AirfieldpathOrlando Sanford International

6.11 · Data and sources · Data

USGS 3DEP lidar

Public lidar over the airfield. Every runway end agrees with the FAA published elevation to within 15 centimetres.

No picture for this component.

Data source · Data and sources · Data

USGS 3DEP lidar

Public lidar over the airfield. Every runway end agrees with the FAA published elevation to within 15 centimetres.

  1. Sources
  2. Interchange
  3. Twin
  4. Screens
  5. Plan
Kind
Data source
Status
not linked
Airport layer
not linked
Airport area
not linked
System
not linked
Sensor type
not linked
How it worksHow it works: Data

Open the detail

Level 2 of 3 This component, complete on its own. Its files, sources and tests are the detail below. Open the detail

Open the airfield in 3D

The web files behind the screen airfield3d ("The Airfield in 3D, from Lidar"). They are built from public USGS lidar by build_sfb.py with lidar-kit (lidar-kit), the lidar module shared by Airfieldpath, Municipath, SprayPath and Field Twin.

The source

ProjectFL_Peninsular_Seminole_2018, a USGS 3DEP work unit of FL_Peninsular_2018_D18
Flown2018-12-04 to 2019-01-09 (metadata begdate and enddate)
ProducerQuantum Spatial, Inc., for USGS 3DEP
QualityQL1, nominal pulse spacing 0.35 m, produced to the ASPRS (2014) 10-cm RMSEz vertical accuracy class
Density measured here22.1 returns per m2, 15.7 of them ground, over the terminal box (458,814 m2, every point the EPT holds)
DatumNAD83(2011) and NAVD88 (Geoid12B) as delivered; the USGS EPT serves it in EPSG:3857 with heights in metres
LicencePublic domain. Metadata: access "No restrictions apply to these data"; use "None"
Read2026-10-30, from https://s3-us-west-2.amazonaws.com/usgs-lidar-public/FL_Peninsular_Seminole_2018/ept.json
Metadatahttps://prd-tnm.s3.amazonaws.com/StagedProducts/Elevation/metadata/FL_Peninsular_2018_D18/FL_Peninsular_Seminole_2018/reports/vendor_provided_xml/Seminole_Classified.xml

It is the only project in the USGS EPT index (hobuinc/usgs-lidar, boundaries/resources.geojson, read 2026-10-30) that covers the FAA airport reference point, 28-46-37.841N 81-14-5.689W.

Checked against the FAA. At each of the eight runway ends, the median of the ground returns within 8 m is within 0.15 m of the elevation NASR publishes, and seven of the eight are within 0.06 m. The table is on the screen and in sfb-analysis.js (runwayEndCheck).

Colour is USDA NAIP orthoimagery from USGS The National Map (USGSImageryOnly, service refreshed June 2024, public domain), read 2026-10-30. It is a picture laid on the points and the ground. It was taken years after the lidar, so what changed since 2019 shows in it.

Footprints are OpenStreetMap's (ODbL), from poc/build/sfb-osm.json.

What is measured and what is not

  • Measured: ground, building heights (303 of 303 OSM footprints, Terminal A 11.2 m to the 95th percentile of its roof). The runway profiles and slope come from the same ground model, as does the ponding analysis.
  • Analysis, not observation: ponding marks closed hollows in the 2018-19 surface where water would stand if nothing drained. Drains and inlets the lidar cannot see show as hollows. Depths near 10 cm sit within the stated accuracy class. The pavement mask uses FAA widths for runways, assumed widths for taxiways (75 ft) and taxilanes (50 ft), and OSM aprons. Cells green in the imagery are removed, and the mask stops 3 m short of every edge.
  • Illustrative: inside the terminal no lidar exists. AHU-3, Escalator A and Elevator 5 are placed by hand inside Terminal A and say "position illustrative" until floor plans or an interior scan arrive.

Files

FileWhatSize
sfb-ground.terrain.js4 m bare-earth surface of the whole box, imagery and hillshade laid on it7.3 MB
sfb-apron.terrain.js1 m bare earth over the apron, the terminal and the middle of 9L/27R, finer imagery9.2 MB
sfb-airfield.*.js2,390,289 returns above the ground (EPT depth 8; buildings in 1.5 m voxels, the rest in 5 m)35.1 MB
sfb-terminal.*.js1,791,932 returns around the terminal (every point, 0.55 m voxels)26.3 MB
sfb-analysis.jsHeights, ponding candidates, runway profiles, the FAA check, the sourceunder 1 MB
sfb-ponding-layer.jpgHillshade of the 1 m ground with ponding depth, for the layered runway map0.7 MB
lidar-viewer.js, three-r149.min.jsCopied from lidar-kit by the build0.6 MB
capture.mjsFilm capture of the fly-in
shots/The screen in light and dark

About 78 MB in all. Every data file is a script, so the page loads it from disk (file://), as the film recorder opens it.

Regenerate

The raw data stays outside git, in ~/.stellarview/lidar-data (about 2.4 GB with the cache).

K=~/Forks/lidar-kit/bin/lidarkit
EPT=https://s3-us-west-2.amazonaws.com/usgs-lidar-public/FL_Peninsular_Seminole_2018/ept.json
D=~/.stellarview/lidar-data
O=-81.23491361,28.77717805                                    # the FAA airport reference point
$K fetch --ept $EPT --bbox=-81.27,28.76,-81.21,28.80 --depth 8 --origin=$O --cache $D/cache --out $D/sfb/airfield-d8.npz
$K fetch --ept $EPT --bbox=-81.2625,28.7635,-81.2125,28.788 --depth 9 --origin=$O --cache $D/cache --out $D/sfb/core-d9.npz
$K fetch --ept $EPT --bbox=-81.2475,28.7735,-81.2390,28.7785 --depth 11 --origin=$O --cache $D/cache --out $D/sfb/terminal-full.npz

The imagery is four exports of https://basemap.nationalmap.gov/arcgis/rest/services/USGSImageryOnly/MapServer/export (bboxSR=3857, 4096 pixels wide, JPEG) into ~/.stellarview/lidar-data/sfb/img/: wide.jpg over the first box above, core.jpg over the second, terminal.jpg over the third, and apron.jpg over -81.2530,28.7700,-81.2330,28.7840. The exact URLs are in img/urls.txt. Then:

~/.stellarview/lidar-venv/bin/python poc/lidar/build_sfb.py    # about 3 minutes
python3 poc/build/assemble.py

The fly-in, for film

?v=airfield3d&film=1&step=1 starts the 32 second fly-in on the film clock: from high over the airfield, down the length of 9L/27R, across the apron, to the terminal's airside face. Once window.__lidar.ready is true, each window.__filmStep(ms) draws the frame at ms, and two runs give byte-identical frames (checked on 2026-10-30). window.__lidar.renderAt(ms) draws a frame directly, for the broadcast.

node poc/lidar/capture.mjs --stills <dir> 0,16000,32000 dark --full    # stills framed as the film frames them
node poc/lidar/capture.mjs --video <out.mp4>                           # 1920 x 1080, 30 fps, picture only

The video has no audio stream. It is a plate for the film to cut in under Sue's voice and the bed.