Why satellites matter for forests
Forests cover huge, often remote areas. Satellites orbit the planet and photograph the same places again and again, so they can observe regions no field crew could reach and can do it on a regular schedule. Many of the world’s maps of tree cover and many of the alerts about clearing rely on satellite data. But they are an indirect way of observing: they record light and radar energy bouncing off the surface, and software has to interpret what that means.
What a satellite sees
Optical satellites record sunlight reflected from the ground in several wavelength bands, including visible light and near-infrared. Healthy leaves reflect strongly in near-infrared and absorb red light, whereas bare soil and dead vegetation behave differently. Combining bands gives indices of greenness, which are widely used to map vegetation and watch seasonal cycles. This signal is also used to follow spring leaf-out, linking to the topic of spring phenology.
Radar satellites send their own microwave pulses and measure what comes back. Radar works at night and can pass through clouds. It responds to the structure and moisture of vegetation rather than color, and some types can help to estimate the amount of woody material.
Some sensors, including space-based lidar, sample canopy height along tracks, which supports work on forest lidar and biomass.
Resolution: the size of the smallest detail
Resolution describes how fine the detail is. Each image is a grid of pixels, and each pixel summarizes the ground it covers.
- Widely used public programs offer pixels in the range of 10 to 30 meters. A 30 meter pixel covers 900 square meters, or 0.09 hectare.
- Commercial satellites provide much finer pixels, down to well below a meter, but images usually cost money and cover smaller areas.
- Coarse sensors with pixels of hundreds of meters see the whole globe daily, though only at a rough level.
Resolution shapes what can be detected. A 30 meter pixel can show a large clearing or a new road cut, but a single tree, a narrow strip of forest or selective removal of a few trees from a stand may disappear within a pixel’s average. The longest archives of consistent imagery go back to the 1970s, which lets researchers look at decades of change, albeit with the coarser tools of the time.
What satellites do well
- Mapping tree cover. Estimating the extent of tree canopy over large areas with consistent methods.
- Detecting abrupt change. Clear-cuts, fire scars, flooding, storm damage and new roads can show up clearly as sudden changes between dates. The scale of forest fires and windthrow events can be mapped soon after they occur.
- Giving early alerts. Frequent images allow near real-time warnings of clearing, which can help respond to illegal logging.
- Following seasons. Greenness over time distinguishes evergreen from deciduous forest and shows drought effects.
- Covering inaccessible areas. Remote regions and areas of conflict can be observed without sending anyone.
What satellites cannot easily show
Species and age. From above, many forests look similar. Distinguishing species from satellite data is difficult, and the age of trees is difficult to read directly. An old natural stand and a dense plantation can have similar signals, which is a key issue when comparing old-growth forests with managed forests.
Understory and what lies beneath. Optical sensors see the top of the canopy. Small trees, shrubs and the ground below are mostly hidden. Degradation caused by selective cutting or hunting within an intact-looking canopy can go unseen.
Condition. Subtle decline may only appear late, and many stresses, from insects to drought, can look alike.
Land use. A satellite sees cover, not purpose. An oil palm plantation, regrowth and a natural stand can resemble each other, and a cleared area may be a temporary harvest or permanent conversion. This is part of why definitions of forest vary.
Carbon and wood volume. These cannot be measured directly. They are estimated using models calibrated with field plots.
Practical challenges
Clouds are a major problem for optical satellites in humid tropical regions, where weeks may pass without a clear view. Methods combine many images, use radar or accept gaps. Shadows from terrain, haze and smoke can also distort measurements. Sensors change over time, and differences between instruments can create apparent change that is not real. Processing large archives requires computing power, though cloud platforms and public tools have made this much more accessible.
Why ground checks still matter
Maps from satellites are produced by classification, meaning software assigns each pixel to a class. Errors always occur, and the standard way to measure them is by comparing a sample of pixels with reference data from field visits, aerial photos or high resolution images. A good map comes with an accuracy assessment, often reported separately for each class.
Reported changes in forest cover estimated this way can carry quite large uncertainty, particularly when the changes are small compared with the total area. A reader should treat a headline value with that in mind. The way gains and losses are counted is another big factor, explored in the article on tracking forest loss and gain over time.
Using satellites sensibly
Satellite data are best for answering “where and when did the canopy change?” They are less suited to “what exactly is there and how healthy is it?” Combining them with plot inventories, drone and airborne scans, and local knowledge produces a fuller picture than any single source. Anyone reading a map should look for how it defines forest, what resolution it uses, which years it covers and how its accuracy was checked.