马云梅,吴培强,潘良浩,王锦锦,王建步,任广波,马毅
1.内蒙古师范大学地理科学学院, 内蒙古自治区 呼和浩特 010010;2.自然资源部第一海洋研究所, 山东 青岛 266061;3.广西科学院广西红树林研究中心, 广西 北海 536000;4.上海欧比特航天科技有限公司, 上海 200000

摘要:
为了掌握茅尾海红树林种类和盖度情况，基于GF-2 PMS1高分辨率卫星遥感影像数据，使用支持向量机（support vector machine，SVM）和回归模型方法开展了茅尾海红树林物种分类和植被盖度估算研究，并对其分布特征展开了详细分析。研究表明：（1）广西茅尾海红树林共有4个种类，分别是桐花树（Aegiceras corniculatum）、无瓣海桑（Sonneratia apetala）、秋茄（Kandelia candel）和茳芏（Cyperus malaccensis）。桐花树作为茅尾海红树林的优势种，面积为1228.612 ha，占总面积72.5%，散布在茅尾海潮间带。秋茄面积最少，仅有1.976 ha，零星分布于茅岭乡和尖山镇沿岸，多数与无瓣海桑和桐花树混生。（2）基于GF-2 PMS1卫星影像的NDVI植被指数建立的二次多项式盖度回归模型效果最好，决定系数最大为0.7644，均方根误差RMSE最小为0.068。（3）植被盖度分布状况与植被类型有密切关系，植被种类纯生区的盖度要明显大于植被混生区的盖度。本研究结果为茅尾海国家海洋公园和自治区级自然保护区管理部门的决策提供理论基础。
关键词:红树林茅尾海GF-2 PMS1影像支持向量机(support vector machine,SVM)植被盖度
DOI：10.11759/hykx20190212001
分类号:TP79
基金项目:国家重点研究开发项目（2017YFC1405100）

Classification and coverage estimation of mangrove species at the Maoweihai Sea in Guangxi based on GF-2 images
MA Yun-mei^1,2, WU Pei-qiang², PAN Liang-hao³, WANG Jin-jin^4,5, WANG jian-bu², REN Guang-bo², MA Yi^2,1
1.College of Geographical Science, Inner Mongolia Normal University, Hohhot 010010, China;2.First Institute of Oceanography, Ministry of Natural Resources, Qingdao 266061, China;3.Guangxi Mangrove Research Center, Guangxi Academy of Sciences, Beihai 536000, China;4.Shanghai Aubit Aerospace Tech & shy;5.nology Co., Ltd., Shanghai 200000, China
Abstract:
To understand the species and coverage of mangrove in the Maoweihai Sea, based on GF-2 PMS1 high-resolution satellite remote sensing image data, support vector machine (SVM) classification and regression model method were used to carry out research on species classification and vegetation coverage estimation of mangrove in the Maoweihai Sea, and the distribution characteristics of mangrove were analyzed in detail. The conclusions are as follows:(1) The Maoweihai Sea mangrove consists of four species:Sonneratia apetala, Aegiceras corniculatum, Kandelia candel, and Cyperus malaccensis. As the dominant species of the mangrove in the Maoweihai Sea, Aegiceras corniculatum covers an area of 1228.612 ha, accounting for 72.5% of the total area and is found throughout the intertidal zone of the Maoweihai Sea. Kandelia candel has the least area, only 1.976 ha, and is scattered in Maolin rural area and Jianshan town along the coast; moreover, it is mostly mixed with the S. apetala and A. corniculatum. (2) The quadratic polynomial coverage regression model established based on the normalized differential vegetation index (NDVI) of GF-2 PMS1 satellite images had the best effect, with the maximum determination coefficient of 0.7644 and the minimum root mean square error (RMSE) of 0.068. (3) The distribution of vegetation coverage is closely related to the vegetation type, and the coverage of pure vegetation is significantly greater than that of mixed vegetation. Based on the above research results, this paper provides a theoretical basis for the decision-making of the management departments of the Maoweihai Sea national Marine park and autonomous region-level nature reserves.
Key words:mangrovethe Maoweihai SeaGF-2 PMS1 satellite imagesupport vector machine (SVM)vegetation fraction coverage





PDF全文下载地址:

点我下载PDF