Abstract
Collecting and analyzing data at the correct scale is paramount to the successful implementation of any remote sensing mapping and monitoring activities, including in the mangrove ecosystem. This study investigates the variation of mangrove structural features, e. g. canopy and patch dimensions able to be detected from different remote sensing image pixel sizes. Semi-variograms were used as a tool to identify the spatial scale domain and thresholds for detecting and measuring mangrove features based on original and resampled WorldView-2 images. Our results show a gradual loss of vegetation structure information, i.e. the ability to delineate crowns and gaps, with increasing pixel size beyond 2 m.