8.2 Control Surveys in the Field
Key Takeaways
- CCS83 is a six-zone Lambert system on NAD 83; field documents must show the realization tag and epoch to two decimal places (example: CCS83 (2011) epoch 2010.00) per Public Resources Code §8815.1.
- After December 31, 2005, a survey that uses or establishes CCS83 values needs field-observed, statistically independent connections to CSRN or equivalent stations (PRC §8813.1), not a single unchecked rover fix.
- CRTN is CSRC’s statewide real-time GNSS utility: dual-frequency receiver, NTRIP, current station health, and the epoch in the RTCM stream—which CSRC has been publishing as CSRS Epoch 2025.00 NAD83 on migrated streams.
- Static GNSS for project primary control uses simultaneous occupations, typically on the order of 30 to 120 minutes depending on baseline, plus repeat observations; a single 5-minute RTN shot is not that network.
- Basis of bearings and the named NAVD 88 benchmark on the plans are the values you occupy and check; mixing epochs in tectonically active California is a systematic error, not a rounding issue.
Why field control is its own exam activity
Domain III, activity 2 is perform control surveys. Knowledge items K, L, and N are the field cluster: California Coordinate Systems, real-time networks (processes, redundancy, accessibility, accuracy), and basis of control values (basis of bearings, benchmark) as they appear on maps and construction plans. Office datum theory is useless if the rover is in the wrong zone or the crew never occupies the named benchmark.
This section is field procedure. It is not a boundary retracement and not a hydrographic reduction.
CCS in the collector and on the ground
Public Resources Code Division 8, Chapter 1 (§§ 8801–8819) defines the California Coordinate System. CCS27 has seven zones. CCS83 has six Lambert conformal conic zones on GRS 80 / NAD 83. Former CCS27 Zone 7 (Los Angeles County) is in CCS83 Zone 5 (§8802). Each CCS83 zone origin is published as Northing (y) = 500,000 meters and Easting (x) = 2,000,000 meters.
Field rules that show up on documents and in the data collector:
- Suffix the system: CCS83, CCS27, or CCS2022 (§8815, as amended by SB 566, effective January 1, 2024).
- When CCS83 coordinates are shown, state the datum realization tag and epoch in decimal-year format to two decimal places, for example
CCS83 (2011) epoch 2010.00(§8815.1). - The epoch of the survey is the published NGS or California Spatial Reference Center (CSRC) realization of a controlling station. After December 31, 1999, that must be NAD83 (1992) epoch 1991.35 or a later published NGS/CSRC realization (§8815.2).
- On or after January 1, 1995, new State Plane work uses CCS83, not CCS27; retracement of a CCS27 survey is still allowed. After January 1, 2025, new surveys may use CCS2022 (§8817). CCS83 remains everyday California practice until a job explicitly holds CCS2022 / NATRF2022 products.
- Feet, when used, are U.S. Survey Feet (1 ft = 1200/3937 m).
- After December 31, 2005, a survey that uses or establishes CCS83 values must have field-observed, statistically independent connections to one or more California Spatial Reference Network (CSRN) stations (or listed equivalents, including certain filed CCS83 stations) (§8813.1). The connected stations, their coordinates, and published or stated accuracies go on the document (§8813.3).
Grid versus ground is a field setting, not an office footnote. Caltrans Surveys Manual Chapter 4 (May 2013 posting) tells crews to convert ground distances to CCS83 grid by the combined grid factor (CGF), expressed to at least seven decimal places. Example: CGF = 0.9999321. A 1,000.00 ft ground distance = 1,000.00 × 0.9999321 = 999.93 ft on the grid (999.9321 ft before 0.01 ft rounding). Staking a long street from CCS83 coordinates while the contractor is building ground distances—without applying the factor the plans use—walks the job off by several tenths in a mile.
Wrong-zone is worse than a factor error. A rover in Zone 6 on a Zone 5 job is not “a little off.” It is a different projection.
Basis of bearings and benchmarks in the field
Read the control sheet, then occupy what it names.
- Basis of bearings — commonly the CCS83 grid bearing defined by field ties to named CSRN/NGS stations. CSRC’s sample language still expects those stations, northings, and eastings on the map. If you occupy different marks, or a different epoch, you do not hold the plans’ basis of bearings.
- Benchmark — a published or project NAVD 88 (or stated) orthometric height. GNSS ellipsoid height becomes an orthometric height only through a geoid model (or a local fit) that the project documents. Caltrans Chapter 4 tells project control to base elevations on at least two monuments with NAVD 88 heights of third order or better when that is the project datum.
- California tectonics — monuments move relative to the CCS83 grid. Holding epoch 2007.00 hubs against a CRTN stream publishing epoch 2025.00, without a documented transformation, is a systematic easting/northing shift. That is why the epoch is a field check, not decoration.
Real-time networks: process, redundancy, accessibility, accuracy
The California Real Time Network (CRTN), managed through CSRC, rebroadcasts high-rate GNSS from partner CORS (EarthScope/NOTA, BARD, USGS, Caltrans CTSRN, SOPAC, and others). Field process:
- Access — survey-grade dual-frequency GNSS, internet or cell modem, NTRIP login, then pick a live mountpoint/base from the caster. CSRC documents RTCM streams at 1 Hz with latency under one second when the station is healthy.
- Epoch in the stream — CSRC has been migrating CRTN RTCM coordinates to CSRS Epoch 2025.00 (NAD83). Some worksheets still warn that a subset of stations carried Epoch 2017.50 values during transition. Match the stream to the plans, or transform.
- Health and accessibility — read the CRTN status map before relying on a mountpoint. A listed station can be down, high-latency, or on the wrong epoch worksheet. Cell coverage is part of accessibility: no NTRIP, no RTN.
- Accuracy — network RTK models spatially correlated errors. It is still a real-time kinematic solution. Multipath, wrong antenna height, wrong geoid, and a dead reference station will not be fixed by a pretty RMS.
Redundancy is the item L word that fails single-occupation habits. CLSA/CSRC GNSS Surveying Standards and Specifications v1.1 (December 10, 2014) treats repeat occupations at different sidereal times and satellite geometry as the defense against systematic GNSS error. A Los Angeles County Public Works testimonial in CSRC’s 2021 plan described CRTN use as 5-minute measurements three times at different times of day—an agency practice example, not a Board-published cutoff. The exam idea is the same: independent occupations plus checks into known CCS83 stations, not three 10-second averages without breaking down the tripod.
CSRC states that CRTN is intended to support PRC requirements for GPS-derived geodetic coordinates and orthometric heights (CSRC cites §§ 8856, 8857, 8858). That is a network claim. Your project still has to show the field-observed connections and checks §8813.1 expects.
Worked example: RTN versus static GNSS for project control
A 1.2-mile arterial will be designed and staked on CCS83 Zone 3, NAD83(2011) epoch 2017.50, NAVD 88. You must set four new primary hubs (A–D).
| Method | Typical occupation (guidance, not a Board cutoff) | Redundancy you must still design | Defensible field use |
|---|---|---|---|
| Static GNSS | 30 to 120 minutes, simultaneous receivers | Repeat session at a different sidereal window; CSRN/filed CCS83 ties | Primary project control at the plans’ epoch |
| Fast-static | Often 5 to 10 minutes, dual-frequency | Still a network of simultaneous occupations and repeats | Shorter primary; weaker than static |
| Single RTN occupation | About 5 minutes, one rover | None | Not primary; production topo or staking only after hubs are proven |
| Redundant RTN | Independent occupations (agency example: 5 minutes, three times of day) | Station-health check plus known CCS83 stations at the plans’ epoch | Primary only when those checks close |
Static (defensible primary): Two or more receivers occupy hubs and at least one published CSRN/CSRC station simultaneously. CLSA/CSRC typical static sessions are 30 to 120 minutes depending on baseline length, long enough to resolve ambiguities and change satellite geometry. Repeat at least one occupation on another session (different sidereal window). Process baselines, run a constrained adjustment to the published marks at epoch 2017.50, and report residuals. Run a closed level loop (or equivalent) to two NAVD 88 benches. Fast-static (often 5 to 10 minutes, dual-frequency) is shorter than static and weaker; it is still a network method, not a single RTN fix.
Single RTN occupation (not primary): One rover, NTRIP to CRTN, 5-minute fixed solution on Hub A, RMS 0.02 ft, no published-mark check, no second occupation. You have an accessible, precise-looking coordinate in whatever epoch the stream is broadcasting. You do not have a redundant CCS83 control survey, and you may not even be on 2017.50.
Hybrid used on many California jobs: Prove primary hubs with static/fast-static or with redundant RTN occupations plus CSRN checks at the plans’ epoch. Then use RTN/RTK for topography and staking from those hubs. Do not constrain a least-squares adjustment solely to whatever coordinates the RTN caster is advertising that morning without checking the NSRS/CSRS values the plans name.
Numeric sketch: Hub A static-adjusted CCS83 position versus a same-day CRTN epoch-2025.00 fix differs by 0.08 ft north and 0.11 ft east (resultant 0.14 ft) because the stream epoch is not 2017.50. That 0.14 ft is not “GNSS noise.” It is an epoch mismatch. Transform or occupy the correct realization before you call Hub A primary.
Error sources you actually catch in the field
The test plan lists multipath, data input, and instrument calibration. In control work that means: tripod not over the mark, antenna height 2.000 m entered as 2.000 ft, geoid model off, wrong zone, uninitialized inertial tilt, and a level vial that was never checked. Redundancy is how those blunders appear. One pretty RTN coordinate hides them.
Field sequence: read plans (zone, epoch, benchmark) → recover published marks → choose static network or redundant RTN → check residuals → only then densify working control for topo and staking.
A 1.2-mile arterial must hold CCS83 Zone 3, NAD83(2011) epoch 2017.50 for primary project control. Which field approach is the defensible way to set that primary control?
When CCS83 coordinates are shown on a California map, corner record, or similar document, Public Resources Code §8815.1 requires which statement of the basis of those values?
Which CRTN / real-time-network field practice supplies the redundancy the PLS test plan emphasizes?