Global tree-cover percentage is an important parameter used to understand the global environment. However, the available global percentage tree-cover products are few. Producing a new global-scale data set facilitates comparison analysis among maps. Our study was undertaken to map tree-cover
percentage on a global scale with better accuracy than previous studies. In this study, we estimated the tree-cover percentage on a global scale at a pixel size of 500 m using a modified supervised regression tree algorithm from the Moderate Resolution Imaging Spectroradiometer (MODIS)
data of 2008. Training data were derived from high-resolution images displayed in Google Earth and created by linear mixture simulation. The estimation model was modified to fit the reference data, which were randomly collected from the global area. The estimation result was validated at 1106
random sample points. The root mean square error (RMSE) between estimated and observed tree cover was 13.8%. The produced map was also compared with existing global data sets. The RMSE of our result was better than that of two existing global percentage tree-cover data sets (about 1.8% and
9.2% lower than Vegetation Continuous Fields MOD44B data set and Global Map – Percent Tree Cover data set, respectively). In our result and all other data sets, the accuracy was lower for forests than for other areas. When the produced map was compared with the Vegetation Continuous
Fields MOD44B data set, the low agreement between them was concentrated in closed forests of Colombia, India, and Indonesia, and closed to open forests of Central Africa.
No Reference information available - sign in for access.
No Citation information available - sign in for access.
No Supplementary Data.
No Article Media
Document Type: Research Article
Centre for Environmental Remote Sensing, Chiba University, Chiba, Japan
Department of Natural Resources and Environmental Management, University of Hawaii at Manoa, Honolulu, USA
Publication date: February 16, 2016
More about this publication?