RTCM Correction Data in RTK GNSS Surveying

Surveyor in a high-visibility vest checking a data collector next to a GNSS rover pole

Two receivers standing ten metres apart can be looking at the same satellite and disagreeing about where it is by several metres. Not because either receiver is faulty, but because the raw signal carries baggage: atmospheric delay, orbital drift, clock error, all riding along with the data every GNSS receiver depends on.

RTCM is the format that lets a base station tell a rover exactly what that baggage looks like, in a language both machines can parse in real time. Without it, RTK doesn’t work. Every centimetre-accurate fix a survey crew pulls in the field traces back to a stream of RTCM messages doing quiet, unglamorous work in the background.

What Is RTCM?
Why RTCM Data Exists
How RTCM Corrections Move From Base to Rover
Single-Baseline vs. Network RTCM Delivery
What Degrades RTCM Correction Quality
Choosing a Correction Source in Canada
RTCM Compatibility: What to Check Before You Buy
The Correction Nobody Sees, But Every Fix Depends On

What Is RTCM?

RTCM stands for Radio Technical Commission for Maritime Services, the organization that defined the messaging standard used to encode GNSS correction data. In practice, “RTCM” refers to the message format itself, specifically RTCM 3.x, the version nearly every modern RTK system runs on.

An RTCM message isn’t a position. It’s a structured data packet describing carrier-phase observations, satellite identifiers, and reference station coordinates, formatted so any compliant receiver can decode it regardless of manufacturer. That standardization is what makes RTK interoperable across brands. A rover from one manufacturer can pull corrections generated by a completely different manufacturer’s base station, because both sides speak RTCM.

Each RTCM message type serves a specific role. Observation messages carry the raw carrier-phase and code-phase data for a given constellation, GPS, GLONASS, Galileo, and BeiDou each get their own message types. Station coordinate messages tell the rover exactly where the base sits, which is what lets the rover calculate the correction delta in the first place. A modern multi-constellation setup is typically streaming several message types simultaneously, bundled into one correction feed.

Why RTCM Data Exists

GNSS satellites broadcast timing signals, and receivers use those signals to calculate distance to each satellite. That calculation is affected by a handful of predictable error sources:

  • Ionospheric delay. Charged particles high in the atmosphere slow the signal, and the effect varies with solar activity and time of day.
  • Tropospheric delay. Water vapour and temperature shifts in the lower atmosphere add further delay to the signal path.
  • Satellite orbit and clock drift. Even with regular updates, satellite position and onboard timing carry small residual errors.
  • Multipath. Signals bouncing off buildings, vehicles, or wet ground before reaching the antenna create false distance readings.

These errors aren’t random from one receiver to the next. Two receivers sitting reasonably close together experience nearly identical atmospheric and orbital error at any given moment, because they’re looking at the same satellites through the same slice of sky. A base station at a known location isolates that shared error and packages it as an RTCM correction. The rover applies it, and the errors that would otherwise throw off both receivers cancel out.

How RTCM Corrections Move From Base to Rover

The path from raw satellite signal to a fixed centimetre position follows a consistent sequence:

  1. The base station, sitting on a precisely known point, continuously logs carrier-phase observations from every satellite it can see.
  2. It compares what it’s observing against what it should be observing and isolates the difference.
  3. That difference gets packaged into RTCM messages.
  4. The messages are transmitted to the rover, most commonly over NTRIP (Networked Transport of RTCM via Internet Protocol), though radio link and satellite delivery are used where cellular coverage isn’t available.
  5. The rover combines the incoming RTCM data with its own observations, resolves the carrier-phase ambiguity, and lands on a centimetre-accurate position.

NTRIP is the transport layer that gets RTCM data from a base station to a field receiver over the internet. Bench-Mark equipment relies on standard NTRIP connections to pull correction streams from Canadian network providers, so a properly configured RTK GPS and GNSS receiver is really just an RTCM decoder with a very good antenna attached.

Surveyor holding a GNSS rover pole and data collector on a residential street

Single-Baseline vs. Network RTCM Delivery

Not every RTCM stream comes from the same kind of source, and the difference affects both range and reliability.

Single-baseline RTK pulls RTCM data from one physical base station. It’s straightforward and effective within roughly 30 km, but accuracy erodes with distance, since atmospheric conditions at the base start to diverge from what the rover is experiencing.

Network RTK, sometimes delivered as a Virtual Reference Station (VRS), draws on multiple interconnected base stations across a region. The network models atmospheric conditions spatially and synthesizes an RTCM stream tailored to the rover’s approximate location, rather than relying on one fixed point. This keeps the effective distance short even when the nearest physical station is far away, which matters across the kind of geography Canadian crews regularly work in.

The density of the underlying station network determines how well this works. A sparse network produces coarser atmospheric modelling and slower convergence to a fixed solution. A dense one gets a rover locked in faster and holds it there more reliably.

What Degrades RTCM Correction Quality

An RTK receiver pulling in a clean RTCM stream and one struggling to hold a fix aren’t always separated by hardware quality. A few variables determine how well the correction data performs:

  • Correction age. RTCM data is time-sensitive. Most receivers can still use corrections up to roughly 60 seconds old, but quality drops as latency climbs. Corrections arriving within a second or two of being generated give the cleanest results.
  • Baseline distance. Accuracy degrades by roughly 1 to 1.5 cm for every 10 km between the rover and the reference source, which is the practical argument for network corrections over a single distant base.
  • Station calibration. A base station with a properly calibrated antenna and a stable, well-monumented mount produces cleaner corrections than one assembled with inconsistent hardware.
  • Constellation coverage. An RTCM stream carrying observations from all four major constellations gives the rover more satellites to resolve against, which speeds convergence and improves performance under partial sky obstruction.

A crew troubleshooting a shaky fix on an otherwise clean site is usually running into one of these, not a defective receiver. That’s a common diagnostic conversation at Bench-Mark: the equipment is fine, and the correction stream feeding it is the variable worth checking.

Choosing a Correction Source in Canada

Canadian surveyors generally have three practical ways to get RTCM data into a rover: a private single base station set up on a known point for a job, useful on remote sites without network coverage; a subscription network RTK service, which handles the base infrastructure and delivers a modelled correction stream over NTRIP, the most common approach for production survey and construction work; or a public CORS-based source, continuously operating reference stations available in some regions, with coverage and reliability that vary depending on where you’re working.

Most crews default to a subscription network for consistency and keep single-baseline capability in reserve for jobs that fall outside network coverage.

Two surveyors with GNSS rover poles on a grassy hillside above a valley

RTCM Compatibility: What to Check Before You Buy

Not every receiver handles RTCM data the same way, and mismatched hardware is a common source of field frustration. Before pairing a new receiver with a correction service, confirm it supports RTCM 3.x, since older RTCM 2.x formats are largely obsolete; that it decodes observation messages for all constellations you intend to track, not just GPS; and that it has NTRIP client capability built in, or can be paired with a data collector that provides it.

The Hemisphere S631 and similar survey-grade receivers handle this natively, decoding multi-constellation RTCM streams and applying corrections without extra configuration once the NTRIP connection is set.

The Correction Nobody Sees, But Every Fix Depends On

RTCM never shows up on a data collector screen. Nobody reads it, nobody quotes it in a deliverable, and most surveyors go entire careers without needing to know a message type number. But it’s the format doing the actual work behind every fixed solution, translating shared satellite error into a correction a rover can use in real time.

Getting reliable results in the field comes down to the same handful of variables every time: a clean, low-latency RTCM stream, a receiver that decodes it properly, and an antenna good enough not to waste the accuracy the correction data is providing. Get those three aligned, and the rest of the survey workflow takes care of itself.

About the Author

Portrait of Nolan Betzner, Bench Mark team member
Nolan has been working in the surveying field since 2017, starting as a part-time student at Bench-Mark while attending the University of Calgary. He now works in technical support and sales helping customers find the right product for them.

In this article