Detecting areas disturbed by gold mining activities through JERS-1 SAR images, Roraima State, Brazilian Amazon

Three JERS-1 SAR images ( 1993, 1994, 1996) were evaluated for detection of areas disturbed by gold mining activities in the region of Serra Tepequem, Roraima State, Brazilian Amazon. Subtle tonal contrast between eroded areas and savannah grass fields in the Tepequem plateau, constitutes a serious constraint on the identification of the disturbed areas, using SAR original images. However, normalized difference index (NDI) images allowed identification of cover changes related to mining activities. NDI images indicated increase in areas disturbed from 1993 to 1994 and from 1993 to 1996. The 1994-1996 NDI image did not show significant changes, indicating that most areas were disturbed prior to 1994. Results showed that processed JERS-1 SAR images may be used to detect land cover changes even in non-forested terrain with low spectral contrast among scene components.