTK is an extremely effective method, but it is sensitive to field conditions. Therefore, do not treat the mere appearance of numbers as success; instead, cultivate the habit of judging whether those numbers are trustworthy. Experienced operators often make this judgment unconsciously. Conversely, relying solely on high-spec equipment with the mindset “it should be fine” makes repeated mistakes more likely.

In practice, what matters more than measuring itself is how the results are used. RTK coordinates become the reference for downstream tasks: stakeout, as-built verification, earthwork volumes, drawing overlays, or point cloud utilization. A few centimeters of initial error can translate into differences in construction decisions or quantity assessments later. Thus, sparing the extra check on site is ultimately the most efficient approach.

If you want to balance field speed and accuracy, set up an environment that allows handling measurement, verification, and sharing as one workflow. Making high-precision position information easy to use on site and enabling immediate checks rather than “measure and stop” helps detect errors sooner. If you want stable RTK operation in practice, consider not only receiver selection but also how easy the system is to operate on site.

Summary

The causes of RTK failing to achieve expected accuracy are not limited to a single factor. Poor sky view and multipath, unstable correction information, unfavorable satellite geometry times, setup and initialization mistakes, mismatched coordinate or height settings, and variation in field operations combine to produce the phenomenon of “unexplained drift.” Therefore, reducing errors requires reviewing not only equipment performance but also observation conditions and operational procedures as a set.

In practice, it is more effective to have an order of suspicion before measuring rather than chasing causes after large errors appear. Check whether the sky is open, whether corrections are stable, whether the fixed solution can be maintained, whether setup is correct, whether reference settings match, and whether verification points have been used. Enforcing these basics will greatly improve RTK reproducibility.

If you want to make high-precision positioning easier to handle in the field and streamline measurement through verification, operability is also an important decision factor. LRTK, as an iPhone-mounted high-precision GNSS positioning device, pairs well with situations where you want to quickly handle high-precision location information on site, making daily positioning tasks more approachable. If you are troubled by RTK errors and want an easier way to bring high-precision positioning into the field, consider the option of simple surveying using LRTK.