Introduction
Your receiver gives you coordinates. But how do you know they're correct? Quality control is the process of verifying, validating, and documenting GNSS measurements to ensure you can trust them.
The QC Mindset: Never trust a single measurement. Every GNSS observation has potential errors. QC is about detecting problems, quantifying quality, documenting confidence, and making decisions on whether to accept, reject, or remeasure.
Real-Time Quality Indicators
- Fix status: Fixed (best, ambiguities resolved) → Float (OK but degraded, decimeter–metre) → DGNSS → Standalone → Invalid (don't use)
- DOP values: PDOP <4 = Good; PDOP 4–6 = Acceptable; PDOP >6 = Consider waiting
- Number of satellites: 7+ = Good; 5–6 = Marginal; <5 = Risky (unless fixed RTK)
- Age of corrections: RTK/DGNSS ideally <10 seconds, older corrections mean less accuracy
- RMS/Standard deviations: Per-position quality estimates; usually optimistic, but useful for comparison
Post-Processing Quality Metrics
- Ambiguity resolution rate: Percentage of epochs with fixed solution, 95% = excellent; <80% = investigate
- Ratio factor: Quality of ambiguity resolution, >3 = good fix; <2 = weak fix
- Residuals: Post-fit measurement residuals, large residuals indicate problems with a satellite
- Repeatability: Measure same point multiple times, spread indicates precision
Field QC Procedures
Before starting
- Check antenna setup (plumb, measured height correctly)
- Verify receiver settings (datum, logging rate, elevation mask)
- Allow initialization time and check DOP predictions
During work
- Revisit known points periodically
- Log more than minimum needed
- Monitor fix status and watch for suspicious jumps
At end of day
- Download and backup data
- Run initial processing check
- Document any issues noted
Statistical QC
Typical RTK expectations (1 sigma): horizontal 1–2 cm, vertical 2–4 cm. Reject measurements beyond 3 sigma; investigate 2–3 sigma. DGNSS typical: 20–50 cm (1 sigma).
Common Problems and Solutions
| Symptom | Likely Cause | Check |
|---|---|---|
| Can't get fix | Obstructions, long baseline | Sky view, distance to base |
| Fixes but jumps | Multipath, high DOP | Environment, DOP values |
| Consistent bias | Wrong antenna height | Measure again carefully |
| Drifting | Bad base coordinates | Verify base station position |
| No RTK | Radio/cellular issue | Data link connectivity |
Documentation: The Metadata You Must Keep
- Date and time (UTC)
- Antenna type and serial number
- Antenna height and measurement method
- Base station used (or CORS network)
- Receiver settings
- Photos of setup, sky plots, any issues noted
Verification Methods
- Known point check: Measure a point with known coordinates and compare, should match within expected accuracy
- Redundant measurements: Measure same point multiple times at different times, spread indicates true precision
- Independent check: Different equipment or technique (total station, level), gold standard for critical work
- Network adjustment: Multiple points adjusted together; reveals inconsistencies; required for control networks
Decision Matrix
| Application | Required QC | Action if Failed |
|---|---|---|
| Hiking | None really | Ignore, it's fine |
| GIS mapping | Fix status, DOP | Reject points with poor QC |
| Construction stakeout | Fixed RTK, known point check | Remeasure before staking |
| Boundary survey | Fixed, redundant, independent check | Full investigation, possible rework |
| Control network | Everything + adjustment | Complete reprocessing |
Vital Points
- Never trust a single measurement, always verify
- Real-time indicators (fix, DOP, satellite count) give immediate quality sense
- Post-processing metrics (ratio, residuals) reveal deeper issues
- Field procedures prevent many errors before they happen
- Documentation is essential for traceability
- Different applications require different QC levels, know what yours requires