Field Quality Control

Overview#

Quality control in the field is the practice of detecting and correcting errors before they leave the field. Every measurement contains some combination of random error, systematic error, and the ever-present risk of blunders (mistakes). The purpose of field QC is to catch blunders immediately, minimize systematic errors through proper procedures, and quantify random errors through redundancy.

A survey that looks good on paper but was never checked in the field is a liability. A survey with documented QC checks is a defensible professional work product. The difference is not talent or expensive equipment -- it is discipline and procedure.

"The purpose of redundant measurements is not to get a better answer -- it is to prove that the answer is correct. Redundancy is the surveyor's insurance policy." -- Ghilani, C.D., Adjustment Computations: Spatial Data Analysis (6th Ed.), Ch. 1, p. 4

The Principle of Redundancy#

Why Redundancy Matters

A single measurement tells you nothing about its quality. Two measurements that agree give you confidence. Three measurements allow you to quantify precision and detect a blunder. Redundancy is the foundation of all quality control in surveying.

Minimum redundancy requirements:

The Concept of Independent Checks

True redundancy requires independence -- the second measurement must not share the same error sources as the first. Examples:

• Measuring a distance twice from the same setup is not fully independent (shares centering error, HI error, atmospheric error).
• Measuring a distance from station A to station B, then from station B to station A, is more independent (different centering errors, checks for prism constant errors).
• Measuring a point by total station and then by GNSS is highly independent (completely different instruments, methods, and error sources).

Check Shots#

What They Are

A check shot is a measurement to a point of known position taken specifically to verify that the instrument setup, orientation, and calibration are correct. Check shots should be taken:

• At the beginning of every instrument setup (backsight check).
• After any disturbance to the instrument (bumped tripod, releveled, changed settings).
• Periodically throughout the day (every 30--60 minutes of continuous work, or every 20--30 points).
• At the end of the day before tearing down.
• Whenever results seem questionable.

Evaluating Check Shots

Compare the measured position of the check point to its known position:

$\Delta H = \sqrt{(\Delta N)^2 + (\Delta E)^2}$

$\Delta V = |Z_{\text{measured}} - Z_{\text{known}}|$

where $\Delta N$ and $\Delta E$ are the northing and easting differences, and $\Delta V$ is the vertical difference.

Acceptable Check Shot Tolerances

If a check shot exceeds tolerance, stop work immediately. Do not continue collecting data until the cause is identified and resolved. Common causes include: bumped tripod, incorrect backsight, wrong HI, wrong prism constant, atmospheric change, or instrument malfunction.

Closure Checks#

Angular Closure

For a closed traverse or a round of angles, angular closure provides an immediate field check on the quality of the angle measurements.

Closed traverse (polygon):

$\text{Misclosure} = \sum \alpha_{\text{measured}} - (n - 2) \times 180°$

Horizon closure (direction method):

$\text{Misclosure} = \text{Final reading on initial target} - \text{Initial reading}$

Distance Closure

In a traverse or control network, distance closure is checked by comparing measured distances to computed (adjusted) distances. Reciprocal distance measurements (A to B, then B to A) should agree within:

$|\Delta D| \leq 2\sqrt{2} \cdot \sigma_D$

where $\sigma_D$ is the standard deviation of a single distance measurement.

Level Loop Closure

A level loop that returns to the starting benchmark provides the most direct quality check in leveling:

$\text{Misclosure} = \sum BS - \sum FS - (\text{Elev}_{\text{end}} - \text{Elev}_{\text{start}})$

For a loop (start = end), the misclosure should be zero. The allowable misclosure depends on the survey order (see Measurement Techniques guide for standards by order).

Quick field estimate for third-order leveling:

$\text{Allowable misclosure} = \pm 12 \text{ mm} \sqrt{K}$

where $K$ is the total loop distance in kilometers. For a 2 km loop: $\pm 12 \times \sqrt{2} = \pm 17$ mm.

GNSS Closure Checks

For GNSS RTK surveys, closure checks include:

• Re-initialization check: After the rover loses and regains a fixed solution, re-measure a previously observed point. The two observations should agree within the expected RTK accuracy ($\pm$ 15--20 mm horizontal).
• Control point check: Occupy a known control point with the rover and compare the measured position to the published coordinates.
• Base station verification: At the start and end of each RTK session, verify the base station is broadcasting the correct coordinates by occupying a second known point with the rover.

Acceptable Tolerances by Survey Type#

The following table summarizes field accuracy standards for common survey types. These represent the total allowable error in the final product, not individual measurement tolerances:

The ALTA/NSPS standard uses Relative Positional Precision (RPP): the uncertainty in the position of any point relative to any other point on the survey, computed at the 95% confidence level. The current standard requires RPP $\leq$ 2 cm (0.07 feet) plus 50 parts per million.

Instrument Checks in the Field#

Quick Checks (No Special Equipment Needed)

These checks can be performed on-site without returning to the office:

Setting Up a Field Check Distance

If an NGS calibration baseline is not nearby, establish a field check distance:

1. Set two stable points approximately 100--300 m apart with clear line of sight.
2. Measure the distance with the EDM using proper procedures (FL/FR, multiple measurements).
3. Record the distance and atmospheric conditions as the reference value.
4. Re-measure this distance periodically as a field check. Changes exceeding $\pm 3$ mm + ppm suggest instrument problems.

Error Detection vs. Error Correction#

Detection

The field surveyor's primary QC role is detection -- finding errors before they contaminate the dataset. Detection tools include:

• Closure checks (angular, linear, level loop)
• Check shots to known points
• Internal consistency checks (re-measuring a line from both ends)
• Comparison of redundant measurements
• Visual reasonableness checks (does this elevation make sense for this location?)

Correction in the Field

Some errors can and should be corrected immediately:

Errors That Require Office Processing

• Systematic angular errors are removed by FL/FR averaging, which happens during data processing.
• Traverse misclosure is distributed through adjustment (compass rule, least squares).
• Leveling misclosure is distributed proportionally by distance.
• Network adjustment (least squares) identifies outliers and distributes residual error optimally.

Never adjust data in the field to force closure. If a traverse or level loop does not close within tolerance, the correct action is to re-measure until the source of the error is found, not to change numbers to make them fit.

Documentation of Field QC#

What to Record

Every QC check performed in the field should be documented:

• Check shots: Point ID, measured vs. known coordinates, computed differences, pass/fail.
• Closures: Misclosure value, tolerance, pass/fail.
• Instrument checks: Date, type of check, result, any adjustment made.
• Re-measurements: Why the re-measurement was necessary, the original and new values, which value was accepted.
• Environmental conditions: Temperature, pressure, wind, visibility -- anything that may affect measurement quality.

QC Log Format

A simple field QC log can be maintained in the field book or data collector:

Check Ties to Control#

Purpose

Check ties are measurements from the survey to independent control points that were not used in the survey setup. They provide an external verification that the survey is correctly positioned and oriented.

Best Practices

• Include at least two check ties per survey, from different parts of the project area.
• Check ties should be to points with coordinates from an independent source (NGS control, separate GNSS observation, county monuments).
• Measure check ties using the same methods and accuracy as the rest of the survey.
• If a check tie fails, investigate before continuing. The failure may indicate a systematic error affecting the entire survey.

Interpreting Check Tie Results

• Pass (within tolerance): The survey is consistent with independent control. Document and continue.
• Marginal (near tolerance): Investigate. The issue may resolve with a simple correction (HI, prism constant) or may indicate a deeper problem.
• Fail (exceeds tolerance): Stop work. Identify the source of the discrepancy before collecting more data. Common causes: wrong control coordinates, datum mismatch, blunder in setup, disturbed monument.

Blunder Detection#

What Blunders Look Like

Blunders are gross errors -- mistakes, not measurement imprecision. Common blunders include:

• Reading or recording the wrong value (transposed digits, wrong rod reading).
• Setting up on the wrong point.
• Using the wrong control coordinates.
• Entering the wrong HI or prism height.
• Sighting the wrong target (especially in robotic operation -- the instrument may lock onto a passing vehicle).
• Wrong prism constant.
• Wrong atmospheric correction (or none entered).
• Recording a float GNSS solution as fixed.

Detection Methods

The best time to find a blunder is before you leave the station. Plot your data in the field on the data collector map view. If a point looks wrong, it probably is. Investigate immediately.

Field Verification Procedures#

Before Leaving a Station

1. Review all points collected from this station on the data collector map.
2. Verify that all planned features have been shot.
3. Check the backsight one final time (direction and distance).
4. Confirm that the data file has been saved.
5. Update the field sketch with any new features or notes.

Before Leaving the Site

1. Review the day's work on the data collector -- scan for obvious outliers or missing data.
2. Verify that all control points have been checked.
3. Close any open traverses or level loops if possible.
4. Confirm that check shots are within tolerance.
5. Take any remaining photographs.
6. Note any points that need follow-up or re-measurement.

Back in the Office (Same Day)

1. Download raw data from the data collector.
2. Back up the data to the project server.
3. Perform a quick review of the data in processing software -- check for outliers, missing codes, and unreasonable values.
4. If problems are found, flag them for re-measurement on the next field visit.
5. Plot the data against the field sketch. Verify that the electronic data matches the field notes.

Summary Statistics for Field QC#

When enough redundant observations exist, compute summary statistics to characterize the quality of the day's work:

Standard deviation of repeated measurements:

$\sigma = \sqrt{\frac{\sum v^2}{n - 1}}$

where $v$ is the residual (difference between individual measurement and mean) and $n$ is the number of measurements.

Standard error of the mean:

$\sigma_{\bar{x}} = \frac{\sigma}{\sqrt{n}}$

95% confidence interval:

$\bar{x} \pm 1.96 \cdot \sigma_{\bar{x}}$

These statistics allow you to verify that the achieved precision meets the project requirements and to identify observations that may contain blunders (outliers beyond $3\sigma$).

Key Takeaways#

• Redundancy is not optional. Every critical measurement must have an independent check. A single unchecked measurement is an assumption, not a fact.
• Check shots should be taken at every instrument setup and periodically throughout the day. If a check shot fails, stop work and investigate.
• Closure checks (angular, linear, level loop, GNSS) provide the most direct evidence of measurement quality. Always close when possible.
• Know your tolerances. Different survey types have different accuracy requirements. Understand the standard that applies to your project and verify compliance in the field.
• Detect errors before leaving the station. Plot data, review codes, check the backsight. It is always faster to fix a problem in the field than to return for re-measurement.
• Document all QC checks. A QC log provides defensible evidence that the survey meets professional standards. Undocumented checks may as well have not been performed.
• Never force closure. If the data does not close within tolerance, the answer is to find the error and re-measure, not to adjust the numbers.
• Blunder detection is a mindset. Always be asking: "Does this result make sense?" The surveyor who stops questioning the data is the one who misses the blunder.

References#

1. Ghilani, C.D. Adjustment Computations: Spatial Data Analysis (6th Ed.). Wiley, 2017. Chapters 1--3.
2. Ghilani, C.D. & Wolf, P.R. Elementary Surveying: An Introduction to Geomatics (13th Ed.). Pearson, 2012. Chapters 3, 9, 11.
3. Federal Geodetic Control Subcommittee. Standards and Specifications for Geodetic Control Networks. FGCS, 1984.
4. ALTA/NSPS. Minimum Standard Detail Requirements for ALTA/NSPS Land Title Surveys (2021). Section 3.E -- Relative Positional Precision.
5. U.S. Department of the Interior, Bureau of Land Management. Manual of Surveying Instructions (2009). Chapter 4 -- Measurements.
6. Mikhail, E.M. & Ackermann, F. Observations and Least Squares. University Press of America, 1976.
7. Kavanagh, B.F. & Mastin, T.B. Surveying: Principles and Applications (9th Ed.). Pearson, 2014. Chapters 3, 7.