Objectives Large-scaled remote sensing extraction of mangrove plants can be easily affected by the factors, such as ground objects, sea flooding and the degree of mangrove plants' coverage. To solve this problem, the most suitable method to extract the growth area of mangrove plants was explored. In April 2019, we conducted a field survey on the mangrove growth area located in Tongming Bay, Zhanjiang.
Methods Using remote sensing image of GF-1(Gaofen-1) satellite, the distribution range of mangrove plants was extracted. According to the spectral characteristics of ground objects, normalized difference water index was used to extract the land-and-water boundary for establishing a suitable growing area of the mangrove plants, while object-oriented method for gradually pinpointing the growth area.Besides, regional distribution of mangrove plants was secondary extracted through the methods of spectral normalized difference index and vegetation index. The union of the two extraction results was obtained to maximize the precise growth area of mangrove plants. Based on the empirical model of Chlorophyll a of mangrove plants in the water environment, the retrieval of Chlorophyll a was carried out.
Results Experimental results show that the concentration of Chlorophyll a is high in the growth area of mangrove plants, and in the nearshore water with mangrove plants, the concentration of Chlorophyll a is higher than that without mangrove plants. Furthermore, correlation analysis shows that there is a good correlation between Chlorophyll a and mangrove plants, and the concentration of Chlorophyll a decreases sharply from high concentration to low concentration, which proves that the concentration and distribution characteristics of Chlorophyll a in the water around mangrove plants indicate a significant law of change.
Conclusions Therefore, this method can be applied to accurately extract the growth range of all the mangrove plants.