4.3 GIS Applications for Civil Engineering Surveys

The 2022 CES test plan includes use and applications of GIS under Survey Planning. For a California civil engineer, a geographic information system (GIS) is not a substitute total station. It is a database of layers with locations and attributes. Those layers are either consumed from your survey or overlaid before you ever occupy a hub. The planning skill is knowing which is which.

GIS as a consumer of survey data

A completed engineering survey—control, breaklines, contours, as-built inverts, photographed utilities—often becomes GIS. The CE (or the surveyor working with the CE) delivers coordinates in a documented coordinate reference system (CRS), with metadata: datum, epoch if relevant, units (U.S. survey foot versus international foot), vertical datum, date, and a statement of how accuracy was tested. BPC 6731.1(c) and (d) expressly include creating or modifying electronic data for engineering-surveying products and rendering a statement regarding accuracy of maps or measured survey data for those products. A shapefile or CAD export with no CRS and no accuracy statement is a weak consumer product even if the field work was excellent.

Once in GIS, the survey supports asset inventory (manholes, valves, monuments you set for the project), exhibits for public meetings, and operations (which invert is 0.15 ft above the hydraulic grade). Those uses are downstream of the survey. They do not define the field method by themselves, but they do define the attribution you should collect (pipe size, material, invert, photo ID) while the crew is on site.

GIS as a planning overlay

Before fieldwork, GIS is how you scope the engineering survey:

• Utilities: city or district GIS, USA/811 tickets, and old as-builts show where to look. They do not show the invert you will design against until someone measures it.
• Parcels and rights-of-way: assessor maps, tax lots, and GIS centerlines are unsurveyed background. They help you see likely ownership and possible easements so you can plan access and conflicts. They are not field-located boundaries.
• Floodplains: Federal Emergency Management Agency (FEMA) National Flood Hazard Layer polygons and base flood elevations are regulatory overlays. They tell you whether finish floor and grading need a tighter vertical survey; they are not a site topographic model.
• As-builts and prior projects: scanned plans and old CAD may exist as GIS attachments. Vintage and unknown CRS make them clues, not control.
• Terrain: county lidar or USGS 3DEP may exist as a raster. That is a method choice (Section 4.1) plus an accuracy choice (Section 4.2), not an automatic design surface.

The planning deliverable is a base map for scoping: print or screen the overlays, then write a field list—ties to control, utility designation, drainage structures, canopy holes, and pad grid—that GIS cannot finish.

Coordinate-system pitfalls

The most expensive GIS mistake in civil survey planning is overlaying layers that do not share a coordinate framework and believing the picture.

A common job setup is an assumed local system: northing 10,000, easting 5,000, benchmark elevation 100.00 ft, with a rotation that makes a property line 'due north' on the sheet. County GIS parcels, floodplains, and utilities are typically in a published projected CRS—in California often a California Coordinate System of 1983 (CCS83) zone, U.S. survey feet—on NAD83 (or a successor realization) horizontally and NAVD88 (or another published vertical datum) vertically. Those two worlds do not share an origin. 'Both are in feet' is not a transformation.

If you drop the assumed-grid CAD on top of CCS83 GIS without a similarity or georeferencing step that you document, hydrants can land in the street, flood zones can miss the pad, and a designer can 'see' a 40-ft easement conflict that is only a coordinate fiction. The reverse error is staking from GIS vertices that were never tied to your control.

Planning rules for mixed CRS

1. Name the project CRS in the scope (assumed local versus CCS83 zone, units, vertical datum).
2. Treat every GIS layer as having its own CRS. Reproject with known parameters, or do not overlay.
3. If the job stays assumed-local, do not use GIS coordinates as hold points. Use GIS only as pictures, then survey the features you will design.
4. If you need GIS and survey to live together long-term, put control on a published system (or publish a documented transformation) instead of hoping the GIS operator 'shifts it until it looks right.'

Grid-versus-ground and combined factors belong with the datums chapter. The GIS-planning point is smaller: do not mix an assumed job with mapped GIS and call it a survey.

What GIS is good for versus what still needs a field survey

GIS is good for: inventory of already-surveyed assets; exhibits; alternative-route screening; identifying which easements might exist; floodplain and environmental flags; estimating whether 3DEP lidar is even in the project area; communicating the survey footprint to the client.

A field engineering survey is still required for: design-grade existing ground; invert and rim elevations you will compute hydraulics on; as-built of new work; construction layout; locating marked utilities and potholed conflicts; any accuracy statement you will stand behind under 6731.1(d); configuration of the ground and position of fixed objects you are actually measuring under 6731.1(a)–(b).

Photogrammetry and lidar from a GIS portal can be the field-derived surface if they meet Section 4.1 and 4.2 specs and you add control and checks. Downloading a 10-meter DEM and labeling it 'existing ground for grading' is not that product.

Topology, layers, metadata, and source scale

Layers separate themes (parcels, water, storm, contours) so they can be turned on and off. That convenience hides source scale. A water main digitized from a 1 inch = 200 ft atlas is still a 1 inch = 200 ft line when you zoom to 1 inch = 20 ft. The GIS will draw it sharply. The sharpness is a lie.

Topology in GIS means connectivity rules: pipes meet at a node, polygons close, parcels do not overlap if the dataset claims a coverage. Topology is excellent for network inventory (can water flow from this main to that hydrant in the model?). It is not evidence that the pipes were surveyed in those positions. A topologically 'clean' parcel fabric can still be 8 ft off the iron pipes in the field.

Metadata should answer: Who compiled this? When? What CRS and datum? What accuracy test? What scale? FGDC-style metadata is professional practice for federal and many local datasets. If metadata is missing, plan as if the layer is an unlabeled photocopy—useful for reconnaissance, useless as control.

Scale of source data is an accuracy problem (Section 4.2) wearing a GIS costume. Interpolating 1-ft design contours from 10-ft USGS contours, or staking a curb from a parcel GIS vertex, violates the use-of-data rule: the application is finer than the source.

Civil-engineer limitation: parcels are not boundary surveys

This is the CES-relevant legal line, not a Professional Land Surveyor (PLS) exam course.

BPC 6731.1 authorizes a registered civil engineer to locate alignment or elevation for fixed works listed in BPC 6731, to determine configuration or contour of the earth's surface and the position of fixed objects, to create electronic data for those activities, and to state the accuracy of those maps and data. That is engineering surveying.

Determining property boundaries, setting missing lot corners, and preparing a Record of Survey are land surveying under the Professional Land Surveyors' Act. GIS assessor lines do not change that. A CE may not treat a GIS parcel as a field-surveyed boundary and stamp it as such. Civil engineers registered prior to January 1, 1982 (BPELSG notes the last such number as 33,965) may practice land surveying as defined in the PLS Act; that grandfathering is a license-status fact, not a GIS feature. BPC 6731.2 lets a CE offer to procure land surveying incidental to civil practice if the land-surveying work is performed by a person authorized to practice it.

Planning language that belongs in a CE survey scope: 'Assessor GIS shown for reference only; not a boundary survey.' Planning language that does not: 'Set the northeast lot corner from the GIS vertex.'

Exam traps: overlaying assumed local CAD on CCS83 GIS because both say feet; using FEMA polygons as pad contours; calling GIS topology a cadastral survey; citing parcel GIS as 6731.1 authority to set property corners; treating missing metadata as 'close enough for construction.'