How to Convert Coordinates into Terrain Data

Aki Olafsson

— 7 min read
Share
How to Convert Coordinates into Terrain Data

A pair of coordinates can identify a site precisely, but they do not yet describe the ground your building needs to sit on. To convert coordinates into terrain for BIM, you need three things: the correct horizontal reference system, reliable elevation coverage, and an export that your modelling software can interpret without moving, distorting or overloading the model.

For architects and landscape teams, this is not a GIS exercise for its own sake. It is the practical route from a survey reference, planning drawing or project address to an editable site model that supports levels, drainage studies, access routes and coordination.

Start by defining what the coordinates represent

Coordinates are only useful when their reference system is known. A number such as `55.6761, 12.5683` is likely latitude and longitude in WGS 84, while a pair such as `724000, 6179000` is more likely projected map coordinates in metres. Treating one as the other can place a site hundreds or thousands of kilometres from its true location.

Before generating terrain, establish whether you have geographic coordinates, projected coordinates, or a local project grid. Geographic coordinates use latitude and longitude in degrees. They are ideal for finding a location on a map, but they are not usually the best working system for measuring a site model. Projected coordinate systems convert the curved earth into eastings and northings, normally in metres or feet, so distances and areas can be modelled more predictably.

For a UK project, this may mean British National Grid data. In the US, the appropriate system could be State Plane Coordinates or a UTM zone. The right choice depends on the source data and the extent of the project. A compact building site can normally use its local projected system comfortably; a large infrastructure corridor requires more care because projection effects become more noticeable over distance.

The key rule is simple: retain the coordinate reference system supplied with the coordinates. Do not guess from the numbers alone where a drawing, survey control schedule or GIS file can confirm it.

Define a boundary, not just a point

A single coordinate tells you where to look. Terrain generation needs an area. Draw or enter a polygon that includes the building footprint, immediate landscape works and sufficient surrounding ground to understand how water and access approach the site.

The required margin depends on the decision you need to make. For an early massing study, 20 to 50 metres beyond the plot may be enough. For retaining walls, parking gradients or surface-water routing, extend the boundary to include the uphill catchment, downhill outfall direction and adjacent highway levels where available. If the model ends at the property line, you may miss the terrain that controls the site.

Avoid generating a very large area by default. More coverage means more points, larger files and slower BIM performance. A focused boundary usually produces a more useful terrain model than an oversized rectangle containing roads, woodland and distant slopes with no bearing on the design.

Check the site before exporting

Review the selected area against recognisable site features: road bends, existing buildings, watercourses and field boundaries. This quick visual check catches an incorrect longitude sign, swapped latitude and longitude, or a mismatch between a local grid and a global coordinate system.

Also check the elevation datum where the project depends on fixed levels. Terrain data may be referenced to a national vertical datum, while a survey may use a local benchmark or an assumed datum. The horizontal position can be correct while every elevation is offset. That may not matter for a conceptual model, but it matters when comparing finished floor levels, drainage falls or survey spot heights.

A worked example

Say a planning drawing gives you 55.6761, 12.5683.

  1. Identify the system. Two values near 55 and 12, to four or more decimal places, are latitude and longitude in WGS 84. This is a point in central Copenhagen.
  2. Turn the point into an area. For a feasibility study, draw a boundary roughly 150 m across, centred on the point, taking in the street frontage and the fall of the ground towards it. Keep it tight; you can widen it later.
  3. Choose a working grid. For a Danish site the natural projected system is ETRS89 / UTM zone 32N, in metres. Generate the terrain in that system, not in latitude and longitude.
  4. Pick a density. 2 m spacing is enough to read the slope and the street levels for a feasibility model. That is roughly 5,600 points over a 150 m square, well inside Revit's limit.
  5. Export. A CSV with X, Y, Z columns, comma delimited, period decimals, elevations in metres above Danish Vertical Reference (DVR90).
  6. Import and check. Bring it into the Toposolid, drop a spot elevation on the street centreline, and compare it against the level on the planning drawing. If they agree within a few centimetres, the coordinate system and datum are right.

The same six steps apply whether the input is a lat/long pair, a British National Grid reference or a US state plane coordinate. Only steps 1 and 3 change.

Retrieve elevation data at a suitable resolution

Once the boundary is correctly located, the next task is to retrieve elevation values across it. This is where generic 3D map views often fall short. They may look convincing, yet provide no editable points, no stated coordinate system and no dependable basis for design decisions.

A usable terrain dataset contains XYZ values: an easting or X coordinate, a northing or Y coordinate, and a height or Z coordinate. The point density should reflect both the available source data and the purpose of the model.

LiDAR-derived data can represent terrain in considerable detail, particularly on open ground. Photogrammetry and official national mapping datasets may provide useful wider coverage where LiDAR is unavailable. However, source resolution is not the same as the point spacing you should import into Revit, SketchUp, ArchiCAD or Rhino. A dense cloud may be technically accurate but unnecessarily heavy for an architectural site model.

For most building-scale BIM work, select a spacing that preserves meaningful breaks in slope while keeping the model responsive. A broad, gently sloping site can often use fewer points. Stepped ground, embankments, ditches, kerbs and drainage channels may need closer sampling, or a separate detailed model in the area being designed.

Topo-grapher is useful here because it turns a drawn boundary or supplied coordinates into editable terrain data from available authoritative elevation sources, rather than leaving the team with a visual web model.

Clean the data before it enters BIM

Raw elevation points are not automatically ground points. Depending on the source, they can include trees, roofs, bridges, walls or other surface features. A terrain model generated from surface data rather than bare-earth data may show a wooded site as a series of high mounds, or carry a road bridge across a valley as though it were the ground level.

Use terrain or ground-classified data where the workflow requires existing landform. Then inspect the results at known locations. Compare a few elevations against survey spot levels, published contours or visible site conditions. This is a validation step, not a substitute for a site survey.

Where you have a current topographic survey, it should normally control construction-critical geometry. Public elevation datasets are valuable for feasibility, planning context, early-stage grading and areas outside the survey extent. They may be older than recent earthworks, and their stated accuracy may not be adequate for setting out or certifying levels.

You should also decide whether to preserve absolute coordinates in the model. Real-world coordinates make civil and consultant coordination easier, but very large easting and northing values can cause precision or display issues in some software. The practical approach is often to retain the source coordinates in the dataset and establish a shared coordinate workflow in the authoring model, rather than arbitrarily moving terrain to an unrecorded local origin.

Export XYZ data in the format your software expects

A clean CSV or plain-text XYZ file is a dependable interchange format because it is transparent. Each row represents one point, usually in the order X, Y, Z. Before import, confirm the delimiter, decimal format, unit and coordinate order. A file that uses commas as decimal separators, for example, may be read incorrectly by software expecting commas between fields.

Keep units consistent. If your terrain is exported in metres but the receiving project uses millimetres or feet, the resulting model may be wildly too small, too large or vertically exaggerated. The same applies to elevation: establish whether heights are metres, feet, or a local project unit before import.

For Revit, rename the file to .csv and use Massing & Site > Toposolid > Create from Import > Create from Points File; keep it under 10,000 points, and verify shared coordinates before placing it alongside consultant models. If the point count is high, reduce it before import rather than asking the project file to carry more detail than it needs.

In SketchUp, XYZ points need an importer extension (SketchUp has no native point import); once triangulated, inspect the mesh around sharp grade changes and site edges. In Rhino and Grasshopper, points can be used to create a surface or a triangulated mesh, giving more control over filtering, contours and later analysis. ArchiCAD Mesh workflows similarly benefit from an organised point file and a clear decision about which critical levels need manual confirmation.

Validate the terrain model in context

The final check is not whether the file imported successfully. It is whether the terrain makes spatial sense alongside the design. Place a section through the site and test the levels that matter: pavement to entrance, car park fall, retaining-wall height, accessible route gradient and the relationship between proposed floor level and existing ground.

Look for warning signs such as isolated spikes, flat plateaux where slopes should continue, drainage flowing uphill, or terrain that appears offset from roads and buildings. These problems usually point to one of four causes: an incorrect coordinate system, a vertical datum mismatch, unsuitable source data or an import setting error.

When the terrain is correctly located and proportioned, it stops being background context. It becomes a working design input: something the team can cut sections through, coordinate with civil information and use to make earlier, better decisions about the ground beneath the project.

From coordinates to terrain in one step

Topo-grapher does steps 2 to 5 for you: enter a coordinate or draw a boundary, pick the point spacing, and download a clean X, Y, Z CSV in the projected system and vertical datum for that country, from authoritative elevation datasets. No GIS software, no tile hunting, no manual reprojection. The datum and unit checks above still belong in your authoring model.

Generate terrain from a coordinate →

Read more