Surveying Deep Dive: Traverses, Control Networks, and Adjustments

Introduction

Beyond basic RTK, professional surveying uses sophisticated techniques to achieve the highest accuracy: static observations, traverses, control networks, and least squares adjustments. This is how sub-centimetre accuracy is achieved and legally defended.

Control Networks: The Foundation

Control is a set of monumented points with known coordinates that form the reference for all other surveys, tied to national and global datums.

• Primary (national/CORS): 1–5 mm accuracy
• Secondary (regional network): 5–10 mm accuracy
• Tertiary (project control): 1–2 cm accuracy

Static GNSS Surveying

The gold standard for control work: multiple receivers occupy points simultaneously, logging data for extended periods (1 hour to multiple days). Baselines between receivers are processed and the network is adjusted. Why static? Long observations average out errors, ambiguities are reliably fixed, highest accuracy is achievable, and results are traceable to national standards.

Traverses with GNSS

A GNSS traverse observes vectors (3D distance and direction) between control points, forming loops tied to existing control. Advantages over conventional traverses: direct 3D vectors, no orientation errors, and faster work. The procedure: establish control points, observe GNSS vectors between them, form loops, and adjust the network.

Least Squares Adjustment

Redundant measurements always conflict slightly, least squares finds the best estimate of true coordinates while providing rigorous quality measures.

• Inputs: Observed vectors (ΔX, ΔY, ΔZ), estimated standard deviations, fixed control coordinates
• Outputs: Adjusted coordinates, residuals (how well observations fit), standard deviations (quality), statistical tests (blunder detection)
• Method: Minimise the sum of squared residuals, weighted by observation quality

Network Design

Observation Specifications

Blunder Detection

• Loop closures: Measure around a closed figure, should sum to zero; large closure indicates a problem
• Baseline repeatability: Same baseline measured multiple times should agree within tolerance
• Residual analysis: Large residuals indicate a suspect observation, investigate before accepting
• Statistical tests: Tau test for outliers, global test for overall model fit

Combining with Conventional Surveys

GNSS alone isn't always sufficient. Under trees, inside buildings, or near tall structures, combine with total station traverses, levelling (for precise heights), and terrestrial laser scanning. An integrated least squares adjustment combines all observation types with appropriate weights for a single consistent solution.

Practical Workflow

1. Reconnaissance: Identify point locations, check sky visibility, plan occupations
2. Monumentation: Set permanent marks with documentation and photos
3. Observation: Schedule sessions, deploy receivers, log data
4. Processing: Download data, process baselines, check quality
5. Adjustment: Least squares, blunder detection, final coordinates
6. Reporting: Coordinate list, uncertainty estimates, metadata

Vital Points

• Control networks are the foundation of accurate surveying
• Static GNSS provides the highest accuracy achievable
• Least squares adjustment combines observations optimally
• Redundancy detects blunders and improves quality
• Observation specifications depend on required accuracy class
• GNSS is often combined with conventional methods for complete coverage
• Quality control throughout, not just at the end