Distribution_and_Drivers_of_Global_Mangrove_Forest_Change_1996-2010

This study undertook an analysis of 1168 1^ox1^o radar mosaic composite images, combining 1996 JERS-1 and 2007 and 2010 ALOS PALSAR mosaic imagery across the geographical range of mangrove forests. The composite imagery was visually interrogated to determine the location and distribution of changes in extent over the period 1996-2010. In addition, these composites were used to establish the perceived drivers of the change in Land Cover and Land Use Change (LCLUC). Validation of the observed changes and their drivers was undertaken through reference to the dense time-series of Landsat optical imagery and high-resolution imagery via Google Earth.The data for this work was made available for use to the Global Mangrove Watch (GMW), an international coalition conceived through JAXA’s Kyoto & Carbon (K&C) Initiative to demonstrate the potential of time-series Japanese L-band Synthetic Aperture Radar (SAR) for characterising, mapping and monitoring mangroves at regional to global scales.Time-series of JERS-1 imagery from 1996 and ALOS PALSAR imagery data from 2007 and 2010 were compiled, in KML format for Google Earth, to generate temporal colour composite (1996, 2007 and 2010 in Red, Green and Blue bands) tiles measuring 1^ox1^o in extent at 25 m pixel resolution. Only the HH polarisation channel was used. The total geographical extent of the colour composite dataset was defined using the mangrove map of Giri et al 2011. The tiles were imported into Google Earth for the visual analysis of change visible as colours in the RGB composite images. Each region was manually interpreted to identify evidence of changes in mangrove extent to a predefined set of categories. These categories were selected using reference to the literature on the causes of mangrove forest loss and expert knowledge. The categories of change identified were: 1) Intact, 2) Prior Disturbance, 3) Erosion, 4) Deposition/Regrowth, 5) Aquaculture/Agriculture, 6) Dieback 7) Logging.