Satellite Imagery
Definition
Satellite imagery refers to pictures and measurements of the Earth captured by instruments on satellites. Sensors span optical, thermal, radar, and microwave, each sensitive to different physical properties. Images come as raw digital numbers, calibrated reflectance or brightness temperature, and derived products. Key attributes include spatial resolution, spectral bands, revisit frequency, swath width, and radiometric depth. Because satellites provide systematic coverage, imagery underpins mapping, monitoring, and forecasting across the globe.
Application
Public missions provide open archives used for science, education, and many operational tasks. Commercial providers offer very high resolution and rapid tasking that serve defense, media, and infrastructure monitoring. Applications include crop yield estimation, disaster mapping, shoreline change, urban growth, and climate indicators like snow and fires. Developers stream imagery into cloud platforms for scalable processing and for near real time dashboards.
FAQ
How do optical and radar imagery differ in strengths and limitations?
Optical relies on sunlight and struggles with clouds or darkness, yet it offers intuitive color and rich spectral detail. Radar supplies its own energy, penetrates clouds, and senses surface roughness and moisture, but requires specialized processing and interpretation. Many projects benefit from combining both.
What metadata are essential to keep with imagery for later analysis?
Acquisition time, sensor geometry, calibration coefficients, processing level, cloud and shadow masks, and geolocation accuracy are critical. Without these, change detection and quantitative retrievals become unreliable. Storing lineage links to upstream scenes ensures reproducibility.
How should one manage the storage burden of long time series?
Use cloud optimized formats, compression, and tiling. Keep derived products rather than every intermediate. Where possible, query public archives on demand and cache only regions of interest to avoid duplication.
What are common misconceptions about very high resolution imagery?
More detail does not guarantee better analysis. Noise, shadows, and perspective effects can reduce model performance. Method choice should match the question, and coarser sensors can be superior for regional indicators and stable time series.
SUPPORT
© 2025 GISCARTA