Ground Location Accuracy

This figure, which shows a site in North Korea, is an example of an image that can be used as a map (e.g., locations of roads, railroads, bridges, buildings) if the spatial accuracy is high.

There are two types of spatial accuracy: absolute accuracy and relative accuracy.

• Absolute accuracy reflects how well the locations of features in an image match their true geographic locations on the surface of the earth.
• Relative accuracy relates to how closely the distance between objects within an image match their true distances.

For images taken at nadir, relative accuracies are normally quite good for both the old and new systems.

In general, the old systems lacked good, absolute accuracies producing significant practical considerations. In order to correct these inaccuracies, it is necessary to relate features in the image to the corresponding objects on the ground where the location is known to have a high degree of accuracy. These locations are called ground control points (GCP). The entire image is then adjusted around these known points so that features fit more accurately to their correct locations.

This process of identifying of GCPs and registering an image can be expensive, time consuming, and introduce error. In the worst case, it may not be possible to find enough GCPs to correctly register an image to the earth's surface.

The new satellites are agile and stable enough to permit relatively high absolute spatial accuracy without resorting to GCPs. For instance, the Space Imaging System quotes a map accuracy comparable to 1:24,000 scale U.S. maps (12.2 meters; 90% CE) without using any GCPs for images taken at nadir. If greater accuracy is required, the use of GCPs with Space Imaging System images will result in image maps with absolute spatial accuracies that meet U.S. National Map accuracy standards for 1:2,400 (2.0 meters; 90% CE) scale maps.