Sentinel-2: read landscapes across spectral bands

Sentinel-2 carries a multispectral imager with thirteen bands spanning visible, near-infrared and shortwave-infrared wavelengths. The first spacecraft launched in 2015, followed by other members that support continuing land observations. Its design draws on earlier SPOT and Landsat experience while providing its own combination of coverage, spectral channels and repeat visits.

Different surfaces reflect different fractions of incoming light across those bands. Vegetation, water and built areas can therefore be investigated using combinations beyond a natural-looking color view. The Semarang image in this lesson is a processed true-colour display from Sentinel-2B. It helps orient the reader, but the displayed colors do not expose every band or every derived land product.

The thirteen bands do not all have the same spatial resolution: four use 10-metre sampling, six use 20-metre sampling and three use 60-metre sampling. Their roles include land observations and atmospheric correction. A statement that Sentinel-2 is a 10-metre mission is incomplete if it suggests that every band independently measures detail at that scale.

Suppose a fictional analyst combines a 20-metre band with a 10-metre band. Resampling can place both on a common grid for calculation, but it does not create new independent measurements in the coarser band. A useful result documents which original bands were used, how they were aligned and what the resulting spatial detail can actually support.

Land-reflectance observations depend on illumination and can be obscured by clouds or affected by atmosphere and shadows. Processing levels describe different corrections, and quality information helps identify unusable pixels. A nominal repeat opportunity is not a promise of a cloud-free surface observation each time the satellite can view an area.

For a fictional vegetation-change exercise, compare compatible products acquired at meaningful times in the growing cycle. A difference could reflect actual land change, seasonal development or different observing conditions. The single coastal scene shown here can reveal spatial patterns; establishing change needs further dated observations. This makes the archive’s calibration and acquisition information as valuable as the image itself.