Ecosystem Analysis

Deforestation Studies

Objective
Identify changes in forest tree cover due to human activities or pest infestation. Changes to the forest cover must first be identified and then these changes can be characterized at the subpixel level for an accurate assessment of deforestation.

Challenges
· On-the-ground survey methods are prohibitively expensive and time consuming
· Many large forest tracts are remote and inaccessible
· High-resolution aerial photography not viable due to the small scale and expense required
required to identify tree species
· Change detection alone does not characterize the type of change

Approach
DeltaCue software can be used to identify areas of change over broad forest areas. IMAGINE Subpixel Classifier can then be used to characterize that change. For example, bare soil detections indicate areas of recent clear cut activity.

Areas detected represent new clear-cut areas in South America.

Solution and Results
The combination of automated change detection analysis and subpixel classification make it possible to use inexpensive low resolution Landsat imagery to characterize deforestation activity over very broad areas. The resulting coverage analysis can be very accurate and can also be used to identify particular types of activity such as clear cutting or slash and burn.