Native Forest Mapping in Patagonian Andes based on Optical Satellite Imagery

LENCINAS JOSÉ DANIEL; FARIAS MARÍA

PHOTOGRAMMETRIE, FERNERKUNDUNG, GEOINFORMATION

E SCHWEIZERBARTSCHE VERLAGS

Año: 2005 p. 147 - 156

1432-8364

Data derived from optical satellite sensors have been increasingly utilized as a source of data to inventory native forests in Argentinean Patagonia, mostly due to the difficulty of accessing vast, rugged areas, where native forests are placed. This study focuses on the classification process applied to map native Nothofagus forest, as well as the identification of the critical factors that affected it. The study area corresponds to the watershed of La Plata and Fontana lakes, covering 1,256 km2. Both types of SPOT 4 sensor data, multispectral and panchromatic mode, were used. A combination of supervised classification and a binary hierarchical procedure was utilized. The classification revealed a total area of 44,484 ha of Lenga (Nothofagus pumilio) forests and 5,791 ha of ñire (Nothofagus antarctica). The classification accuracy for Lenga class was better than that for Ñire (85% and 80% of user’s accuracy respectively). Some justifications for these results and recommendations for future classifications are given.