PNOA-LiDAR: what it is, the third coverage and how to use it

What PNOA-LiDAR is

PNOA-LiDAR is the project within Spain’s National Aerial Orthophotography Plan (Plan Nacional de Ortofotografía Aérea) that uses aircraft-mounted LiDAR sensors to capture three-dimensional information across the whole of Spain. The result is point clouds with X, Y, Z coordinates, plus the digital terrain and surface models derived from them.

It is coordinated by Spain’s National Geographic Institute (Instituto Geográfico Nacional, IGN) and co-funded by several bodies: the IGN itself and its National Centre for Geographic Information (CNIG), ENAIRE (the air navigation service provider), FEGA (the Spanish agricultural guarantee fund), the Dirección General del Catastro (the land registry), the Dirección General del Agua (the water directorate) and the autonomous communities (Spain’s regional governments). That is why the same data serves the land registry, forest inventory, flood zones and mapping.

If this is your first time working with points, start with what a point cloud is: here we take density, accuracy and class as read.

First, second and third coverage

A coverage is a complete pass over the whole country. There have been three so far, and each one multiplies the density of the last:

Source: PNOA-LiDAR technical specifications.

The third coverage began in 2021 in Catalonia and in 2023 in the rest of the autonomous communities. According to the IGN status table of August 2026, the flying is finished (the last blocks, in Galicia, Asturias, Castilla y León and Castilla-La Mancha, were flown between January and March 2026), but publication is still in progress: every community now has a downloadable cloud, though most with the basic classification and only Cantabria, the Balearic Islands (Illes Balears) and Navarre with the enhanced one.

What each file contains

• LAZ format (LAS 1.4, point data format 8, compressed), in 1×1 km tiles.
• Reference system ETRS89 (REGCAN95 in the Canary Islands), UTM projection in the official zone and orthometric heights on the EGM2008-REDNAP geoid. Tiles that fall on the boundary between two zones are delivered in both.
• Colour and infrared: the flight carries a simultaneous camera, and first-return points carry RGB and near infrared, which is useful for separating vegetation.
• Up to four returns per pulse: the laser passes through gaps in the tree canopy and also records the ground beneath.
• Classification at three levels, depending on how far publication has progressed in each area:
  • NPC01, automatic: ground, low, medium and high vegetation, building, noise and overlap.
  • NPC02, edited: adds water, bridges, vehicles, wires, transmission towers, wind turbines, solar panels and antennas.
  • NPC03, enhanced: adds walls and dams, and splits buildings into flat, pitched or green roof, façades and courtyards.

The classes follow the ASPRS LAS numbering (2 ground, 3-5 vegetation, 6 building, 7 noise, 9 water…) plus the IGN’s own codes; they are listed in the class definition for the third coverage. Besides the cloud, the IGN publishes the DTM (digital terrain model, ground only) and the DSM (surface model, with buildings and trees), both on a 50 cm grid.

What it is for

Its strength is coverage: any plot in Spain, free, georeferenced and with the terrain measured beneath the vegetation. Where it performs best:

• Reference and context. The surroundings of a project in real coordinates: the topography of the plot, the heights of neighbouring buildings, the trees. It fits together with a denser cloud of your own that covers only the area that genuinely needs measuring.
• Planning fieldwork. Before going out, you can study access, level changes and obstacles. For a drone flight it is especially useful: the wire, transmission tower and wind turbine classes, along with tree height, help set the flight altitude and take-off areas.
• Preliminary studies and outline design. Profiles, slopes, catchments and approximate earthwork volumes with the 50 cm DTM, before commissioning a detailed topographic survey.
• Comparing over time. With three coverages, you can see changes in terrain, woodland or quarry faces over a scale of years.
• Territorial studies: flood risk, forest inventory, rooftop solar potential, visibility or shading at municipal scale.

The rule of thumb: PNOA-LiDAR answers “what is the terrain like and what is around it” well. It answers “exactly how big is this” badly as soon as “this” is a building, a facility or a structure.

How to download it

1. Go to the CNIG Download Centre (Centro de Descargas del CNIG), LiDAR third coverage series.
2. Find the area by autonomous community, province or municipality, by MTN50 map sheet, by cadastral reference or by coordinates.
3. Select the 1×1 km tiles you need and download them. If your area sits on the boundary between UTM zones, choose the version in the zone you work in.
4. If you only need the terrain, download the DTM instead of the cloud: it is much lighter and already comes clean.

To view the data without downloading it, the IGN publishes DTM web services (WMS, WMTS and WCS) that load straight into QGIS or any GIS. There is no official 3D viewer for the point cloud, so to explore it you have to download it.

Licence: CC BY 4.0. It can be used even for commercial purposes, crediting the source alongside the data (short form: “CC-BY 4.0 scne.es” and the date). If you modify it, state that it is a derivative work.

How to open it in your tools

A 1×1 km LAZ tile opens in almost any point cloud software. The important thing is not to lose the coordinates along the way:

• VisionHub: opens LAS and LAZ in the browser, alongside the IFC model and a more detailed cloud of your own. It is the most convenient way to share the PNOA tile with a team that has no point cloud software.
• QGIS: reads LAS/LAZ natively, in 2D and 3D, and lets you filter by class and overlay it with the cadastre or orthophoto. The natural choice if you work in GIS.
• CloudCompare: free, for cropping, filtering by class, cleaning noise or comparing two coverages.
• Revit, AutoCAD and Civil 3D: first convert it to RCP with Autodesk ReCap. Because it is in UTM, the coordinates are very large: link it using shared coordinates or a base point, not “at origin”.

You will find the full software comparison in BIM and point cloud viewers, and the formats (LAS, LAZ, E57, RCP) explained in the point cloud guide.

Noise, gaps and other limits

All of the above comes with small print. What follows is our technical reading after working with these clouds, not an assessment by the IGN:

• Density. Around 5 points per m² works out at roughly one point every 45 cm. Enough for terrain, not enough for a cornice, a window opening or a pipe.
• Lack of vertical information. From an aircraft at high altitude you see roofs and ground. Façades, arcades, the undersides of eaves and bridges, and the inside of narrow courtyards end up with few points or none.
• Noise. Stray points in the air (birds, particles, reflections) or below ground from multiple bounces. Much of it is flagged as class 7, but not all of it, and in areas with automatic classification it is worth checking before calculating anything.
• Gaps. Water, low-reflectance materials and areas hidden by other objects can end up with no data, and the specification itself acknowledges this.
• Classification still under review. Where only the NPC02 level has been published, the IGN warns that work is ongoing to complete the classes. In mountain areas the required accuracy for vegetation is low, so ground and scrub get confused.
• Overlaps. Where two strips cross there are more points and sometimes a small step between them; that excess is flagged as class 12.
• Date. Depending on the area, the cloud dates from between 2021 and 2026. A construction site, a stockpile or an industrial building built afterwards will not appear.

When it is not enough: high-density point clouds

If the project needs genuine measurement (a survey for design, an as-built, a clearance check, construction monitoring), PNOA-LiDAR stays as reference and you need your own capture. That is what we do at LAURIA ESCÁNER:

• Drone LiDAR: the same principle as the IGN flight, but a few dozen metres above the ground. More than a million points per hectare, 2-3 cm accuracy with RTK and up to 600 ha in a day. It sees through vegetation and captures embankments, linear works or quarries in a detail PNOA cannot provide.
• Terrestrial laser scanning and SLAM: where the drone cannot reach. Façades, interiors, beneath roofs, building services and structure, with accuracy in the order of ±2 mm.
• Drone photogrammetry: when colour and a high-resolution orthophoto of roofs, façades or faces are the priority.

The usual approach is to combine them: drone outside, scanner inside and on façades, all registered into a single cloud in ETRS89 / UTM that fits without extra work with the PNOA tile you already have as context. That cloud is the basis for BIM modelling, extracting drawings or building a digital twin, and it is delivered in LAZ, E57 and RCP, ready to open in VisionHub.

If you have downloaded the PNOA-LiDAR for your area and it is not enough, tell us what you need to measure and to what accuracy: we will tell you which capture you need and which you can do without.