Course · ← 0015 LADM In Depth · 0017 Adjustment → · depth track

Lesson 0016 · ~15 minutes

The Third Dimension

A flat parcel cannot express a flat. Apartments, basements, tunnels, air rights and utility corridors are all rights over volumes — and the model that handles them is the one you already know, extended by one class.

Why 2D runs out

Stack twelve apartments on one footprint and the 2D cadastre records one parcel with twelve owners and no way to say who owns what. Run a metro tunnel under a private plot and the surface owner's rights and the transit authority's rights coexist in the same map coordinates. Sell the air above a railway for development and there is nothing on a flat map to sell.

These are not exotic cases. They are urban land administration, and they are where donor programmes go once rural coverage is done.

A correction to something said earlier

3D is not a separate part of LADM. It lives in the core model — Edition II was explicitly designed with 3D aspects throughout.1 Part 3 is marine geographic information; Part 4 is valuation. If you want the 3D machinery, you want Part 1, the same document you already needed.

One class extends the model

You met the boundary classes in lesson 0015. The 3D story is entirely contained in the difference between two of them:

LA_Point               a corner, in 2D or 3D
LA_BoundaryFaceString  represents a 2D LA_SpatialUnit   ← a parcel outline
LA_BoundaryFace        represents a 3D LA_SpatialUnit   ← a face bounding a volume

LADM uses LA_BoundaryFaceString to represent a 2D spatial unit and LA_BoundaryFace to represent a 3D spatial unit such as a 3D parcel.2 Everything else — party, RRR, BAUnit, sources, versioning — is unchanged. A 3D parcel is a spatial unit like any other; only its bounding representation differs.

Two further classes give the common cases names. LA_LegalSpaceBuildingUnit defines a building unit with legal spaces that may serve different purposes — living or commercial — where a building, or part of it, is not equivalent to the physical separation.2 And LA_LegalSpaceUtilityNetwork covers pipes and cables, which is lesson 0014's network reappearing as a rights-bearing volume rather than a graph.

Where the geometry comes from

3D legal spaces have to be sourced from somewhere, and the two candidates come from a different industry than GIS.

StandardIsRole here
IFCOpen BIM — how buildings are designed and documentedThe as-designed geometry. Cartesian points map to LA_Point; polyloops formed from those points equate to LA_BoundaryFaceString; connected faces to LA_BoundaryFace.3
CityGMLCity modelling, with levels of detailThe city-scale context — how the registered volume sits among its neighbours
LADMThe rights modelWhat is owned, by whom, on what basis

That IFC mapping is the practical route into 3D registration: developers already produce IFC models for permitting, so the geometry exists before the registry needs it. The work is extracting legal spaces from a physical model — and deciding who certifies that extraction.

What actually gets hard

The data model is the easy part. Real implementations report difficulty in creating 3D legal spaces, extracting boundary faces, representing vertical extents such as airspace rights, and applying virtual boundaries.4 Unpacked, that is four distinct problems:

  1. Vertical extent is often unstated Surface ownership rarely says where it stops going up or down. Countries answer this differently, sometimes not at all — and your model must represent "unbounded above" without pretending to a number.
  2. Virtual boundaries have no physical referent The plane dividing floor 3 from floor 4 is legal, not material. Nothing to survey, nothing to photograph — the document is the only evidence there is.
  3. Volumes must close A 2D polygon that fails to close is obvious. A solid with one missing face is not, and it will pass every check you did not write. Validation is genuinely harder here.
  4. Almost no tooling assumes 3D PostGIS handles 3D geometry, but topology, most web clients and most editing tools are planar. Budget for this rather than discovering it.

Which leads to the honest scoping advice: for a fit-for-purpose rural programme, 3D is out of scope and saying so plainly is a strength. For an urban formalisation or condominium register, it is the requirement — and pretending 2D will do just defers the rewrite.


Drill 1 — 2D or 3D?

Drill 2 — Recall


Your primary source

"Design of the new structure and capabilities of LADM Edition II including 3D aspects", Land Use Policy — how 3D was worked into the core model rather than bolted on:
sciencedirect.com/science/article/pii/S0264837723004696

Then, for the BIM route into it, "Working with Open BIM Standards to Source Legal Spaces for a 3D Cadastre" (open access) and "Utilizing BIM and GIS for Representation and Visualization of 3D Cadastre".

Ask your teacher. Worth pushing: "how do I store and validate closed solids in PostGIS?", "who certifies that a legal space matches the IFC model?", "what does a condominium register look like as tables?"