LiDAR HD IGN and Urban Canopy Estimation

Since 2021, IGN has been progressively releasing LiDAR HD, an aerial dataset that covers the whole of France with a density of 10 points per square metre. For local authorities, consulting firms and urban environment researchers, it is a public resource of considerable value, available free of charge on geoservices.ign.fr. Yet many municipalities are still unaware that they hold, over their own territory, operational 3D mapping that can be used to quantify their canopy, identify their urban heat islands or monitor the tree-related obligations set out in their local urban planning regulations.

This article summarises how to move from raw LiDAR data to a rigorous estimation of the urban tree cover, and why this approach changes the way a canopy strategy is steered.

LiDAR HD IGN: an under-used public resource

The LiDAR HD programme targets nationwide coverage by 2026. Data is distributed as one-square-kilometre tiles in compressed .laz format, under the open Etalab licence. Each point carries X, Y, Z coordinates in Lambert 93, a return intensity and, most importantly, an automatic ASPRS classification: ground (class 2), low, medium and high vegetation (classes 3, 4, 5), buildings (class 6), water (class 9). This classification, already computed by IGN, saves weeks of processing compared with a raw point cloud.

In practice, downloading the tiles for a 5 km² municipality represents a few hundred megabytes and thirty minutes of connection time. The real work begins afterwards, when these points are transformed into raster models that can be exploited in a GIS.

From raw data to operational models

Three models structure any canopy analysis. The DTM (Digital Terrain Model) represents the bare ground, generated from points classified as ground only. The DSM (Digital Surface Model) represents the visible surface seen from above, the first return of each cell. The difference DSM minus DTM gives the nDSM (normalized Digital Surface Model), which isolates the actual height of objects above the ground. This nDSM is the central model for canopy analysis: a 12-metre pixel corresponds to a mature tree, a 3-metre pixel to a shrub, a 0-metre pixel to bare ground.

The classical processing chain combines PDAL for point cloud filtering and rasterisation, GDAL for raster operations, laspy for fine-grained manipulations in Python, and geopandas to cross the result with the cadastre or urban planning zoning. Once the nDSM is produced, a threshold at 3 metres extracts the tree footprint, and a per-parcel area calculation yields the canopy rate.

Application: estimating the urban canopy of a municipality

The study I carried out on the municipality of Grigny illustrates this complete workflow. Seven thematic maps were produced: raw nDSM, thresholded canopy height, canopy cover rate per IRIS block, intersection with public spaces, cross-analysis with fuel poverty zones, identification of potential urban heat islands and priority effort zones. The i-Tree Eco model from the US Forest Service was then applied to estimate the ecosystem services delivered: tonnes of CO2 sequestered annually, cubic metres of rainfall intercepted, average cooling effect over the dense urban area. All technical and visual details are on the project page Grigny LiDAR study (i-Tree Eco).

The work was the subject of an academic defence on 4 May 2026, as part of my ongoing Master 1 Geomatics. A scientific co-publication proposal was subsequently received from Prof. Nowak; the discussion on scope and schedule is in progress.

Use cases for local authorities

A canopy estimate is not a style exercise. It feeds several operational documents: the green and blue infrastructure plan written into the SCoT, a local canopy strategy with a quantified ten-year cover target, the monitoring of tree compensation obligations set in the PLU, or the prioritisation of plantings in the neighbourhoods most exposed to thermal stress. Cross-referencing the canopy with urban heat islands derived from Landsat or Sentinel also makes it possible to identify where each newly planted tree will deliver the greatest public-health benefit. For a complete approach combining LiDAR, satellite imagery and vegetation indices, see my work in remote sensing.

Technical tools and lead times

For a medium-sized municipality (5 to 15 km²), a LiDAR study targeted on a precise perimeter (a neighbourhood, a park, a development zone) takes two to three weeks, starting from 1,500 €. A full canopy study at the municipal level, with thematic maps, i-Tree Eco computations and printable cartographic deliverables, takes three to four weeks, starting from 4,000 €. The final price depends on the area covered, the density of the point cloud and the level of cross-analysis with local datasets (cadastre, PLU, social data): personalised quote within 48 hours after a free scoping call. To discuss a concrete project, the contact page is open.

For a full breakdown of every service and price range, see the Services & Pricing page.