S p a t i a l   R e s o l u t i o n

One of the key concepts in remote sensing is that of spatial resolution. The amount of detail we can see depends on the character of the data we use. The following images illustrate the differences in spatial resolution achieved by different sensors aboard different satellites. The images illustrate the increasing detail achieved when moving from continent-scale images recorded by weather satellites (with low spatial resolution) up through AVHRR images (medium spatial resolution) to Landsat images and aerial photographs (with much higher spatial resolution).

Image of Brazil

Image of a part of South America recorded by the GOES weather satellite from a geosynchronous orbit nearly 22,000 miles above the Earth's surface. Spatial resolution (pixel size) of ~10 km provides enough information for general weather forecasts. (Pixel=picture element)

Satellite image of Central Amazonia

Satellite image depicting part of Central Amazonia acquired by the Advanced Very High Resolution Radiometer (AVHRR) aboard a NOAA weather satellite. From a height of about 520 miles we can see the confluence of two large rivers (the Rio Negro and the Rio Solimoes) near the city of Manaus (magenta). The forest is depicted in green, clouds in cyan (bright bluish-green), and evidence of human activities in orange.

Close-up of Central Amazonia

A close-up view of the preceding AVHRR image. The city of Manaus can be seen at the right bottom of the picture (magenta). Light patches of green (upper right corner) show areas of regenerating forest.

Note the coarseness of the image features caused by the relatively low spatial resolution of the data (pixel size = ~1.1 km) which fails to detect fine details within the scene.

Landsat of the city of Manaus

The same area as above recorded by Landsat from a lower orbit of about 438 miles above the Earth's surface reveals many more details. This is due to Landsat's much finer (higher) spatial resolution (pixel size = ~30 m).

Close-up of Landsat image of Manaus

Zoom to the top part of the above image. Now we can clearly identify roads, rivers, fresh clearings (magenta), and patches of regenerating forest (bright green). The dark green is mature tropical forest.

Note the different texture of the regenerating forest (with a smooth and simple canopy architecture) compared to the rougher and more irregular texture of mature forest (with its complex multi-layer canopy architecture).

The various dark patterns within the mature forest are caused by a drainage system of streams and small rivers. The road going across marks one of INPA's reservations and study sites, and is quite close to several of the INPA/Smithsonian forest fragments and Camp 41.

Close-up view of a landsat of the same area.

A close-up Landsat view of the same area. Here INPA has built a 45-meter-high observation jungle tower in the middle of mature tropical forest. Note that spatial resolution (pixel size = ~30 m) is not quite enough to distinguish between individual tree canopies, nor can we see the jungle tower located near the center of the image.

Aerial view of jungle close to tower

An aerial view close to the jungle tower. This image was taken by an ordinary photographic camera from a height of about 100 m from a light aircraft. The photograph was scanned into the computer and is used in research into how best to describe tropical tree species' diversity automatically. (See the page which explains spectral information.)

Remote Sensing: Seeing the Forest and the Trees

  Spatial Resolution
  Spectral Information
  Morphometric Information
  Tree Phenology
  Jungle Tower