Linear spectral mixture model as a tool for monitoring deforestation and timber exploitation in the Brazilian Amazon

2003 ◽  
Author(s):  
Joao R. dos Santos ◽  
Yosio E. Shimabukuro ◽  
Valdete Duarte ◽  
Paulo Mauricio Lima de Alencastro Graea ◽  
Patricia Guedes da Silva
Keyword(s):  
Mixture Model ◽  
Brazilian Amazon ◽  
Spectral Mixture

Spectral Mixture Model Inspired Network Architectures for Hyperspectral Unmixing

2020 ◽  
Vol 58 (10) ◽  
pp. 7418-7434
Author(s):  
Yuntao Qian ◽  
Fengchao Xiong ◽  
Qipeng Qian ◽  
Jun Zhou
Keyword(s):  
Mixture Model ◽  
Network Architectures ◽  
Hyperspectral Unmixing ◽  
Spectral Mixture

Ecosystem Structure throughout the Brazilian Amazon from Landsat Observations and Automated Spectral Unmixing

2005 ◽  
Vol 9 (7) ◽  
pp. 1-31 ◽  
Author(s):  
Gregory P. Asner ◽  
David E. Knapp ◽  
Amanda N. Cooper ◽  
Mercedes M. C. Bustamante ◽  
Lydia P. Olander
Keyword(s):  
Land Use ◽  
Land Cover ◽  
Land Cover Change ◽  
Brazilian Amazon ◽  
Canopy Gap ◽  
Amazon Forest ◽  
Ecosystem Structure ◽  
Gap Fraction ◽  
Cerrado Savanna ◽  
Spectral Mixture

Abstract The Brazilian Amazon forest and cerrado savanna encompasses a region of enormous ecological, climatic, and land-use variation. Satellite remote sensing is the only tractable means to measure the biophysical attributes of vegetation throughout this region, but coarse-resolution sensors cannot resolve the details of forest structure and land-cover change deemed critical to many land-use, ecological, and conservation-oriented studies. The Carnegie Landsat Analysis System (CLAS) was developed for studies of forest and savanna structural attributes using widely available Landsat Enhanced Thematic Mapper Plus (ETM+) satellite data and advanced methods in automated spectral mixture analysis. The methodology of the CLAS approach is presented along with a study of its sensitivity to atmospheric correction errors. CLAS is then applied to a mosaic of Landsat images spanning the years 1999–2001 as a proof of concept and capability for large-scale, very high resolution mapping of the Amazon and bordering cerrado savanna. A total of 197 images were analyzed for fractional photosynthetic vegetation (PV), nonphotosynthetic vegetation (NPV), and bare substrate covers using a probabilistic spectral mixture model. Results from areas without significant land use, clouds, cloud shadows, and water bodies were compiled by the Brazilian state and vegetation class to understand the baseline structural typology of forests and savannas using this new system. Conversion of the satellite-derived PV data to woody canopy gap fraction was made to highlight major differences by vegetation and ecosystem classes. The results indicate important differences in fractional photosynthetic cover and canopy gap fraction that can now be accounted for in future studies of land-cover change, ecological variability, and biogeochemical processes across the Amazon and bordering cerrado regions of Brazil.

Predicting water content using linear spectral mixture model on soil spectra

2013 ◽  
Vol 7 (1) ◽  
pp. 073539 ◽  
Author(s):  
Huan Li ◽  
Changkuan Zhang ◽  
Ying Zhang ◽  
Dong Zhang ◽  
Jay Gao ◽  
...  
Keyword(s):  
Water Content ◽  
Mixture Model ◽  
Spectral Mixture
Sign in / Sign up

Export Citation Format

Share Document