Civil Earth-observation satellites have a fifty-year tradition that began with Landsat-1, launched in 1972. Today many public and commercial constellations imaging in the optical bands make it possible to observe the same area at different times and at different resolutions.

Orbit types

Most Earth-observation satellites operate in a sun-synchronous orbit (SSO). An SSO is a near-polar orbit, typically at 600–800 km altitude with an inclination of 96–99°, in which the orbital plane keeps a constant angle with respect to the Sun. As a result, every pass occurs at the same local mean solar time (e.g. 10:30) and illumination conditions remain consistent for time-series analysis. Geostationary (GEO) orbit, at 35,786 km altitude, is preferred for meteorological satellites.

Imaging principle

The vast majority of modern optical satellites use push-broom scanners: a linear CCD/CMOS sensor array oriented perpendicular to the flight direction continuously records the ground track as the satellite advances. Swath width directly determines the achievable revisit time — a wider swath means more frequent revisits, but usually at lower spatial resolution.

Common constellations

  • Sentinel-2 (ESA, 2015 and 2017): the A and B satellites operate at 786 km altitude with a 290 km swath, 13 spectral bands and 10/20/60 m resolution. With both satellites the equatorial revisit is 5 days, dropping to 2–3 days at mid-latitudes.
  • Landsat 8/9 (NASA-USGS, 2013/2021): 705 km altitude, 185 km swath, 11 bands (OLI + TIRS), 30 m (visible/NIR) and 100 m (thermal) resolution, 16-day revisit. Combined with Sentinel-2, the harmonized HLS product brings the revisit down to 2–3 days.
  • MODIS (Terra 1999, Aqua 2002): 2330 km swath, daily global coverage, 36 bands, 250–1000 m resolution. For broad-scale and climate applications.
  • Planet (Dove constellation): roughly 200 microsatellites; about 3 m spatial resolution and near-daily global coverage. Requires a commercial subscription.
  • SPOT 6/7, Pléiades Neo, WorldView, GeoEye: very-high-resolution (0.3–1.5 m) commercial satellites; revisit is tasking-driven because of the narrow swath.

The trade-off between resolution types

In satellite system design four resolution dimensions constrain one another:

  • Spatial resolution: the ground footprint a single pixel represents (e.g. 10 m).
  • Temporal resolution (revisit): the interval between repeat acquisitions of the same area.
  • Spectral resolution: the number and width of the recorded wavelengths.
  • Radiometric resolution: the brightness-level granularity, expressed in bits (8/12/16-bit).

High spatial resolution implies a narrow swath and infrequent revisits; broad coverage implies low spatial resolution. Mission design prioritises the application requirement (for crop monitoring, for example, spectral breadth tends to outweigh spatial detail).

Cloud cover and data availability

Optical satellites cannot see the surface beneath clouds, fog or smoke. The annual average cloud-free window depends on season and region; for Türkiye it is roughly 30–60%. The concept of cloud-free probability expresses the chance of obtaining a cloud-free image on a given date and is used in service-level agreements (SLAs). When optical data is unavailable, microwave sensors such as Sentinel-1 SAR provide a cloud-independent alternative; they measure different physical quantities, however.

Data levels: most satellite providers classify their data by processing level. For Sentinel-2, the L1C product is top-of-atmosphere reflectance (TOA reflectance), while the L2A product is surface reflectance (atmospherically corrected with the Sen2Cor algorithm). L2A is preferred for the accuracy of indices such as NDVI.