MAPS & PLACES / SATELLITES
What a satellite image
actually shows.
A colorful satellite layer has been measured, processed, and styled before it appears in your browser.
THE MAIN IDEA
Pixels are instrument measurements, not transparent photographs of truth.
An Earth-observation sensor records reflected or emitted energy in particular parts of the electromagnetic spectrum. Visible-band images approximate what a camera might see from above, but infrared and other bands reveal properties the human eye cannot see. A displayed image often combines bands, adjusts brightness and contrast, removes cloud-contaminated pixels, and mosaics observations from different dates. False-color composites deliberately map invisible bands into visible colors to make vegetation, moisture, or other features easier to study. Those colors do not mean the ground literally looks that way. Spatial resolution describes the ground area represented by a pixel; a pixel does not always describe a homogeneous patch. It may blend roofs, trees, and roads together. Spectral and temporal resolution also matter: the sensor measures a defined set of wavelengths and revisits the same place on a particular schedule. Cloud, smoke, haze, season, sun angle, shadows, and viewing angle affect what is recorded. A dark area might be water, a shadow, a burned surface, or an imaging artifact; appearance alone is not classification. Machine-learning systems can label land cover, but their errors depend on training data, chosen classes, and local validation. A classification of “forest” cannot reveal species or ecological quality at a glance. Check the acquisition date, sensor, resolution, processing level, and legend before interpreting change. A side-by-side comparison must account for seasons and sensor differences; a spring image and an autumn image may show vegetation changes unrelated to construction or damage. Ground validation uses trustworthy observations from the place and time to test an image interpretation. Even high-resolution images cannot reliably reveal every human activity or private detail; avoid using remote imagery to infer sensitive personal behavior. Satellite data are powerful in climate, land-use, agriculture, disaster response, and ocean research because repeated measurements reveal large patterns, but a model is only as robust as its data and assumptions. For urgent hazards, use current official alerts and local response guidance, not an unfamiliar commercial image mosaic. Responsible interpretation states what the sensor measured, what was inferred, when it was observed, and what uncertainty remains. The image is evidence from a particular vantage, not an all-seeing window.
TRY THIS
Check the date before comparing colors.
Open a public satellite image layer and find its acquisition date, legend, and stated resolution. Compare two images only if seasons and processing are reasonably similar. Note at least one explanation for a color difference other than a real surface change.