研究目的
To evaluate the use of MODIS NDVI and EVI time series for detecting deforestation in the Cerrado biome.
研究成果
The study demonstrated the feasibility of using MODIS EVI and NDVI time series for detecting deforestation in the Cerrado biome, with NDVI showing better performance than EVI. However, a single deforestation threshold for each vegetation formation across the entire biome is not feasible due to regional differences.
研究不足
The methodology requires a reliable land cover map for operational implementation. The detection of deforestation in campestre formations was not effective due to low biomass and high soil interference.
1：Experimental Design and Method Selection:
The study analyzed time series (2000-2013) of MODIS EVI and MODIS NDVI to detect deforestation in the Cerrado. The time series were smoothed by the double logistic filter available in the TIMESAT program. Kruskal-Wallis statistics was used to determine the equality of temporal signatures.
2：Sample Selection and Data Sources:
The study areas corresponded to the municipalities of Jataí/GO, Luís Eduardo Magalh?es/BA, Mateiros/TO and S?o Miguel do Araguaia/GO.
3：List of Experimental Equipment and Materials:
MODIS EVI and NDVI data, TIMESAT program for smoothing time series, and Landsat images for validation.
4：Experimental Procedures and Operational Workflow:
Extraction of temporal signatures, smoothing of time series, statistical analysis to define deforestation thresholds, and validation with Landsat images.
5：Data Analysis Methods:
Kruskal-Wallis test for statistical equality of temporal signatures and visual analysis of deforestation events.
独家科研数据包，助您复现前沿成果，加速创新突破
获取完整内容
