Data fundamentals
Revisit, resolution, and cloud cover: what shapes satellite data freshness
Satellite imagery is collected on orbits, through an atmosphere, at a fixed pixel size. Those three facts decide how often you can see a place, how small a change you can detect, and how often the view is blocked.
2 minute read. Written by the GeoVeyra team.
Revisit time
Revisit time is how often a satellite or constellation can observe the same location. It depends on the orbit, the width of the imaging swath, and how many satellites share the job.
For example, the Copernicus Sentinel-2 mission was designed for a revisit of about five days at the equator with two satellites in operation, and Landsat 8 and 9 together offer an eight-day revisit. Higher latitudes are often seen more frequently because orbits converge toward the poles.
Spatial resolution
Spatial resolution is the ground size represented by one pixel. Sentinel-2 records its visible and near-infrared bands at 10 metres and other bands at 20 or 60 metres. Landsat 8 and 9 record most multispectral bands at 30 metres.
As a rule of thumb, a feature needs to span several pixels to be identified reliably. A 10-metre image can show a new building cluster, but not an individual car. Commercial imagery can offer finer detail where licensing allows.
Cloud cover
Optical sensors measure reflected sunlight, so clouds and their shadows block the view. In persistently cloudy regions and seasons, usable optical observations can be far less frequent than the revisit time suggests.
Synthetic aperture radar, such as Sentinel-1, sends its own microwave signal and can image through cloud and at night. It shows surface structure and moisture rather than color, so it complements optical imagery rather than replacing it.
Processing and product levels
After acquisition, raw data is processed into products. Sentinel-2, for instance, is distributed as Level-1C top-of-atmosphere reflectance and Level-2A surface reflectance, which corrects for atmospheric effects and is usually better for comparing dates.
Processing and distribution take time, and that time varies by provider and product. Freshness is best described by the acquisition timestamp, not by the moment an image appears in a viewer.
What this means for monitoring
Plan monitoring around the physics, and communicate it honestly.
- Show acquisition dates on every comparison
- Expect gaps in cloudy seasons and plan for radar where it helps
- Match the resolution to the smallest change that matters
- Treat alerts as new evidence to review, not instant ground truth
Want to see these ideas applied to your area of interest? Book a demo or browse the glossary.