ABSTRACT
Remotely sensed data, with its unique characteristics in repetitive data acquisition, synoptic view, and digital format suitable for computer processing, make it the primary data
CONTENTS
7.1 Introduction ........................................................................................................................ 127 7.2 Methods ............................................................................................................................... 128
7.2.1 Data Collection and Preprocessing ..................................................................... 129 7.2.2 Development of Fractional Images with Linear Spectral Mixture Analysis ................................................................................................... 130 7.2.3 Vegetation Gain/Loss Detection ......................................................................... 130 7.2.4 Land Use/Cover Classification and Evaluation ................................................ 132 7.2.5 Detection of Detailed “From-To” Vegetation Change Trajectories ................ 133
7.3 Resultant Analysis and Discussion ................................................................................. 133 7.3.1 Analysis of Vegetation Gain/Loss Results ......................................................... 133 7.3.2 Analysis of Detailed Vegetation Change Trajectories ...................................... 135
7.4 Conclusions ......................................................................................................................... 138 Acknowledgments ...................................................................................................................... 139 References ..................................................................................................................................... 139
source for examining land use/cover change at different scales (Lu et al. 2004). Since the first earth observation satellite was launched in the early 1970s, different kinds of sensor data have been readily available for land use/cover change detection (Althausen 2002; Lefsky and Cohen 2003). The optical sensors aboard the Landsat satellite platform series may be the most common data source for change detection because they offer a relatively long history of data collection (from the early 1970s to 2011), with suitable spectral and spatial resolutions. In particular, because of recent free public access to these images, the use of time-series Landsat images for detecting land use/cover change has recently attracted great interest (Cohen et al. 2010; Huang et al. 2010; Kennedy et al. 2010; Thomas et al. 2011; Hansen and Loveland 2012).
