Imagery & point clouds
Data that arrives as measurements of a surface rather than as features on it.
Everything in this section arrives as a measurement of a surface rather than as a feature on it. There are no boundaries and no attribute table: a pixel has a value, a footprint and a date, and a LiDAR return has a position, an intensity and a classification. Turning that back into something with edges is usually the job.
The recurring difficulty is that the numbers are not what they appear to be until you know the product. A satellite band can be a stored integer, top-of-atmosphere reflectance or surface reflectance, and only the metadata distinguishes them. A NetCDF variable can be packed into sixteen bits with a scale factor that some readers apply and others do not. A LiDAR file can hold three returns for one pulse and a classification somebody else assigned.
Four topics: optical imagery, LiDAR and point clouds, radar, and gridded data with a time axis.
Topics in Imagery & point clouds
Each topic page brings that subject's concepts, how-to guides and fixes onto one page.
- Satellite imagery Bands, reflectance, cloud masks and composites — turning scenes into an array you can trust.
- LiDAR & point clouds LAS, LAZ and COPC files: returns, classes, density, and the surfaces you derive from them.
- Radar & SAR Sentinel-1 and radar imagery: backscatter, polarisation, speckle, terrain geometry, floods and change through cloud.
- Gridded & climate data NetCDF, GRIB and Zarr grids with a time axis: opening, subsetting, averaging and regridding them with xarray.
- Weather data Historical, station and forecast weather in Python: Open-Meteo, Meteostat, NOAA GHCN and ERA5, climate normals and anomalies, heat index, rainfall totals and weather maps, with the unit, time-zone and rate-limit traps.