热带地理 ›› 2019, Vol. 39 ›› Issue (4): 583-596.doi: 10.13284/j.cnki.rddl.003161

• 专刊:无人机在生态学和地理学中的应用 • 上一篇    下一篇

基于无人机影像的北部湾典型岛群红树林生态系统净初级生产力估算

田义超1ab,黄远林1a,陶 进1a,张 强1a,吴 彬1a,张亚丽1a,黄 鹄1b,梁铭忠1b,周国清2   

  1. (1.北部湾大学 a. 资源与环境学院;b. 海洋地理信息资源开发利用重点实验室,广西 钦州 535000; 2.桂林理工大学 广西空间信息与测绘重点实验室,广西 桂林 541004)
  • 出版日期:2019-07-10 发布日期:2019-07-10
  • 通讯作者: 黄远林(1971―),男,湖南龙山人,副教授,博士,主要从事资源环境遥感与GIS的相关研究,(E-mail)huangyuanlin@yeah.net。
  • 作者简介:田义超(1986―),男,陕西西安人,工程师,博士,主要从事资源环境遥感及海岸带生态环境监测的相关研究,(E-mail)tianyichao1314@yeah.net。
  • 基金资助:

    广西自然科学基金联合资助培育项目(2018JJA150135);广西创新驱动发展专项(AA18118038);广西教育厅基金资助项目(ZD2014138)

Estimating the Net Primary Productivity of Typical Mangrove and Archipelago Ecosystems in the Beibu Gulf Based on Unmanned Aerial Vehicle Imagery

Tian Yichao1ab, Huang Yuanlin1a, Tao Jin1a, Zhang Qiang1a, Wu Bin1a, Zhang Yali1a, Huang Hu1b, Liang Mingzhong1b and Zhou Guoqing2   

  1. (1. a. College of Resources and Environment,; b. Key Laboratory of Marine Geographic Information Resources Development and Utilization in the Beibu Gulf, Beibu Gulf University, Qinzhou 535000, China; 2. Guangxi Key Laboratory for Geospatial Informatics and Geomatics Engineering, Guilin University of Technology, Guilin 541004, China)
  • Online:2019-07-10 Published:2019-07-10

摘要:

提出以无人机季节航拍影像为数据源,采用ENVI5.3软件中的CART方法对广西北部湾的典型岛群红树林景观类型进行解译,借助于Python语言实现CASA模型并将其引入到对岛群红树林生态系统的研究中,得出了不同海岛和红树林生态系统净初级生产力(NPP)的空间分布特征。结果表明:1)可见光波段差异性植被指数(VDVI)可以很好地区分海岛及红树林植被等典型地物,可应用在岛群红树林生态系统NPP的估算上;2)研究区NPP的总量为127.09 t·C/a,NPP的年均碳密度值介于0~1 437.12 g·C/m2,年均碳密度为399.85 g·C/m2。红树林生态系统的NPP值较高,而海岛生态系统的NPP值较低;3)季节尺度上NPP碳密度的大小与年内太阳辐射有直接的关系,夏季的NPP碳密度最大,冬季最小;4)白骨壤的单位面积NPP最大,为1 123.09 g·C/m2,而桐花树最小,为649.65 g·C/m2。红树林生态系统NPP平均碳密度低于实测样地和深圳福田红树林计算结果,这与不同红树林群落的植被光谱特征有关。

关键词: 无人机遥感, 景观格局, CASA模型, 净初级生产力, 岛群红树林生态系统, 广西北部湾

Abstract:

The net primary productivity (NPP) of vegetation is an important component of the carbon cycle and carbon budget balance in global ecosystems. Its accurate assessment depends on high-resolution satellite data. Previous studies have mostly estimated NPP using remote sensing data from moderate resolution imaging spectroradiometers (250-8000 m) and thematic mappers (30 m). Due to the limitations and influences of spatial resolution and weather factors in the Beibu Gulf, it is difficult to accurately assess the NPP of mangrove and archipelago ecosystems. This study is the first to use aerial images from unmanned aerial vehicles (UAV) as a data source and to use the CART method in ENVI5.3 to interpret the landscapes of mangrove and archipelago ecosystems. Using python language, the CASA model was implemented and introduced in our study to obtain spatial NPP distribution characteristics in different islands and mangrove ecosystems. We found that: 1) The VDVI index ranged from -1 to 1 and its vegetation index form was similar to the normalized differential vegetation index (NDVI). The threshold value for vegetation and non-vegetation segmentation was 0, which should be selected as the input parameter for the vegetation index in the CASA model. 2) The NPP carbon density in the study area ranged from 0 to 1437.12 g·C/m2, with an average value of 399.85 g·C/m2. NPP spatial distribution was divided by the mangrove ecosystem in the north and Guixian island in the south. The NPP of the mangrove ecosystem in the north was higher than that of the south. 3) The seasonal order of NPP carbon density was as follows: summer (191.35 g·C/m2) > autumn (116.55 g·C/m2) > spring (62.98 g·C/m2) > winter (28.94 g·C/m2). The percentage of the average annual NPP for each season was 47.86%, 29.15%, 15.75%, and 7.24%, respectively. 4) The maximum carbon density per unit area of mangrove ecosystem in the study area was 545.56 g·C·m-2, followed by Leeward Pier Island (535.90 g·C/m2), whilst Guixian Island was the lowest with a value of 413.54 g·C·m-2. The average NPP carbon density in typical mangrove and archipelago ecosystems estimated by this method was similar to the national average, lower than that of coastal provinces on the same latitude, but higher than that of provinces such as Hunan and Yunnan. The results of this study provide a reference and basis for further small-scale studies of NPP and its rapid estimation and also provide a new international research perspective. In the future, the Li-6400XT portable photosynthesis measurement system could be used to determine the maximum light use efficiency of different vegetation types to examine the NPP of mangroves in island groups and accurately assess the model. At the same time, high resolution satellite data and UAV data could be used to combine spatial and temporal data, thus improving both the spatial resolution and range of the vegetation index.

Key words: Unmanned Aerial Vehicle remote sensing, landscape pattern, Carnegie Ames Stanford Approach model, Net Primary Productivity, mangrove and archipelago ecosystems, Beibu Gulf of Guangxi