6.3 Planning Control Networks to Accuracy Standards

Key Takeaways

  • On and after January 1, 2023, a survey that establishes CCS83 values must have field-observed statistically independent connections to two or more qualifying horizontal reference stations (PRC 8813.1(a)(2)).
  • If an accuracy is claimed, PRC 8813.1(b) and 8813.2 require an FGDC or FGCS published standard; FGDC accuracies must be identified as local or network accuracy and justified with equipment, diagram, adjustment, and residuals.
  • Professional Land Surveyors’ Act section 8771.5 requires a Record of Survey that shows California Coordinate System coordinates to show the control scheme from points of known coordinates.
  • Plan GNSS occupations against multipath and against uncalibrated antennas or instruments; those error sources, not just least-squares cosmetics, decide whether a claimed FGDC accuracy is honest.
  • Earth movement is a network-design input: PRC 8815.3 requires consistent epochs, HTDP or CSRC published values supply the shift, and a single RTN fix does not replace independent connections or an epoch note.
Last updated: September 2026

Domain II Activity 7 asks you to plan control networks. Independent OpenExamPrep teaching in this section is the method side of Knowledge items on producing control inside the Public Resources Code and federal or state accuracy standards, plus error sources and earth movement. A beautiful GNSS loop that is connected to one CORS datasheet, occupied beside a metal warehouse, and labeled "NAD83" without an epoch is not a planned network. It is a future retracement problem.

Decide what you are producing before you occupy anything

Start with the product, then the standard, then the stations.

  1. Purpose. Boundary control, topographic mapping, construction layout, or geodetic densification each tolerate a different error budget. Construction staking that must match an existing site grid is not the same job as establishing CCS83 values that later appear on a Record of Survey.
  2. Named system. If the product will show State Plane Coordinates, PRC 8817 still points new work to CCS83, with CCS2022 permitted after January 1, 2025. If the product is geodetic, PRC Chapter 4 (CGC83) and Chapter 3 (CSRN) apply. If the product is orthometric, PRC Chapter 5 (COH88) distinguishes leveled from derived heights.
  3. Claimed accuracy, or none. You are not required to claim an FGDC or FGCS accuracy. If you do claim one, PRC 8813.1(b) says the claimed accuracy shall be a standard published by FGDC or FGCS. Homemade "±0.04 ft GNSS" language is not an FGDC standard.

Federal Geographic Data Committee (FGDC) Geospatial Positioning Accuracy Standards, Part 2 (FGDC-STD-007.2-1998), report local accuracy and network accuracy at a 95 percent confidence level for each geodetic component (horizontal, ellipsoid height, orthometric height). Local accuracy describes how well a point fits nearby control—the closure a surveyor feels between two adjacent marks. Network accuracy describes how well coordinates approach an ideal, error-free datum. PRC 8813.2 requires FGDC accuracies to be identified as local or network. Federal Geodetic Control Subcommittee (FGCS) orders (first order, second order, third order and their classes) remain available; CSRN membership tests in PRC 8856–8859 still cite both families.

Connections: using CCS83 versus establishing CCS83

PRC 8813.1 is the connection statute for surveys on and after December 31, 2005.

A survey that uses or establishes CCS83 values must be referenced to and have field-observed statistically independent connections to qualifying horizontal reference stations. Qualifying stations are a CSRN station; an out-of-state station that meets CSRN tests except for being outside California; or an existing CCS83 station shown on a qualifying public map that meets CSRN requirements except NGS/CSRC publication and that states a CSRN-conforming accuracy.

On and after January 1, 2023, a survey that establishes a CCS83 value or values must connect to two or more such stations (PRC 8813.1(a)(2)). One connection can be enough when you are only using existing CCS83 values. Two independent connections are the floor when you are creating new CCS83 values. "Statistically independent" is the statutory phrase: two vectors that are merely two radio hits on the same RTN mount point in the same session are not two independent connections.

Professional Land Surveyors’ Act section 8771.5 adds a map-facing rule: when California Coordinate System coordinates are shown for points on a Record of Survey, the map may not be recorded unless it also shows, or is accompanied by a map showing, the control scheme through which the coordinates were determined from points of known coordinates. PRC 8813.2 and 8813.3 tell you what that scheme documentation must include if accuracy is claimed: the CCS83 values, the FGDC or FGCS standard, the control diagram of required independent connections, equipment, adjustment methodology and software, procedures or a cited specification, final residuals or closures, the reference stations, and those stations' published or stated accuracies.

Worked planning example: you will set two new CCS83 control marks for a 2026 Santa Clara County (Zone 3) subdivision map that will show CCS83 coordinates. Plan field-observed, statistically independent GNSS sessions to two CSRN stations (or two qualifying county CCS83 stations). Recover those stations, check their published realization and epoch—if one is NGS NAD83(2011) epoch 2010.00 and the other is CSRC Epoch 2025.00, PRC 8815.3 requires an epoch reduction using NGS or CSRC published procedures before the adjustment. Design two or more independent occupations per new mark, with different satellite geometry. After adjustment, if you claim accuracy, pick a published FGDC local and/or network statement at 95 percent and attach the 8813.2 narrative. Show the 8771.5 control scheme on the Record of Survey. Show PRC 8815.5 mapping angle, combined grid factor, and elevation at a representative point. Chapter 15 owns the full grid-versus-ground derivation; the planning duty here is to occupy a representative point whose elevation you actually know so that factor is not fiction.

Error sources you plan for, not errors you discover in the office

Knowledge items on this exam still name multipath and calibration. They belong in the reconnaissance, not in the residual plot.

  • Multipath. GNSS signals reflecting from buildings, vehicles, chain-link, water, or solar panels bias the range. Plan occupations away from those surfaces, use adequate antenna height, and reject short RTK fixes next to construction trailers when the product is geodetic control.
  • Antenna and instrument calibration. NGS absolute antenna calibrations (phase-center offsets and variations) belong in the processing. Mixing an uncalibrated generic antenna model with a CSRN CORS that has a published calibration is a systematic bias. Terrestrial networks have the same issue: EDM and total-station calibration, collimation, and leveling-collimation checks are FGCS-era duties that did not disappear because the box also has GNSS.
  • Other planned errors. Troposphere and ionosphere on long GNSS baselines, incorrect antenna-height measurement, centering and plumbing over the mark, and using a geoid model with the wrong ellipsoid-height epoch (GEOID18 with a non-NAD83(2011) epoch 2010.0 height) all masquerade as "network error."

PRC 8819 does not prohibit new techniques that lack published FGCS specifications. That is permission to use a method. It is not permission to skip independent checks or to claim an FGDC accuracy you cannot justify.

Earth movement is part of the network, not a footnote

Western California is not a rigid NAD83 plate. HTDP models interseismic velocities plus coseismic and postseismic displacements. CSRC's CSRN Epoch 2025.00 publication exists because secular plate-boundary motion, earthquakes, fault creep, subsidence, and uplift make a 2017.50 coordinate stale. Planning a control network therefore includes:

  • Selecting controlling stations that have published velocities or that CSRC/NGS can move to one epoch.
  • Stating that epoch on the map (PRC 8815.1, 8854, 8876).
  • Not mixing CRTN Epoch 2025.00 streams (transmitted beginning August 11, 2025) with 2010.00 datasheets as if they were the same physical point.
  • Re-observing after a significant earthquake instead of assuming last year's control still sits on last year's coordinates. HTDP can estimate a displacement if NGS has a model for that event; if it does not, you do not invent one.

A real-time network such as CRTN is an access method to epoch-dated CSRN coordinates. It does not, by itself, satisfy 8813.1(a)(2) for establishing new CCS83 values, and it does not waive 8771.5 control-scheme drafting. Plan redundancy: independent sessions, opposite-sky GNSS windows or a terrestrial check, and at least the statutory two reference stations when you are establishing CCS83.

If the project later needs ground distances for construction, carry the PRC 8815.5 combined-factor metadata forward. The full grid-versus-ground analysis is Chapter 15. The control-network plan is finished only when the next surveyor can recover the datum, realization, epoch, zone, units, accuracy claim, and control scheme from the filed map without calling you.

Loading diagram...
Control-network planning from purpose through statutory documentation
PRC 8813.1 field-observed independent horizontal reference stations
Test Your Knowledge

On and after January 1, 2023, what connection rule applies to a survey that establishes new CCS83 values?

A
B
C
D
Test Your Knowledge

If the surveyor will claim an accuracy for newly established CCS83 values, which accuracy statement matches PRC 8813.1 and 8813.2?

A
B
C
D
Test Your Knowledge

Which planning statement treats error sources and earth movement the way California control statutes expect?

A
B
C
D