4.1 Data Collection Methods: Leveling and LiDAR

Data collection method on the California Civil Engineering Surveying (CES) exam is a planning decision: match the sensor to the use of the data, the site, and the accuracy you can actually support with control. The 2022 CES test plan lists leveling and LiDAR among Survey Planning knowledge items, together with the broader topographic toolkit a civil engineer uses when scoping work under Business and Professions Code (BPC) 6731.1. This section compares methods as choices. Running a level loop, reducing trigonometric heights, and checking notes are field procedures (later chapters). Here the question is when a civil engineer (CE) specifies each method.

What you are actually buying

Every method produces coordinates or elevations. They differ in density (how many points per acre), canopy penetration, ability to see utilities, corridor versus site logistics, and vertical versus horizontal strength. A method that is excellent for an open freeway corridor can be the wrong choice for a wooded 4-acre pump-station pad with unmarked laterals.

Differential leveling as a planning choice

Differential leveling measures height difference with a level and rod (or digital bar-code level). Plan it when the product is a vertical datum you will trust for design grade, hydraulic grade, or construction. A closed loop from a published or project benchmark is still the default way to establish project vertical control on a compact site. Do not specify a 500-foot grid of level shots as a substitute for topographic mapping of a 200-acre site—the method is accurate but slow, and it does not capture planimetric features.

Planning questions: How far is the nearest usable benchmark? Will the loop close at a tolerance consistent with the use of the elevations (Section 4.2)? Are turning points stable? On a long corridor, leveling is often reserved for control and critical profiles, while airborne methods fill the surface. The CES trap is treating 'LiDAR' as automatically tighter than a level. Airborne laser scanning (ALS) at roughly 10 cm root-mean-square error in z (RMSEz) is coarser than a closed level loop on a building pad.

Trigonometric leveling as a planning choice

Trigonometric leveling computes height from a measured zenith (or vertical) angle and a slope or horizontal distance, usually with a total station. Plan it when you already occupy a total-station scheme and need elevations on topographic shots, building corners, or invert shots a level cannot see. It is not a free upgrade from a level. Longer sights amplify angular error and refraction. For a design that needs hundredths of a foot on a pad, keep trig sights short, use a well-adjusted instrument, and tie the scheme to differential control. Full reduction (instrument height, target height, curvature, and refraction) is a field and analysis topic. The planning takeaway is distance-dependent vertical uncertainty.

GNSS heights as a planning choice

Global Navigation Satellite System (GNSS) gives three-dimensional positions in a geodetic frame. Two planning traps dominate civil work. First, ellipsoid height is not orthometric height (the height you use with the North American Vertical Datum of 1988 (NAVD88) or a local benchmark) until a geoid model or a local fit is applied. Second, canopy, urban canyons, and multipath destroy vertical quality faster than horizontal quality. Specify GNSS for open-sky control, airborne-survey control, and sparse topography of open pads or fields. Do not plan GNSS topography as the sole existing-ground model under oaks or in a downtown utility corridor.

Total-station topographic survey as a planning choice

A total-station topographic survey is still the default site method when the CE needs selected civil features: building corners, finished floors, invert elevations, trees, curb, and utilities that a locator has marked or that the crew can see. Density is a written specification. A 25-foot grid plus breaklines is a different product from 'enough shots to draw 1-foot contours.' The crew must occupy or see the point. Vegetation and walls create holes that LiDAR or photographs may fill—or may not, if the ground is invisible.

For utilities, no airborne method locates a buried water service. Planning a topographic survey without a subsurface utility engineering (SUE) pass, or at least an Underground Service Alert (USA/811) locate-and-survey pass, is a scope error, not a sensor error. Potholing at conflict points remains a field measurement, not a GIS guess.

Aerial photogrammetry and the photo-scale formula

Aerial photogrammetry recovers three-dimensional measurements from overlapping photographs. Plan it for large, relatively open areas where you need both a surface and photographic texture (pavement limits, stains, vegetation outlines). It requires photo control, or GNSS/inertial-assisted aerotriangulation that is still checked by survey. Canopy hides the ground; shadows and water complicate matching.

The fundamental planning formula for a vertical photograph over reasonably flat terrain is:

S = f / H_flying

S is the photo scale as a representative fraction (photo distance divided by ground distance), f is the camera focal length, and H_flying is flying height above the terrain—not above the ellipsoid unless elevation is being handled explicitly. Units of f and H must match. If flying height above a datum is H and ground elevation is h, the scale at that point is S = f / (H − h). Higher ground produces a larger scale (objects look bigger).

Worked photo-scale example

A mapping camera has f = 152.4 mm. The aircraft flies 1,830 m above mean terrain. Convert focal length to meters: 152.4 mm = 0.1524 m.

S = 0.1524 / 1830 = 0.00008333… = 1/12,000, written 1:12,000.

Because 12,000 inches equal 1,000 feet, this is also 1 inch on the photo ≈ 1,000 feet on the ground.

Reverse planning: you want 1:6,000 photography with a 6-inch (0.50 ft) focal-length camera. Then H_flying = f / S = 0.50 / (1/6000) = 3,000 ft above terrain. If mean ground is 420 ft on the project datum, the flight plan must target about 3,420 ft above that datum. CES items often hide a unit conversion (millimeters versus meters, inches versus feet). Fail the conversion and the scale is wrong by a factor of 10 or 12.

Photo scale is not compiled map scale. Enlargement can produce a 1 inch = 40 ft map from smaller-scale photographs, but you cannot invent ground detail the original resolution never captured. Pick a flying height and camera that support the intended contour interval and smallest feature.

LiDAR: airborne and terrestrial

LiDAR (Light Detection and Ranging) ranges with a laser and timestamps returns. Airborne laser scanning (ALS) mounts the scanner in an aircraft or uncrewed aircraft with GNSS and an inertial measurement unit. Terrestrial laser scanning (TLS) occupies ground setups, often with targets or a total-station/GNSS tie.

Plan ALS when you need a dense surface over a corridor or large site, including many vegetated areas where photogrammetry cannot see the ground. Multiple returns and classification can yield a bare-earth model, but dense canopy, wet ground, and clutter still produce voids and vegetation bias. Plan TLS when the object is a plant, bridge, facade, or tight urban site you can surround with setups. TLS point density is typically orders of magnitude higher than ALS; it is the wrong primary tool for 12 miles of rural alignment. UAV (drone) LiDAR follows the same planning logic as manned ALS—control, density, canopy, and check shots—on a smaller footprint.

LiDAR does not replace utility designation. It also does not, by itself, create a property boundary.

Point density versus contour interval

Point density (points per square meter, pts/m²) and vertical accuracy are related but not the same. Sparse, very accurate points cannot support a fine contour if the ground undulates between shots. Dense but noisy points can produce fake wiggles in the contours.

U.S. Geological Survey (USGS) 3D Elevation Program quality levels are professional-practice specifications for airborne lidar. They are not Board-published CES numeric cutoffs. As published by USGS:

A legacy office rule of thumb is that a suitable contour interval is on the order of three times RMSEz in open terrain—about 1-ft contours from ~10 cm RMSE—if breaklines and classification are competent. The American Society for Photogrammetry and Remote Sensing (ASPRS) now reports metric RMSE classes rather than contour-interval classes. On the exam, know the split: density supports interpolation and breakline capture; RMSE and independent check shots support the accuracy claim.

For a highway corridor needing 1-ft contours, QL2 (≥ 2 pts/m², ~10 cm RMSE non-vegetated) is a common starting specification, with breaklines at curb, flowline, and drainage, plus conventional survey at structures. For a 0.5-ft contour site plan, density and accuracy both step up (often QL1 density and tighter local control), or you switch to total station or TLS.

Nominal pulse spacing is about 1 / √(density). At 2 pts/m², spacing is about 0.71 m (~2.3 ft); at 8 pts/m², about 0.35 m (~1.1 ft). If you must honor a 1-ft grade break, 2-ft pulse spacing is a weak plan without surveyed breaklines.

Corridor versus site, canopy, and utilities

Corridors (highways, channels, pipelines) favor ALS or photogrammetry plus control, because occupying every station with a total station is slow. Sites (plants, buildings, intersections) favor total station, TLS, and leveling because of occlusions, utilities, and grade tolerance.

Canopy: photogrammetry maps the treetops; ALS may reach the ground through gaps; a total station maps what you can see or occupy; GNSS fails. Utilities: field locate and survey, then pothole at conflict points. GIS utility layers are planning overlays only (Section 4.3).

How a CE chooses

1. Write the use: design grade, earthwork, drainage, mapping exhibit, or construction.
2. Write the obstacles: canopy, traffic, water, buried utilities, access.
3. Size the area: acres versus miles.
4. Choose a primary method and a tie method (usually differential or GNSS control).
5. Specify density, contour interval, and a check-shot plan together—not density alone.

Exam traps: treating LiDAR as universally tighter than a level; using S = f/H with mixed units; assuming airborne data locates underground utilities; specifying TLS as the only tool for a multi-mile corridor.