How accurate can RTK-GPS be? Typical error margins and 6 checkpoints

For practitioners considering the introduction or use of RTK-GPS, the biggest concern is likely “how much accuracy can actually be achieved.” On site, if decisions are made based solely on the numbers in the catalog, you may encounter problems such as accuracy not being as stable as expected, lack of reproducibility, or large errors only in elevation. RTK-GPS is widely known as a high-precision positioning method, but it does not deliver the same accuracy at every site. To consistently obtain good results, it is essential to correctly understand the expected error margins and to identify, on a site-by-site basis, the conditions that affect accuracy.

Many people who search for "RTK GPS accuracy" are likely more interested in practical, on-site accuracy and points to check than in theoretical explanations. In situations where RTK-GPS is used in practice—surveying, construction management, as-built verification, stakeout, maintenance, integration of positional data with photographic records, etc.—differences of a few centimeters can directly affect work quality and lead to rework. Therefore, it is not enough to simply understand it as "centimeter-level." A realistic perspective is needed: how much difference tends to occur between horizontal and vertical, what changes depending on the reception state of correction information, and how to approach locations with many trees or buildings.

In this article, after summarizing how much accuracy can be expected from RTK-GPS, we explain from a practical perspective the typical magnitude of errors and six on-site points to check. It is organized to be useful not only for those considering introducing it, but also for those already using it who are wondering, "Is this error normal?" or "What should I review to make it more stable?"

Table of Contents

• How accurate can RTK-GPS be expected to be?

• How should one assess the expected error of RTK-GPS?

• Checkpoint 1: Confirm the conditions for the correction information

• Checkpoint 2: Confirm whether the positioning status is truly stable

• Checkpoint 3: Confirm sky visibility and satellite configuration

• Checkpoint 4 Do not overlook reflections and obstructions caused by the surrounding environment

• Check item 5: Standardize observation procedures and operational rules

• Checkpoint 6: Clarify the handling of height and the verification method

• Approach to Utilizing RTK-GPS Accuracy in the Field

How accurate can RTK-GPS be?

RTK-GPS is a positioning method that can determine positions with substantially higher accuracy than standalone positioning. Typically, errors are on the order of a few centimeters in the horizontal (planar) direction and somewhat larger in the vertical (height) direction, and under favorable conditions it enables high-precision positioning that is fully suitable for practical use. This is a major reason RTK-GPS is highly valued on surveying and construction sites.

However, what matters here is how the phrase "a few centimeters (a few inches)" is interpreted. This is a level that is likely to be expected only when a number of conditions are met: good reception environment, stable correction information, proper initialization, sufficient satellite acquisition, and surrounding conditions with little reflection. Conversely, if those conditions deteriorate, errors can widen from a few centimeters (a few inches) to tens of centimeters (tens of inches). In particular, the vertical direction tends to be less stable than the horizontal plane and is an aspect that often causes unexpected discrepancies on site.

In practice, it is appropriate to understand the accuracy of RTK-GPS not as a mere theoretical value but as a method that, while influenced by field conditions, can provide high reproducibility when used properly. In other words, accuracy is determined not only by the performance of the equipment itself but also by the site environment and operational procedures.

When considering the accuracy of RTK-GPS, its relationship to the work purpose is also important. For example, if you are conducting a wide-area condition survey or obtaining approximate positions, differences of a few centimeters to a few tens of centimeters may not be a major issue. On the other hand, for tasks that require strict precision—such as stakeout, as-built verification, or checking clearances from existing structures—even slight offsets can affect the results. Even with the same RTK-GPS, you must change how you use it and how you verify results according to the accuracy level required for the deliverables.

Therefore, the first step in correctly evaluating RTK-GPS accuracy is not only to consider "how much accuracy can be achieved" but also "under what conditions that accuracy can be achieved" and "whether it is sufficient for the accuracy required at your site." If you introduce it while leaving these points vague, you are likely to be troubled by the gap between the ideal and reality.

How Should RTK-GPS Error Estimates Be Considered?

When trying to understand RTK‑GPS errors, the first point to grasp is that errors manifest differently in the horizontal plane and in height. In many field situations, the horizontal direction tends to be relatively stable, whereas the height (vertical) direction tends to show more variability. This is because they are influenced by factors such as satellite constellation, observation geometry, and the effectiveness of corrections, and on site this often appears as the characteristic “the horizontal aligns but only the height is off.”

Therefore, when checking the margin of error, it is important not to look at a single number alone but to consider planimetric (horizontal) accuracy and vertical (height) accuracy separately. From a practical standpoint, under good conditions planimetry tends to be stable at the centimeter level (half-inch-level), and height should be treated more cautiously. In particular, in situations where differences in height directly affect quality judgments—such as slopes, developed land, road longitudinal profiles, drainage gradients, and areas around structures—handling height with the same mindset used for planimetric measurements is dangerous.

Another important point is that “getting a match once” and “being reproducible whenever you measure” are not the same. RTK-GPS can sometimes produce a good value by chance at a given moment. However, what matters in practice is reproducibility. You must evaluate whether measuring the same point at different times yields nearly the same value, whether measurements by different operators do not produce large differences, and whether observations across multiple days are consistent. A guideline for error should be regarded not as a single positioning result but as an indication of how stable the operation is as a whole.

Furthermore, errors consist of random variations and systematic biases that shift results consistently in one direction. Random variations become partly visible through repeated measurements, but systematic biases are hard to detect and can affect the outcomes across an entire site. For example, if the way references are established, the handling of elevations, the coordinate transformation settings, the input of instrument height, or the placement of observation points is inappropriate, the whole dataset can be offset even if it appears stable. In other words, when considering error estimates, you must check not only the smallness of the numbers but also whether those numbers are referenced to the correct standard.

If you adopt a practical mindset on site, it is useful to treat RTK‑GPS errors as two sets of values: "values under ideal conditions" and "values you should expect in typical field conditions." Although very high accuracy can be expected under ideal circumstances, everyday sites present many disturbances such as trees, vehicles, buildings, slopes, overhead objects, changes in communication conditions, and variations in observer posture. Therefore, when making decisions about deployment and designing operations, you should not assume only favorable conditions but instead plan for errors with some margin. Whether you can make this realistic estimate greatly affects how well you can use RTK‑GPS.

Checkpoint 1: Verify the conditions of the correction information

The first thing to check that determines RTK-GPS accuracy is the condition of the correction information. Since RTK uses reference information to achieve high precision, whether the correction information is being received stably has a large impact on the results. If this is unstable, even if positioning appears to be possible, accuracy and repeatability are likely to vary.

In practice, simply knowing whether correction information has been received is not sufficient. What matters is whether it can be received stably and continuously, whether there are any interruptions, whether reinitializations are occurring frequently, and whether the status is maintained when moving around the site. Even if conditions are good at the starting point, if they deteriorate after moving just a little, it becomes difficult to guarantee the quality across the entire site.

Also, when considering the conditions for correction information, you must not overlook the size of the work area and changes in the environment. Even if there are no problems in open areas, the situation can change near structures, under trees, beside slopes, in mountainous areas, or in places with heavy vehicle traffic. Therefore, rather than simply checking a representative point of the site before starting and feeling reassured, it is important to check the condition at locations where high accuracy is required or where reception conditions are likely to be poor.

A common mistake on site regarding correction information is to assume "it's okay because there are numbers." However, numbers appearing and being stable with high accuracy are not the same. In practical work, in particular, it is necessary to understand status changes during observation, the behavior immediately after a correction drops out, and the time it takes to become stable again. If this awareness is lacking, you may not be able to reconcile coordinates later, and it can take time to determine the cause.

The purpose of verifying the conditions of correction information is not to understand the mechanism in theory, but to determine whether it can truly be operated stably on-site. Rather than ending with a brief pre-start check, observing its behavior in representative on-site environments and identifying in advance the conditions that cause problems will ultimately lead to the greatest improvement in accuracy.

Checkpoint 2: Verify whether the positioning state is truly stable

For RTK-GPS, confirming whether the positioning status is stable is very important. In the field, people sometimes proceed with recording based only on the coordinate values, but you should actually check and judge the positioning status behind them. Even when numerical values are displayed, if the status is unstable you should be cautious about adopting those values.

What is particularly important is whether the solution is sufficiently fixed and remains continuously stable. Immediately after starting observations, right after passing near an obstruction, or immediately after an interruption in corrections, values may appear fine at first glance but may not have settled. If you hastily record data at that time, it can cause a later re-measurement at the same point not to match.

When assessing stability, it is effective to check not only a single reading but also how much the values fluctuate over time. Continue observing the same point for a short while, and if the positional variation is small and the displayed accuracy indicators are stable, you can judge the reliability of the site to some extent. As you become more experienced, you will be able to detect poor conditions from subtle fluctuations in the numbers or slow recovery.

Also, in practice it is effective to decide in advance "which conditions to accept" and "which conditions to re-survey" to reduce differences in judgment among personnel. If left to individual discretion, some staff will accept a measurement while others will re-measure, creating variability. This leads to uneven accuracy across the entire site. RTK-GPS is a high-precision technology, but ultimately quality is largely determined by operational rules.

By carefully verifying stability, you can more easily prevent the impression of a device that "occasionally gives incorrect readings." In fact, there are many cases where large errors are caused not by the device itself but by omitting an assessment of its condition. That is why you should not simply accept positioning results but always make a point of confirming that the values were obtained under stable conditions.

Check 3: Confirm sky visibility and satellite configuration

The accuracy of RTK-GPS is strongly affected by how the sky is visible overhead. This is because, to receive satellite signals stably, it is desirable to have as wide an open sky as possible. Near buildings, under trees, beside elevated structures, in valley terrain, along cut or fill slopes, and in mountainous areas, parts of the satellite constellation can be obscured or reception conditions can become uneven. This is a major cause of reduced accuracy and instability.

A common misconception on site is the idea that "it's fine as long as the sky directly overhead is open." In reality, not only the area directly above but also the visibility of the surrounding sky is important. Because satellites are distributed across various directions in the sky, even if only one direction is obstructed the observation geometry can become biased, which can in turn affect positioning accuracy. The vertical (height) component in particular tends to be more susceptible to this effect.

In areas with poor sky visibility, even when positioning appears to be established, the measured values can slowly drift as the satellite configuration changes, or they may experience a large fluctuation at a particular moment. Because such phenomena are easy to miss with only a brief check, it is effective to wait a little and re-observe, or to approach from a different direction and compare. What is important for field personnel is to grasp, through on-site experience, the difference between favorable locations and challenging ones.

Furthermore, because the satellite constellation changes with the time of day even at the same site, measurements that were stable in the morning can become dispersed in the afternoon. For large sites or work that spans multiple days, you need to be aware of this temporal variation. If accuracy is unusually poor at a particular location, in addition to nearby obstructions, consider the possibility that the satellite geometry at that time is affecting the results.

Checking sky visibility is fundamental to evaluating the accuracy of RTK-GPS. If there are many difficult spots across the site, you need to devise work procedures, secure auxiliary points at locations with good sky visibility, and allocate time to recheck only the critical points. When stabilizing accuracy, it is very important not to overlook sky visibility.

Checkpoint 4: Do not overlook reflections and obstructions caused by the surrounding environment

Reflections and obstructions from the surrounding environment are causes that increase RTK‑GPS errors. This is very likely to occur on-site, yet it is a point that is often overlooked.

In particular, near metal surfaces, glass surfaces, water surfaces, vehicles, temporary structures, fences, handrails, exterior walls, and heavy machinery, signals may arrive not only directly but also by reflection. These reflected signals disturb the true arrival time and cause position variability and offsets.

In practical work, measurements are often taken at the edges of buildings or beside structures. However, these locations are precisely where reflections and obstructions are more likely to coincide, creating conditions in which ideal accuracy is difficult to achieve. What makes things especially troublesome is that not only can the values become clearly erroneous, but sometimes apparently plausible numbers are displayed. If the person responsible adopts them without noticing, they can later become problematic as coordinates that cannot be reproduced.

To reduce the impact of reflections, it is important to first inspect the surroundings and identify "suspicious" locations. Check not only whether the sky is open, but also whether there are nearby reflective surfaces, whether moving vehicles or machinery are close by, and whether temporary structures might affect the observations. Also, simply shifting the observation position slightly can sometimes improve conditions. If that position is absolutely necessary, you should increase reliability by performing repeated observations and comparing results with auxiliary points.

Trees are another factor that should not be overlooked. Leaves and branches can weaken or disrupt satellite signals, so conditions may change between the leaf-on and leaf-off seasons. Even if they do not look like major obstacles, merely obscuring part of the sky overhead can degrade positioning stability. Attention is needed not only in urban areas but also in parks, along rivers, around development sites, and beside roads.

Field operators who are skilled with RTK-GPS pay attention not only to the numbers but also to the surrounding environment. When entering a site, being able to judge "this spot is likely to cause reflections," "the sky view is not clear here," or "this position should be rechecked" greatly reduces accuracy problems. Reflections and obstructions can be unavoidable in some situations, but at the very least you should avoid using the system without noticing them.

Checkpoint 5: Standardize observation procedures and operational rules

The accuracy of RTK-GPS is heavily influenced not only by equipment performance and environmental conditions but also by the consistency of observation procedures. Even at the same site with the same equipment, if procedures differ between operators, the variability of results can increase significantly. In practice, measurements may be stable when taken by one person but show noticeable errors when taken by another. Much of that difference stems from a lack of standardized observation procedures.

For example, small operational differences — such as the waiting time between starting observation and recording, how many times the same point is checked, whether to reinitialize when conditions become disturbed, how strictly the observation posture is maintained, and at what timing the instrument height is checked — accumulate in the results. RTK-GPS is a convenient method that can achieve high-precision positioning in a short time, but precisely because of that, steps that are easy to omit can readily become the cause of reduced accuracy.

On site, it can be tempting to prioritize efficiency by drastically shortening observation times or reducing check points. However, excessive shortening can lead to instability in accuracy. For points that are particularly important or that will affect downstream processes, taking a little extra time to check stability or confirming with round-trip observations will, in the end, help prevent rework. RTK-GPS is attractive for its speed, but pursuing speed alone can compromise quality.

Also, to improve reproducibility, it is useful to have reference checkpoints on-site. By checking the same point at the start, during, and at the end of work, you will more readily notice any shifts or changes in the overall observation system. If work spans multiple days, this also helps confirm consistency with the previous day. Managing such checkpoints may seem like extra effort at first, but considering the risk of having to redo the entire process later, it is well worth it.

Standardizing observation procedures is not about creating rules to restrict field operations. Rather, it is a system to ensure consistent quality no matter who uses it. If you want to operate RTK-GPS reliably on site, it is essential to share decision criteria and procedures instead of relying solely on an operator’s experience and intuition.

Checkpoint 6 Clarify height handling and verification methods

When evaluating the accuracy of RTK-GPS, height is the dimension that requires the most caution. Even if the horizontal (planar) direction is relatively stable, differences tend to appear in the vertical (height) direction, so treating them with the same mindset can lead to misunderstandings. In practice, a position may appear correct, yet there can be inconsistencies when judging slope, longitudinal profile, crown elevation, excavation depth, or embankment height. Many of these problems stem from operating while leaving the handling of height ambiguous.

When it comes to height, the key point is to be clear about what height is required. Depending on whether what you need on site is the ground surface elevation, the top elevation of a structure, or the height relative to the design datum, the method of verification will differ. Simply treating height as just part of the coordinates can cause you to overlook its practical significance. It is important to establish an appropriate definition of height and the method of verification for the required deliverables.

Also, heights can vary depending on the coordinate system and how conversions are handled, so it's necessary to align assumptions among all stakeholders. Even when looking at the same numbers, differences in how the reference is defined can lead to discrepancies in understanding. If this is not shared among the design, construction, and surveying parties, it causes confusion like "it should match but doesn't." Height-related issues tend to take particularly long to resolve on site, so clarifying these matters before work begins is important.

As a verification method, it is fundamental to compare against known points and trusted references, and to be sure to check not only the planar (horizontal) position but also the height. Moreover, rather than doing it only once, rechecking at different times and with different personnel allows for a more practical assessment of accuracy. Because height is difficult to judge by visual inspection on site, it is important to have the verification procedures in place beforehand.

To make RTK-GPS truly usable in the field, do not neglect height. In many practical workflows, height becomes the deciding factor in the final quality assessment. Don’t be complacent just because the horizontal positions line up well; clearly define how you will handle height, how you will verify it, and what tolerances you will accept—this clarity is the quickest way to prevent accuracy problems.

How to Leverage RTK-GPS Accuracy on Site

The accuracy of RTK-GPS can be extremely high when conditions are favorable, and it is at a level that can be fully utilized in practical work. However, that accuracy is not something that will be obtained automatically simply by using the equipment. Correction information, positioning status, sky visibility, reflections and obstructions, observation procedures, and the handling of height must each be checked, and only by operating in a manner suited to the site will stable results be achieved.

What those searching for "RTK GPS accuracy" especially need to know is that accuracy is not merely a specification value but the quality of operation itself. On-site, good and bad conditions coexist, and results can change by location and time even on the same day. To ensure the required accuracy in that context, it is important to be able to judge where you should check carefully and under which conditions it can be accepted. RTK-GPS is a convenient technology, but if used incorrectly it can lead to complaints like "it deviates more than expected." Conversely, if you understand the conditions and handle them appropriately, it becomes easier to achieve both efficiency and quality in your work.

When considering implementation or reassessment, we recommend prioritizing how reproducible the results are on your own site rather than comparing catalog figures alone. The required operation will vary depending on the environment—large development sites, roads, areas around structures, sites with many trees, or narrow construction yards. What matters is not just talk of ideal accuracy, but whether it can be used reliably and comfortably at your own site.

In that sense, if you want to make everyday construction management, position checks, and site records more accessible with high precision, it is important to consider equipment and methods including ease of operation. For example, iPhone-mounted GNSS high-precision positioning devices like LRTK make it easier to incorporate high-precision positioning in a form that is convenient to handle on site, and they also make it easier to verify positioning results and link them with records. If you want to put RTK-GPS accuracy to use not just as desk-derived numbers but in actual field work, it is important to consider not only accuracy but also whether the solution will integrate smoothly into daily operations.