结合潮位信息的红树林遥感识别

Remote sensing identification of mangrove forest combined tidal level information

红树林的滨海湿地生境使得它同陆生植被、水体-陆生植被混合像元难以区分,且红树林在遥感图像上的空间分布还随着潮位的变化而变化,因此基于通常采用的单一潮位遥感图像无法精确提取红树林空间信息.基于高潮位和低潮位TM遥感图像,尝试利用红树林的潮位周期性变化和滨海湿地特征来精确提取红树林空间分布信息.研究结果表明:基于缨帽变换和潮差信息提取的WI L+WI H、GVI L和GVI L-GVI H(WI、GVI分别为Wetness Index、Greenness Vegetation Index,下标L和H分别表示低潮位和高潮位)等指数能使红树林与其他地物之间具有很好的可分性;进一步采用最大似然法对红树林进行分类识别,通过结合潮位信息能精确提取红树林,其中制图精度和用户精度分别为94.57%、98.8%.

It is difficult to distinguish mangrove from terrestrial vegetation or the mixed pixels of water and terrestri- al vegetation due to the coastal wetland habit of mangrove. Moreover,the tide change also causes the change in spa- tial distribution characteristics of mangrove in remotely sensed imagery. Therefore,it is very difficult to precisely ex- tract the spatial mangrove information by means of the remote sensing imagery of single tide,which was usually a- dopted yet. Nevertheless the mangrove resides in coastal wetland,where the tide level varies periodically. In order to solve the problem,it was attempted to make good use of the unique habit characteristics of mangrove based on Land- sat TM remote sensing images of both high tide level and low tide level. The analysis results show that the separabil- ity between mangrove and the other objects are very good through WI L ＋ WI H,GVI L and GVI L- GVI H,which are de- veloped by the tidal range information and tasseled cap transformation. Note that WI and GVI stand for wetness in- dex and greenness vegetation index,respectively; while the subscripted L and H stand for low tide and high tide,re- spectively. The maximum likelihood classifier,an unsupervised classification method,was used to identify mangrove. The classification features based on the tidal range information,greenness index and wetness index can accurately map mangrove forest,and the producer＇s accuracy and user＇s accuracy of mangrove are 94. 57% and 98. 8%,re- spectively.