Use Lidar HD data

[eddy-geek]

• Raw data point (LAS) are available (status acquisitionslidarhd)
• Classification done (for the alps, pyrenées, jura) as of December '23 - status traitementslidarhd
• Couverture

Unfortunately it seems tricky to exploit:

URL example https://wxs.ign.fr/c90xknypoz1flvgojchbphgt/telechargement/prepackage/LIDARHD_PACK_SP_2021$LIDARHD_1-0_LAZ_SP-1036_6295-2021/file/LIDARHD_1-0_LAZ_SP-1036_6295-2021.7z

These 7z files cover 2km² by including 4 laz files of 1km², average 90 Mb (≈.1 Gb).

Downloading 1 files takes ~30minutes (100-200KB/s) so 8 minutes/km² (may improve with the geoplatform?)

Alps are 200K km² of which 40K km² are in France, of which half is nice-to-have from a quick area calculator.

so, 20000 .1 Gb = 2 TB of storage ; and .1hr = 2K hours = 3 months of download time

~Also, because the points are not classified, we would have to only either use data above the tree line (1800m to 2300 N->S), or use heuristics to minimize vegetation impact (eg keep only the points of minimum altitude in each m²?)~ edit: classified data available

So, overall, better to wait for the DEM (MNT) to be available.

[eddy-geek]

seems to have gotten faster now, the classified tiles listed on geoservices lidarhd have better hosting:

wget https://storage.sbg.cloud.ovh.net/v1/AUTH_63234f509d6048bca3c9fd7928720ca1/ppk-lidar/QN/LHD_FXX_0964_6431_PTS_O_LAMB93_IGN69.copc.laz

LHD_FXX_0964_6431_PTS_O_ 100%[=================================>]  85,56M  33,6MB/s    in 2,5s    

(still not referenced in the geoplatform's capabilities)

[eddy-geek]

Le nuage de points est séparé en 11 classes : Code, Désignation: 1 Non classé 2 Sol 3 Végétation basse 4 Végétation moyenne 5 Végétation haute 6 Bâtiment 9 Eau 17 Pont 64 Sursol pérenne 65 Artefact 66 Points virtuels (modélisation)

( source DT_LiDAR_HD_1-0.pdf )