Free GNSS post-processing: online services and software compared

Which free option fits your data?

Start with what you recorded. A single static session on a tripod suits a precise point positioning service or a network solution. A moving rover or a drone needs kinematic processing, which narrows the options. Service rules change, so treat the table as a starting point and confirm the current requirements on each official page.

Technical references: Natural Resources Canada: CSRS-PPP service; Geoscience Australia: AUSPOS online GPS processing service; NOAA NGS: about OPUS; RTKLIB: open-source program package for GNSS positioning; Emlid: Emlid Studio documentation.

How do the online services differ?

CSRS-PPP does not need a nearby base station. It combines your observations with precise satellite orbit and clock products, so it works anywhere in the world. Accuracy improves with session length and with the quality of the products available for that date, which is why a result processed a few weeks after the survey can differ slightly from one processed the next day.

AUSPOS and OPUS compute your point relative to permanent reference stations. AUSPOS draws on the global IGS network and states that it works with data collected anywhere on Earth. It requires at least one hour of static dual-frequency GPS observations per file and notes that centimetre-level estimates are more likely with six or more hours. OPUS states that files must contain static dual-frequency GPS observations of between 15 minutes and 48 hours, and it is built around the US reference network; its national coordinates apply to the United States. For a site in India, check the report to see which stations were used and how far away they are.

• Enter the antenna type exactly as the service lists it, and the vertical height to the antenna reference point.
• Upload one continuous static session per point unless the service says it accepts kinematic data.
• Keep the full report, not only the coordinate line. It records the frame, epoch, products and quality statistics.

Technical references: Natural Resources Canada: CSRS-PPP service; Geoscience Australia: AUSPOS online GPS processing service; NOAA NGS: about OPUS.

When is desktop software the better choice?

Desktop processing suits kinematic work, short baselines to your own base, and data you are not permitted to upload. RTKLIB is an open-source program package for standard and precise positioning. It supports several constellations and positioning modes, and exposes most processing settings. That flexibility is useful for learning and for unusual datasets, and it also means two operators can produce different answers from the same files.

Emlid Studio is a free application with a guided workflow for static, kinematic and stop-and-go processing and for updating drone image positions. It reads RINEX, so it is not limited to one manufacturer's receivers. With either tool, save the configuration alongside the output so that the solution can be reproduced.

Technical references: RTKLIB: open-source program package for GNSS positioning; Emlid: Emlid Studio documentation.

What should users in India check?

The global services report coordinates in an ITRF realisation at the observation epoch, or in another country's national frame. An Indian project usually specifies its own reference, so agree the frame, epoch and height system first and document any transformation that follows. Ellipsoidal heights from a processing report are not orthometric heights; a geoid model or levelling connection is a separate step.

Survey of India operates a national CORS network and portal with registered-user services. If your work must tie to that network, review the portal's current services, data products and terms before choosing a workflow. Our CORS guide describes registration, subscription and rover setup.

Technical references: Survey of India: CORS portal; Natural Resources Canada: CSRS-PPP service.

What should you check before you upload or process?

Most rejected or poor submissions trace back to the file rather than the service. A short review of the RINEX header and a quality-check summary saves a repeat field visit.

• The file is valid RINEX in a version the service accepts, with both frequencies present.
• The session meets the minimum length and does not break the service's rules on day boundaries or sampling interval.
• The antenna model and height in the header match the field sheet.
• The data is permitted to leave your organisation. Some project data should not be uploaded to a third-party service.
• You know which frame and epoch the report will use, and what the project requires.

Technical references: NOAA NGS: about OPUS; Geoscience Australia: AUSPOS online GPS processing service.

How do you know the result is good enough?

Reported standard deviations describe the internal consistency of one solution. We would treat agreement between independent methods, repeat sessions and checks on existing control as separate lines of evidence, and report each one with its limits.

Common questions