Mangrove forests, occupying tropical and subtropical coastal areas, serve as coast protector, water filter, spot of attraction and an exceptional carbon reservoir. Despite these extraordinary ecological and economic functions, our understanding of the mangroves is limited because field survey is hard to conduct in the intertidal mangrove habitats. While satellite remote sensing provides good spatial and temporal coverages globally, data availability at the tropical latitudes is limited due to frequent cloud contamination. The quickly emerging unmanned aerial vehicle (UAV) technique enables data collection under almost all weather conditions, thus provides new opportunities. By mounting light detection and ranging (LiDAR) sensors on UAVs, the 3D structure of mangroves can be accurately characterized even at individual tree level. The basic tree parameters such as tree height and crown size can be easily extracted, which then enables the estimation of biomass, carbon stock and other important ecological indices. This study uses high density (>100 pt/m2) UAV-LiDAR data collected at four consecutive years to detect the growth rate and pattern of mangroves. The change detection is done at individual tree level. The dependence of mangrove growth pattern on tree species and clumpingness are analyzed. The output will allow scientists and environment managers to know mangroves better and to manage them effectively and efficiently.