Last Updated: November 4, 2026
Quick Summary
To perform post-processing for GPS data, import your RINEX observation and navigation files into processing software, select a nearby base station, run baseline processing, and review residuals before exporting corrected coordinates. Learning how to perform post-processing for GPS data starts with clean field collection, since no software rescues bad raw data.
- Online GNSS post-processing typically takes 10 to 20 minutes, per a 2026 Swepos study.
- Set the coordinate system and datum before importing files, or reprocess from scratch.
- Static processing suits stationary receivers logging 20 minutes or more.
- RTKLIB expects antenna height in the RINEX header, not a separate field.
- When a job fails, check file completeness, then antenna height, then baseline length.
Table of Contents
- What GPS Post-Processing Does and When You Need It
- Best Practices for GPS Data Collection Before You Process
- Step-by-Step: Processing Raw GNSS Data
- Static vs Kinematic GPS Processing: Choosing the Right Method
- GNSS Post-Processing Software: Free, Paid and Open-Source Options
- Troubleshooting Common Post-Processing Failures
- From Corrected Coordinates to Field Reports
- Frequently Asked Questions
What GPS Post-Processing Does and When You Need It
According to a 2026 study of the Swepos online GNSS post-processing service, processing times for online post-processing typically run 10 to 20 minutes. That wait buys you something field equipment alone cannot: a corrected position.
You need it in three situations:
- Your rover logged data in an area with no real-time correction signal
- You need survey-grade results from a lower-cost receiver
- You want a defensible record of accuracy for compliance or client reports
Online services can return results in as little as 15 seconds after upload, according to a 2026 GPS World comparison of free post-processing services. But speed is not the same as accuracy.
Best Practices for GPS Data Collection Before You Process
Here is the part most guides skip: no amount of post-processing rescues bad field data. Research published by Taylor & Francis found that pre-processing steps like removing outliers and de-noising are critical prerequisites for effective GPS post-processing.
Log longer than you think you need, so the processing engine has enough data to resolve ambiguities cleanly.
Other rules we would enforce on any crew:
- Keep a clear sky view; trees and buildings cause multipath error that no software fully fixes
- Record the antenna height precisely, in the units your software expects
- Note the observation interval and any receiver settings in your field log
For a deeper look at capturing clean field data, see our guide on improving field survey data accuracy.
Observation and navigation files: what your receiver records
Your receiver writes two file types. The observation file holds the raw satellite measurements: pseudorange, carrier phase, and signal strength. The navigation file holds the ephemeris, which tells software where each satellite was in orbit.
Both matter. Without a valid navigation file, baseline processing cannot compute satellite positions. Most modern receivers export both as RINEX, the standard exchange format.
Step-by-Step: Processing Raw GNSS Data

The workflow is consistent across software: import, select a base, process, review, export. Here is how to perform post-processing for GPS data in five practical steps.
- Create a new project and set the correct coordinate system and datum
- Import your RINEX observation and navigation files
- Choose a base station near your rover
- Run baseline processing and check that ambiguity resolution succeeds
- Review residuals, then export corrected coordinates
Setting up your project and importing RINEX files
Set the coordinate system before you import anything. Changing datums after processing means reprocessing from scratch. Confirm the RINEX version matches what your software supports, and check that the file header lists the correct antenna height.
Base station selection, baseline processing and ambiguity resolution
Base station choice drives your final accuracy. A nearby station with a clean sky view generally outperforms a distant one. Closer baselines mean smaller atmospheric delay errors and better odds of fixing integer ambiguities.
Watch the residual values after processing. Large residuals usually point to a bad base station, a wrong antenna height, or cycle slips in the carrier phase data.
Static vs Kinematic GPS Processing: Choosing the Right Method
Static processing suits a receiver that sits still and logs for 20 minutes or more. Kinematic processing suits a moving rover, and it demands a base station close by and a reliable radio or logging link.
| Method | Receiver State | Typical Occupation | Best For |
|---|---|---|---|
| Static | Stationary | 20+ minutes | Control points, monuments |
| Fast static | Stationary | 5-20 minutes | Dense survey networks |
| Kinematic | Moving | Continuous | Drone flights, vehicle tracking |
| PPP | Either | Varies | Remote sites with no nearby base |
If you have no base station within range, PPP (Precise Point Positioning) uses global orbit and clock products instead. It is slower to converge but needs no local reference.
GNSS Post-Processing Software: Free, Paid and Open-Source Options
The right GNSS post-processing software depends on your file formats, your volume of work, and whether you need a support contract. Choosing software around file formats and workflow also means keeping charting tools current, so the same maintenance discipline applies when you update marine GPS software.
The field breaks into four practical categories:
- Free online services. You upload RINEX files to a web portal and download corrected coordinates. These handle simple single-baseline jobs well and require no installation. They are a poor fit for batch work, custom processing settings, or data you cannot send to a third party.
- Open-source desktop tools. RTKLIB is the reference implementation: it runs on Windows, macOS, and Linux, reads RINEX 2 and 3, and exposes every setting from elevation mask to ambiguity resolution thresholds. The learning curve is real, but nothing else gives you this much control for free.
- Commercial survey suites. These add batch processing, automated reporting, geoid models, and vendor support. They earn their licence cost when you process hundreds of points a week or need a defensible audit trail for a client.
Our advice: start with a free online service to confirm your raw data is sound, then move to RTKLIB if you need control or volume. Only pay for a commercial suite when support, reporting, or batch throughput becomes the bottleneck.
A vendor-neutral RTKLIB workflow
RTKLIB ships as two programs you will use together: RTKPOST for baseline processing and RTKPLOT for visual review. The workflow below assumes you have a rover observation file, a rover navigation file, and access to a reference station’s RINEX data.
- Gather four files. You need the rover’s observation and navigation files, plus the base station’s observation and navigation files. If the base is a continuously operating reference station, download its RINEX for the same time window as your rover occupation.
- Open RTKPOST and set the mode. Choose
Staticfor a stationary receiver orKinematicfor a moving one. Set the positioning mode toKinematicfor rovers andStaticfor control points; leave the frequency setting onL1+L2if your receiver logged both. - Load the files. Point the rover and base fields at the correct observation files, and the navigation fields at the matching ephemeris files. Confirm the time span covers your occupation.
- Set the elevation mask. A mask of 10 to 15 degrees is a common starting point; it drops low satellites whose signals are most affected by atmospheric delay and multipath.
- Choose the ambiguity resolution setting.
Continuousworks for most static jobs;Instantaneoussuits kinematic work where the rover is moving. If ambiguities will not fix, switch toFix and Holdand extend the occupation. - Run the process and open the plot. RTKPLOT shows residuals, satellite visibility, and the fix/float status over time. A clean run shows a stable fix and low residuals after convergence.
- Export. Write out the corrected coordinates in the datum and projection your GIS expects, and keep the processing report alongside the raw files.
Watch out: RTKLIB expects the antenna height in the RINEX header, not in a separate field. If you entered the height only in your field log, edit the header before processing or every coordinate will be offset by that amount.
Matching the tool to the job
A single point collected for a report rarely justifies a commercial licence. A weekly drone mapping program with dozens of images per flight needs batch processing and a repeatable settings file. A team that must defend coordinates to a regulator needs the audit trail a commercial suite produces. Decide which of those three you are before you compare prices, because the cheapest tool that does the job is usually the right one.
Troubleshooting Common Post-Processing Failures
Post-processing failures happen, and knowing which symptom points to which cause helps you resolve them faster. This section is a diagnostic guide for the failures you will actually hit.
Work through the symptoms in the order below. Each one narrows the problem before you change settings.
Ambiguities will not fix
Symptom: the processing report shows a float solution throughout, or the fix status flickers between fix and float.
Likely causes and fixes:
- The baseline is too long. Atmospheric delay decorrelates with distance; move to a closer reference station, ideally within 10 to 20 km.
- The occupation is too short. Extend a static occupation to 20 minutes or more so the engine has enough epochs to resolve integers.
- The elevation mask is too low. Raise it to 10 to 15 degrees to drop noisy low satellites.
Large residuals after processing
Symptom: the solution fixes, but residuals are visibly higher than normal, or the plot shows spikes at specific times.
Likely causes and fixes:
- Wrong antenna height. This is the single most common cause of a consistent offset. Check the RINEX header against your field log.
- Cycle slips in the carrier phase. RTKPLOT will show them as jumps in the phase residual; if they cluster, the rover may have passed under trees or a bridge.
- Multipath. Signals bouncing off buildings or water produce oscillating residuals. Reprocess with a higher elevation mask, or discard the affected epochs.
No navigation data or missing ephemeris
Symptom: the software refuses to process, or reports that satellite positions cannot be computed.
Likely causes and fixes:
- The navigation file was not imported alongside the observation file. Confirm both are loaded.
- The navigation file covers a different time window than the observation file. Download the correct day’s ephemeris.
- The receiver exported a proprietary format instead of RINEX. Convert it before processing.
Coordinates are wrong by a consistent offset
Symptom: every point is shifted by the same distance and direction.
Likely causes and fixes:
- Datum or projection mismatch. Confirm the project coordinate system matches the one your reference station publishes.
- Antenna height entered in the wrong units. A height logged in feet but read as meters produces a large, consistent vertical error.
- A geoid model was not applied where one is required. Check whether your software expects ellipsoidal or orthometric heights.
Loss of lock during collection
Symptom: the processing plot shows gaps in satellite tracking, often around trees, buildings, or under a bridge.
Likely causes and fixes:
- The rover passed through an obstructed area. Reprocess excluding the affected epochs, or re-observe the point with a clearer sky view.
- The receiver lost power or the logging interval changed mid-session. Check the observation file for time gaps.
- A low-grade antenna struggled with weak signals. This is a hardware limit; post-processing cannot recover data the receiver never recorded.
One honest limitation: post-processing cannot compensate for every hardware limitation, as Juniper Systems notes. A low-grade antenna still produces low-grade results, and no setting change will fix a receiver that never locked on.
A diagnostic order that saves time
When a job fails, check in this order: file completeness first, then antenna height, then baseline length, then processing settings. Most failures are resolved by the first two checks, and changing settings before confirming your inputs is the fastest way to waste an afternoon.
From Corrected Coordinates to Field Reports
Corrected coordinates are only useful once they land in a report someone can read. Confirm your datum and projection first; our guide to understanding coordinate systems for Canadian field work covers the common pitfalls.
PhotoLog handles that handoff. For teams digitizing paper notes, our guide on digitizing field notes and photos walks through the workflow, and field documentation best practices covers naming and version control.
Frequently Asked Questions
What is the purpose of GPS post-processing?
Post-processing corrects the positions your receiver recorded in the field by comparing them against a known base station or precise satellite orbit data. This removes much of the error caused by atmospheric delay, satellite clock offsets and multipath. The result is a corrected coordinate that is more accurate than the raw field fix, often improving from several metres to sub-metre or better depending on your equipment and observation time.
Do I need specialized software to post-process GPS data?
You need software that reads RINEX observation and navigation files and runs baseline processing. Options range from free online services that return results in 15 seconds to 20 minutes, to open-source desktop tools, to paid GNSS post-processing software with full project management. For occasional checks a free service works; for repeat survey work, a desktop tool that stores baselines and coordinate systems saves time.
What is the difference between real-time and post-processed GNSS data?
Real-time corrections (RTK) arrive over a radio or cellular link while you survey, so you see corrected positions instantly but depend on a live connection. Post-processing happens after the fact, using logged raw data and a base station or precise ephemeris. Post-processing works where cellular coverage fails and can apply better corrections, but you only learn the final accuracy once processing finishes.
How does post-processing improve the accuracy of field photos?
Corrected coordinates can be tied back to the timestamp on each photo, so geotagged images carry positions that match your survey rather than the raw receiver fix. In PhotoLog, automatic GPS geotagging and date/time stamping mean each image is searchable by session key and exportable in a formatted Field Report.
Corrected coordinates are only half the record. The other half is proof: dated, located images that show what you saw when you stood there. PhotoLog turns your Android camera into a field documentation tool with automatic GPS geotagging, switchable privacy control, and built-in Field Reports you can export and share. Get started with PhotoLog and turn every site visit into a report you can hand to a client without rework.