What Are Forests?

Forests are land areas dominated by trees whose crowns cover a substantial fraction of the ground, forming a vegetation layer tall and dense enough to create its own interior microclimate. The operational definition used for global reporting comes from the Food and Agriculture Organization, whose Global Forest Resources Assessment counts land spanning more than 0.5 hectares with trees higher than 5 meters and canopy cover above 10 percent, excluding land under agricultural or urban use. That threshold matters because it determines what gets classified as deforestation, what counts as afforestation, and how national carbon inventories are compiled.

In engineering terms, forests are studied less as a botanical subject than as a three-dimensional scattering medium and a dynamic carbon reservoir. Geoscience and remote sensing treat the canopy as a structured target whose height, density, and vertical profile can be retrieved from optical reflectance, radar backscatter, and lidar waveforms. Forestry supplies the field measurement conventions, including diameter at breast height, basal area, and stem density, that calibrate those retrievals against ground truth.

Forest Structure and Composition

Structure describes how biomass is arranged in space: canopy height, the number and depth of vertical strata, gap fraction, leaf area index, and the spatial pattern of stems. Composition describes what species make up those strata and in what proportion. The two together control light penetration, water flux, habitat availability, and the radar and optical signatures the forest presents to a sensor. Boreal, temperate, and tropical forests differ sharply in all of these, and the differences constrain which sensing method works. Dense tropical canopies saturate optical vegetation indices and C-band radar backscatter at moderate biomass levels, which is why longer wavelengths and direct height measurement are needed there.

Remote Sensing of Canopy Structure

Spaceborne lidar and radar address the saturation problem by measuring geometry rather than brightness. NASA's Global Ecosystem Dynamics Investigation lidar, operated from the International Space Station since 2019, records full-waveform returns in footprints of roughly 25 meters, resolving canopy top height, vertical profile, and ground elevation beneath the crown. Synthetic aperture radar at L-band and P-band penetrates further into the canopy than shorter wavelengths, and polarimetric interferometry recovers height from the phase difference between scattering centers. Optical time series from Landsat and Sentinel-2 supply the wall-to-wall spatial coverage that sampling instruments lack, and combining Landsat imagery with GEDI samples has become a standard route to continuous maps of forest structure.

Carbon Stocks and Change Detection

Aboveground biomass density, usually reported in megagrams per hectare, converts structural measurements into a carbon quantity through allometric equations fitted to destructive field samples. Gridded products such as the GEDI Level 4B aboveground biomass density map publish these estimates at one kilometer resolution with explicit uncertainty layers, since allometric and sampling error dominate the total budget. Change detection runs on a separate track: dense optical time series flag clearing, fire scars, insect damage, and selective logging by looking for departures from a fitted seasonal trajectory. Attribution of a detected change to a driver remains one of the harder problems, because degradation from selective harvest produces a much weaker signal than clear-cutting.

Applications

Forest observation and modeling have applications in a range of fields, including:

• National greenhouse gas inventories and carbon accounting
• Wildfire fuel load mapping and burn severity assessment
• Timber inventory and sustainable harvest planning
• Biodiversity and habitat conservation planning
• Hydrological modeling of runoff and evapotranspiration
• Supply chain verification for deforestation-free commodity regulations