Seeing the planet from above
Earth observation means using instruments on satellites to measure the surface, oceans and atmosphere from a distance. The field is often called remote sensing. Instead of sending people to measure a forest or a coastline, a satellite records energy coming from it, and analysts turn that record into maps, images and numbers.
Satellites suit this job because they cover large areas consistently and return to the same places again and again. A ground survey can be very detailed but is limited to where people can go and how often they can go there.
What the sensors measure
Satellites do not only take photographs. Different instruments capture different parts of the electromagnetic spectrum, and each reveals something different.
- Optical sensors record visible light, much like a camera. They produce the familiar pictures of cities, fields and coastlines but need daylight and a clear view of the ground.
- Multispectral and hyperspectral sensors record several or many narrow bands of light, including some the eye cannot see. Healthy vegetation, bare soil, water and snow reflect these bands differently, so the combination can help distinguish them.
- Thermal infrared sensors measure heat radiated by the surface and clouds. They can help track surface temperatures, fires and cloud-top heights, and they work at night.
- Radar sensors send out microwave pulses and record the echo. Because microwaves pass through cloud and do not need sunlight, radar can see in darkness and through weather. Synthetic aperture radar can also detect small changes in ground height over time, which is useful for studying subsidence and landslides.
- Altimeters and other instruments measure quantities such as sea level, ice height and atmospheric gases.
Orbits and the trade-offs behind the pictures
Image quality depends on a few linked choices.
Resolution describes the smallest feature a sensor can distinguish, often given as the ground size of one pixel. Finer detail usually means a narrower strip of ground seen at once.
Swath is the width of that strip. A wide swath covers more area but with coarser detail.
Revisit time is how often a satellite sees the same place again. A single satellite in low orbit may need days to return to a given spot, so operators often use several satellites to shorten it.
Many imaging satellites use sun-synchronous polar orbits, passing over each region at roughly the same local solar time so that lighting is comparable between images. Weather satellites commonly use geostationary orbit instead, since a satellite parked about 35,786 kilometers above the equator can watch one large area continuously. The general idea of orbit choice is covered in the overview of satellite technology.
Weather and climate
Weather forecasting is among the oldest and most widely felt uses. Geostationary satellites provide frequent views of cloud movement across a hemisphere, which helps forecasters follow storms as they develop. Polar-orbiting satellites pass lower and see finer detail, and some carry instruments that estimate temperature and humidity at different heights in the atmosphere. Those estimates feed into computer forecast models, which combine many data sources.
Over decades, consistent satellite records of sea ice, sea level, sea-surface temperature and vegetation also contribute to the study of climate. Long, steady records matter more here than any single dramatic image.
Land, water and cities
On land, satellite data supports a wide range of tasks.
- Tracking crop growth and drought stress across large farming regions
- Mapping forest cover and detecting clearing or fire damage
- Monitoring urban growth and changes in land use
- Mapping floods, so responders can see which areas are inundated
- Watching glaciers, lakes and coastlines for change
During disasters such as floods, earthquakes and wildfires, satellite images taken before and after an event help responders judge damage across areas that are hard to reach. Radar is useful when cloud covers the scene.
Turning raw data into information
Raw satellite data is not immediately useful. It must be corrected for atmospheric effects, aligned to a map, and calibrated so that measurements from different dates can be compared. The volume is large, which is why analysts increasingly rely on machine learning to find patterns such as buildings, roads or crop types. Processing and storage commonly happen in data centers rather than on the satellite itself, though some satellites do limited processing on board.
Combining several sources often works best, for example pairing satellite measurements with weather stations and field checks. Interpretation involves judgment, and results can be uncertain, especially for small features or when conditions change quickly.
Limits and questions
Satellites cannot see everything. Cloud blocks optical sensors, very small objects fall below the resolution, and a satellite overhead only briefly may miss short events. Higher-resolution imagery also raises privacy and surveillance technology questions, and governments and companies have different rules about who can collect and share it.
Still, satellite observation has become a standard tool for forecasting, environmental monitoring and emergency response, because it provides a steady, wide view that no other method can match at the same scale.