Understanding GCPs vs PPK vs RTK in Drone Surveys

How to choose (and combine) georeferencing methods for mining and environmental projects

Accurate geolocation is the foundation of reliable spatial data. In mining and environmental monitoring, centimetre-level precision influences engineering design, compliance reporting and long-term ecological analysis. Modern drones can fly almost anywhere, but the data captured is only as good as the way we georeference it.

In the article we explain the three most common georeferencing methods—Ground

Control Points (GCPs), Post-Processed Kinematic (PPK) and Real-Time Kinematic (RTK)—then show you how we combine them in practice, including a guide to how many GCPs you’re likely to need and where to put them.

Ground Control Points (GCPs)

Ground Control Points (GCPs) are targets laid on the ground and observed with high-precision GNSS. They are usually a painted cross, or a physical cross using PVC or plate with a printed cross or similar. They appear in the airborne imagery and are used to anchor the model to the real world during processing. The typical accuracy of data using GCPs is 3–5 cm horizontally and vertically (often better with good distribution and careful survey).

Pros

• High, audit-grade accuracy—especially in elevation.

• Independent of the drone’s onboard GNSS.

Cons

• Time-consuming and sometimes unsafe to install on active pits, waste dumps or fragile habitats.

• Targets must be visible in every photo; canopy, shadows or dirt can hide them.

Best for

• Regulatory or contractual deliverables.

• High-precision stockpile volumes.

• Terrain models over small, accessible areas.

  
  

Post-Processed Kinematic (PPK)

For Post-Processed Kinematic or PPK, the drone logs raw GNSS data during flight. After the mission these logs are corrected against a base-station file (on-site or CORS), producing centimetre-level coordinates for each image. The typical accuracy of data using PPK processing is 2–5 cm horizontally and vertically with good satellite geometry and a stable base station.

Pros

• Works anywhere—no real-time link from the base to the aircraft is required.

• No concern with potential signal dropout , such as in pits, gullies or dense vegetation.

• Usually removes the need for a comprehensive network of GCPs.

Cons

• Requires post-processing software and experience.

• Still benefits from a handful of GCPs/check points for validation.

• Needs a base station or CORS data within about 20 km from the survey area.

Best for

• Remote exploration corridors.

• Vegetated or hazardous terrain where placing targets is risky.

• Repeat surveys where identical flight lines are reused.

Real-Time Kinematic (RTK)

Real-Time Kinematic or TRK is where a GNSS base station (or virtual reference network) streams corrections to the drone during flight, so every photo is geotagged with centimetre-level coordinates in real time. The typical accuracy of data using RTK is 1–3 cm horizontally and 2–4 cm vertically when a reliable correction link is maintained.

Pros

• Fast field-to-deliverable turnaround—no GNSS correction step in processing.

• Minimises GCP labour in open, connected sites.

• Ideal for daily/weekly operational mapping and surveys.

Cons

• Relies on continuous UHF or 4G/5G coverage; dropouts degrade accuracy.

• Deep pits, ridgelines or canopy can block the correction signal.

• Still needs independent check points to prove accuracy.

Best for

• Civil earthworks progress updates and reconciliation surveys.

• Flat, open rehab areas with adequate radio or mobile coverage.

• Projects demanding rapid turnaround.

  
  

Our Preferred Workflow: PPK + A Few GCPs + Independent Check Points

For most of our jobs we prefer a hybrid approach:

1. Use a PPK workflow to achieve centimetre-grade camera positions without the requirement for a continuous base station -> drone link.

2. Install 4-5 GCPs in safe, open areas as a failsafe in case GNSS quality dips.

3. Survey 2-3 Check-Points that are not used in processing. These provide an independent report of horizontal and vertical error (RMSE).

   
   

GCPs tie the model down; check points are used to validate the data and make accuracy statements.

GCP and Check-Point Guide

We are commonly asked how many GCPS and Check-Points should be used. Use the following table as a rough guide.

Survey Area	Ground Control Points	Check-Points
<10 hectare	4-5	2-3
10 -100 hectare	5-10	3-4
>100 hectare	10 + (evenly spaced)	6 + (covering varying terrain)

Placement tips:

• Cover the extents – at least one target near every corner and along long edges.

• Vary elevation – include high and low spots if the terrain undulates.

• Interior support – place one GCP centrally on large sites to prevent sag.

• Distribute check points – put them on critical features: stockpile toes, road crowns, in the base of erosion gullies.

• Avoid clustering – spread points; clusters add little control.

In PPK or RTK workflows, well-placed GCPs plus several check points typically satisfy both accuracy and efficiency goals.

Conclusion

No single technique is perfect. The choice depends on required accuracy, site access, safety, schedule and budget. In practice, a hybrid approach—PPK for primary geotags, a few GCPs for insurance and independent check points for proof—delivers the best balance of accuracy and efficiency for most drone surveys.

Need help designing a control layout for your next project? Contact us and we’d be happy to review your survey design and recommend the optimal georeferencing workflow.