How to Do a Base Shift in RTK GNSS: ApekSurv Guide
A base shift (base station shift calibration) is a single-point translation applied to the RTK correction stream to align the rover's output coordinates with a known control point on the project. When a base station is restarted, moved, or set up in a new position between field sessions, its internal autonomous position shifts slightly — even if it is placed on the same physical monument. This shift propagates into every rover coordinate recorded in that session. The base shift corrects this by occupying one known control point with the rover and forcing the correction stream to match the published coordinate. It takes under five minutes. Without it, data from different sessions on the same project will not align.
The base shift is one of the most underused procedures in RTK survey — and one of the most consequential when skipped. A base station restarted on the same physical monument after lunch will typically initialize 10–50mm away from its morning position. Every point recorded that afternoon carries that offset. On a multi-day project, data from Tuesday and Wednesday will not match Monday's coordinates unless a base shift is applied at the start of each session. This guide explains why the shift occurs, when to apply it, the step-by-step procedure in ApekSurv, and the most common mistakes that cause a shift to fail.
Why the Base Shifts Between Sessions
When an RTK base station is deployed in autonomous mode, it computes its position by averaging satellite observations over a brief initialization window. Because it lacks a known, absolute geodetic coordinate to reference, the receiver calculates its position based on standard standalone GNSS positioning. This standalone position is subject to dynamic atmospheric variables, satellite clock drift, orbital ephemeris errors, and changing satellite geometry (Dilution of Precision). Consequently, the autonomous position computed during setup A will differ randomly from the autonomous position computed during setup B.
The GNSS autonomous position has ±3–5m of absolute random error. However, when a high-precision receiver is restarted on the exact same physical monument, the relative session-to-session drift typically manifests as a 10mm to 50mm vector offset. Because RTK baseline processing calculates the relative vector between the base and the rover, any coordinate displacement at the base station propagates directly into the rover's measurements. The result: all rover coordinates from Session A are internally consistent, and all from Session B are internally consistent — but Sessions A and B are offset from each other by the difference between the two autonomous base positions.
This structural offset introduces significant downstream errors in multi-session engineering environments:
- Multi-day topographic surveys: Terrain models collected over several days will exhibit artificial steps or vertical discrepancies where data edges merge across sessions.
- Construction layouts: Layout stakes or formwork set out on different days will fail to align with the same structural design control.
- Volume calculations: Stockpile or earthwork volume metrics become distorted when comparing surfaces derived from separate, unaligned base setups.
- Any project: Where the base station is power-cycled, moved, or restarted during the active campaign.
This translational error is entirely averted under specific setup configurations. If the base station is initialized using a known coordinate entered manually, it broadcasts from an identical reference origin every time, achieving perfect baseline repeatability. Similarly, network operations utilizing a CORS or NTRIP correction source do not experience session-to-session base shift issues because the network reference infrastructure occupies a fixed, permanently monitored geodetic position.
Base Shift vs Point Calibration — Key Differences
These two software-driven field procedures are frequently confused by operators. They address fundamentally different mathematical problems within spatial coordinate geometry and cannot be substituted for one another without risking data distortion.
| Feature | Base Shift | Point Calibration |
|---|---|---|
| Purpose | Aligns the current receiver session to a single known control point via translation vectors only. | Fits global GNSS ellipsoidal coordinates to a localized arbitrary grid system using translation, rotation, and scale parameters. |
| Control points needed | 1 minimum | 3 minimum for horizontal alignment; 4 minimum for horizontal and vertical alignment. |
| When to use | Each new base session on the same project when using autonomous initialization. | Executed once per project when the local engineering grid does not match the national grid projection. |
| Affects | Current session correction stream and real-time positioning offsets only. | All historical and future sessions that utilize the specific calibration localization file. |
| Corrects | Session-to-session baseline offsets caused by autonomous base station reinitialization. | Systematic datum distortions, projection scale variations, and angular rotations relative to true north. |
| Time required | Under 5 minutes | 20 to 40 minutes to occupy and resolve widely distributed control benchmarks. |
Use an RTK base shift calibration to maintain session consistency on a project that has already been configured with a correct localization or projection framework. Conversely, execute a comprehensive point calibration when the project specifications mandate working within a local arbitrary coordinate system that differs from the GNSS output in rotation, scale, or origin.
When You Must Do a Base Shift
To safeguard structural accuracy across multi-day field operations, a systematic checklist must dictate when to apply a transformation vector. The determination rests entirely on the operational continuity of the base infrastructure.
YOU MUST DO A BASE SHIFT WHEN:
- The base station was shut down, undergone a battery swap, or experienced an involuntary power cycle since the previous survey session concluded.
- The physical base station was moved to a new position (even temporarily to clear site traffic) and subsequently returned to the original monument.
- A different physical base station hardware unit is deployed for today's session than was used during the previous campaign day.
- Any rover data from the current session must align precisely with topographic features, boundaries, or structural elements recorded in a previous session.
YOU DO NOT NEED A BASE SHIFT WHEN:
- The base station has been running continuously since the previous session, maintaining the same power-on status and satellite tracking initialization.
- The base station is set up with a known coordinate entered manually, forcing it to broadcast corrections from a locked reference point rather than an autonomous average.
- You are using CORS/NTRIP as the correction source, meaning no independent physical base station is deployed on site.
- This is the first session on a new project, and there is no prior data or established local reference file to align with.
How to Do a Base Shift in ApekSurv — Step by Step
Execute the following procedural sequence within the ApekSurv field software to calculate and apply an accurate base shift translation vector.
Start the base station and rover normally: Set up your physical base station hardware (AP10, AP20, or the high-constellation MAX5) over your chosen site monument in autonomous mode. Power on the unit and allow it to initialize its position. Connect your rover receiver via the internal UHF radio or network link, and verify that the data link has secured a stable Fixed solution status. At this stage, because the base position is derived from an autonomous average, the rover's real-time coordinate display will exhibit a baseline translation error when held over a known project monument. The base shift calibration will eliminate this offset.
Navigate to the calibration function: Open your primary data controller and launch the ApekSurv application interface. From the primary menu tree, navigate to Survey and select the Calibrate Point (Base Shift) utility. This internal module isolates the active baseline coordinates, allowing you to apply a single-point translation across the active correction stream without altering the underlying local projection parameters.
Select or enter the control point coordinate: Identify a verified, stable control monument from your master project network. Within the ApekSurv calibration screen, you must either enter the published local grid coordinates (Easting, Northing, Elevation) manually or import the point record directly from the pre-existing project library. Crucially, do not select a temporary point measured or staked within the uncalibrated current session, as its position contains the very baseline error you are attempting to correct.
Occupy the control point and record: Advance your rover to the physical control monument. Position the rover pole precisely over the central index point. Adjust the bipod or hand-hold to centre the circular bubble vial precisely, ensuring the pole is perfectly vertical. Within the ApekSurv interface, select the Collect command to initiate an epoch averaging sequence over the monument. The software records the uncorrected GNSS coordinate stream and computes the spatial delta against the stored, published control coordinates. This mathematical variance defines the absolute base shift vector ($dX, dY, dZ$) that will govern the active session.
Apply the shift and verify: Review the calculated translation values ($dX, dY, dZ$) displayed on the confirmation page before saving. Under normal operating conditions on an identical physical monument, these values should range between 10mm and 50mm. If the screen indicates an exceptional delta exceeding 100mm, abort the routine to audit your entry data and pole heights. If the vector is normal, tap Apply. Immediately walk the rover configuration to an independent secondary control monument to perform a check measurement. This verification point must read within ±10mm horizontally and vertically of its long-term published value before beginning production operations.
The Core Problems in Base Shift Operations
Symptom: The ApekSurv software indicates a successful calibration routine, and your primary control point reports zero residual error when measured. However, when the daily field data export is imported into office CAD software and overlaid against previous session maps, a consistent horizontal or vertical alignment gap remains obvious across all topographic features.
Cause: This error occurs when the operator selects a different physical control point for the base shift routine than the one utilized to anchor preceding sessions, or when historical data layers were collected using uncorrected base initializations that were never documented. Using different calibration origins across separate site visits creates conflicting, parallel coordinate reference planes that fail to resolve into a single master framework.
Fix: Enforce a strict site protocol requiring that one primary control point be designated as the sole master base shift reference point for the entire project lifespan. If historical data files are found to contain internal offset variances due to inconsistent reference point selections, you must apply a retrospective post-processing translation within your desktop office suite to align the legacy vector layers with the chosen master project framework.
Symptom: Following the acquisition of the reference point position within ApekSurv, the software calculation screen outputs abnormally large translation metrics, showing linear coordinate shifts such as $dX = 0.45\text{m}$ or $dY = 0.31\text{m}$, which greatly exceed standard initialization tolerances.
Cause: This discrepancy is caused by a physical setup mistake, such as placing the base receiver over the wrong site monument, selecting the wrong point record within the ApekSurv library, or initiating the base unit under severe overhead obstructions that limit tracking to a weak, multi-path-distorted satellite constellation during its autonomous positioning phase.
Fix: Re-verify the physical location of the base station tripod setup to confirm it matches the intended monument. Check the receiver tracking screen to ensure the unit is tracking at least 20 satellites across multiple frequencies. If tracking performance is poor, clear the current autonomous initialization, allow the base receiver 5 to 10 minutes to compile a clean satellite ephemeris profile, and then restart the rover calibration step.
Symptom: The primary control point used for the base shift calibration evaluates correctly with zero residual error. However, when the rover subsequently occupies an independent secondary validation monument, the resulting coordinate readout displays errors of 40mm to 80mm, missing your site tolerance limits.
Cause: A base shift applies a rigid, localized 3D translation vector to your coordinate system. It cannot modify scale factors or correct for angular rotations. If the site layout is built on a localized grid projection that scales or rotates away from true geodetic north, a single-point translation will fail to resolve these structural projection distortions across longer baselines.
Fix: Abandon the simple single-point base shift approach for this specific project layout. Switch your workflow to a full multi-point point calibration routine utilizing 3 to 4 widely distributed control benchmarks across the site boundaries. This allows ApekSurv to compute a comprehensive transformation model that handles translation, angular rotation, and scale variations simultaneously. Review the detailed point calibration guide for complete instructions.
Pro Tips for Base Shift Accuracy
- Designate One Primary Shift Reference: Avoid switching between different site monuments for calibration routines. Select a primary control benchmark at project inception, document its physical location with clear site photos, and ensure all field crews utilize this exact point for every subsequent base shift.
- Verify Rover Pole Height Configuration: Entering an incorrect pole height into your data controller introduces a direct, systematic vertical error into your transformation vector. Double-check your physical pole pin settings and lock the exact value into ApekSurv before initiating data capture routines.
- Always Validate via an Independent Check Point: A single mathematical point cannot validate its own spatial integrity. Always occupy a secondary independent control monument immediately following a base shift calibration to confirm your project alignment.
- Calibrate At the Outset of Every Field Session: Never delay the calibration routine until the end of the day or attempt to resolve shifts retroactively in the office. Execute the base shift calibration immediately after establishing your base station link, before capturing any production survey data.
- Log and Monitor Historic Shift Values: Keep a clear record of the calculated $dX, dY, dZ$ vectors in your daily project field notes. Sudden shifts or shifting trends across identical setups often highlight structural disturbance of your reference control monument.
FAQ
Can I do a base shift using CORS instead of a base station?
No. A base shift is designed specifically to resolve spatial errors introduced by the autonomous initialization of an on-site physical base station receiver. CORS network setups are managed by infrastructure operators who secure their permanent reference antennas to highly stable, mathematically defined geodetic coordinates. Because CORS setups do not use arbitrary autonomous initializations, session-to-session base drift does not occur. If you notice structural coordinate offsets while connected to a CORS network, the issue stems from an incorrect project datum, coordinate system configuration, or geoid model selection inside ApekSurv rather than a base station shift error.
How often should I do a base shift?
You must execute a base shift calibration every single time your physical base station undergoes a power cycle, system reboot, or physical relocation. If your base receiver remains powered on and maintains uninterrupted constellation tracking throughout your mid-day lunch break, its internal reference positioning profile remains unchanged, meaning an afternoon base shift calibration is unnecessary. As a firm operational rule: schedule and complete one base shift calibration per physical power cycle of your base station hardware. For mobile field crews running daily setups on multi-day site campaigns, this requires a minimum of one calibration check every morning.
Does the base shift affect raw GNSS data stored on the receiver?
No. The base shift routine applies a real-time mathematical translation vector exclusively to the incoming RTK correction stream and the calculated coordinates displayed within your active ApekSurv data controller database. The raw GNSS carrier-phase and pseudorange observations recorded internally within the onboard storage memory of your receiver remain unaffected. If you need to correct data that was mistakenly captured without a base shift, you will need to apply a post-processing coordinate transformation or translation step inside your office processing software, as the real-time field calibration tool does not retroactively modify raw satellite data layers.
What is the difference between base shift and point calibration?
A base shift calculates a simple, localized 3D translation vector ($dX, dY, dZ$) to reconcile minor session-to-session initialization drift on projects that already possess an established coordinate system. It uses a single known control point. Conversely, a Point calibration is a multi-point mathematical process that builds a full 2D or 3D similarity transformation model, altering spatial parameters across translation, rotation, and scale factor variables. Use the base shift option for routine session validation on correctly configured projects, and choose point calibration when you need to adapt your global GNSS data to fit an unmapped local site coordinate system.
CONSISTENT COORDINATES. EVERY SESSION.
ApekSurv base shift takes under five minutes and keeps every survey session aligned to the same project control. Works with AP10, AP20, and MAX5 base station configurations.
View APEKS RTK Receivers →References
- ISO 17123-8:2015 — Field Procedures for GNSS RTK
- ApekSurv Field Software User Guide, 2026
- APEKS AP20 Technical Datasheet, 2026
- APEKS MAX5 Base Station Technical Datasheet, 2026
- Unicore Communications UM980 Product Brief