What a Satellite Observes
A satellite instrument records electromagnetic radiation reflected or emitted by the surface and atmosphere. A satellite image is not necessarily an ordinary photograph; it may contain multiple spectral bands, radar backscatter, surface temperature, or atmospheric properties. After geometric and radiometric processing, observations are transformed into maps, indices, and time series. Each product has a method description, units, spatial grid, date, version, and quality information that must be consulted before use.
Optical, Radar, and Thermal Data
Optical sensors use visible and infrared radiation and are well suited to vegetation, land cover, water colour, and burn scars, but clouds, smoke, shadows, and illumination limit their use. Radar systems transmit their own signal and can operate at night and through clouds, making them useful for floods, moisture, deformation, and surface structure. Thermal channels support surface-temperature analysis, heat anomalies, and active-fire detection. Spatial resolution describes pixel size but does not guarantee that every object of that size can be reliably identified. Data selection involves trade-offs: detailed imagery may cover less area or repeat less often, while a frequent global product may be coarser.
Comparing Observations Through Time
Change is detected by comparing imagery, classified maps, or indices across dates — supporting monitoring of deforestation, urban growth, crops, water extent, wildfire effects, floods, and infrastructure damage. Valid comparison requires compatible products, common projection, atmospheric correction, and consideration of season, vegetation cycles, and illumination. Otherwise, a natural seasonal difference may be misclassified as environmental damage.
Main Applications and Limitations
Satellites support mapping of land cover, forests, wetlands, drought, snow, marine conditions, atmospheric pollution, and climate variables. Emergency services create flood, wildfire, and damage maps for response and recovery. For businesses and authorities, these data can support risk assessment, territorial screening, inspection planning, and verification of reported change. They do not replace title documents, engineering inspections, or laboratory analysis. A pixel often contains a mixture of surface types. Clouds, smoke, snow, shadows, and atmospheric conditions create gaps. Some events occur between overpasses or are too small for available resolution. An algorithm may indicate change occurred without proving the cause — a vegetation-index decline could result from drought, harvest, fire, shadow, or a sensor difference.
Ground Validation and Reproducible Workflow
Field observations, sensors, site descriptions, laboratory samples, and very-high-resolution imagery are used to train, calibrate, and validate satellite products. A validation plan should cover different land-cover types, seasons, and levels of difficulty rather than only convenient sites. Results should include accuracy metrics, the date of ground data, and the limits of applicability. A reproducible workflow records the source dataset identifier, download date, algorithm version, processing parameters, cloud masks, study boundary, and manual edits — allowing the analysis to be repeated after data updates. Important decisions benefit from combining optical and radar imagery, ground measurements, documents, and expert assessment.
Satellite monitoring provides scale, repeat coverage, and long time series. Its value is greatest when the data type matches the question, change is analysed with attention to season and resolution, and results are validated with ground information.