| 柘林湾近岸水产养殖区水域叶绿素a浓度反演 |
| |
| 引用本文： | 潘翠红,夏丽华,吴志峰,王猛,解学通,王芳.柘林湾近岸水产养殖区水域叶绿素a浓度反演[J].热带海洋学报,2021,40(1):142-153. |
| |
| 作者姓名： | 潘翠红 夏丽华 吴志峰 王猛 解学通 王芳 |
| |
| 作者单位： | 1.广州大学地理科学学院, 广东 广州 5100062.生态环境部华南环境科学研究所, 广东 广州 5100063.广东省农村水环境面源污染综合治理工程技术研究中心, 广东 广州 5100064.南方海洋科学与工程广东省实验室(广州), 广东 广州 511485 |
| |
| 基金项目： | 南方海洋科学与工程广东省实验室(广州)人才团队引进重大专项(GML2019ZD0301);广东省科技计划项目(2015A020216021);国家自然科学基金项目(41876204) |
| |
| 摘 要： | 海水养殖已成为近海水体环境的重要污染源, 叶绿素a作为水体浮游植物生物量的一个重要参数, 是水质评价的重要指标。本文以广东省柘林湾为研究区域, 采用2018年9月4日的哨兵2号(Sentinel-2)影像与海水养殖区水体中实测的叶绿素a浓度数据构建了叶绿素a浓度的单波段模型、比值模型、三波段模型与归一化叶绿素a指数模型(Normalized Difference Chlorophyll Index, NDCI)等估算模型; 通过对比评价, 以反演精度高的模型估算了2018年多个月份的叶绿素a浓度, 并分析其分布特征。结果显示: 1) NDCI模型的反演精度明显高于其他模型, 其可决系数R2为0.8, 均方根误差(Root Mean Square Error, RMSE)为9.7, 平均绝对百分比误差(Mean Absolute Percentage Error, MAPE)为0.99; 利用实测数据对NDCI模型的时间适用性进行检验, 表明NDCI模型能有效地估算出叶绿素a浓度的空间分布特征。2) 叶绿素a浓度呈现出从近岸向湾外逐步降低的趋势, 养殖区中叶绿素a浓度的总体趋势为池塘养殖区>滩涂插养区>网箱养殖区>浮筏养殖区; 受到水体交换、降雨及养殖活动的影响, 池塘养殖区中的叶绿素a浓度在投放幼苗期的2月最低, 其变化趋势为2月<4月<6月<12月。本文的研究结果可为相关部门对柘林湾养殖水体的环境监测提供参考。
|
| 关 键 词： | 海水养殖 叶绿素a浓度 反演 Sentinel-2影像 NDCI模型 |
| 收稿时间： | 2019-11-06 |
| 修稿时间： | 2020-03-24 |
Remote sensing retrieval of chlorophyll-a concentration in coastal aquaculture area of Zhelin Bay |
| |
| PAN Cuihong,XIA Lihua,WU Zhifeng,WANG Meng,XIE Xuetong,WANG Fang.Remote sensing retrieval of chlorophyll-a concentration in coastal aquaculture area of Zhelin Bay[J].Journal of Tropical Oceanography,2021,40(1):142-153. |
| |
| Authors: | PAN Cuihong XIA Lihua WU Zhifeng WANG Meng XIE Xuetong WANG Fang |
| |
| Institution: | 1. School of Geography Science, Guangzhou University, Guangzhou 510006, China2. South China Institute of Environmental Science, MEE, Guangzhou 510006, China3. Guangdong Rural Water Environment Non-point Source Pollution Comprehensive Treatment Engineering Technology Research Center, Guangzhou 510006, China4. Southern Marine Science and Engineering Guangdong Laboratory (Guangzhou), Guangzhou 511485, China |
| |
| Abstract: | Mariculture has become a major source of pollution in offshore waters. Chlorophyll-a, as a parameter of primary productivity, is an important indicator of water quality evaluation. We took Zhelin Bay of Guangdong Province as our study area. Using Sentinel-2 spectral image on September 4, 2018 and in-situ measured chlorophyll-a concentration, we constructed an estimation model of chlorophyll-a concentration to obtain the spatial distribution of chlorophyll-a concentration. In the chlorophyll-a concentration inversion model, we selected a linear regression model, a three-band model and the Normalized Difference Chlorophyll Index (NDCI) for comparative analysis. Through comparison and evaluation, a model with high inversion accuracy was used to estimate the chlorophyll-a concentration in multiple months of 2018 and analyze its distribution characteristics. The results showed that the inversion accuracy of the NDCI model was significantly higher than that of the other models. The decision coefficient R2 of the NDCI model was 0.8, the root mean square error (RMSE) was 9.7, and the mean absolute percentage error (MAPE) was 0.99. The time applicability of the NDCI model was tested by the measured data, which showed that the NDCI model could more accurately and effectively estimated the spatial distribution characteristics of chlorophyll-a concentration. The chlorophyll-a concentration showed a trend of decreasing from nearshore to the outside of the bay. The overall trend of chlorophyll-a concentration in the aquaculture area was as follows: pond breeding area > tidal flat breeding area > cage culture area > floating raft breeding area. Under the influence of water exchange, rainfall and culture activities, the concentration of chlorophyll-a in the culture area of the fish pond was the lowest in February when the fish was in the seedling stage, and its change trend was February < April < June < December. This study provides a valuable reference for environmental monitoring of marine aquaculture waters in Zhelin Bay. |
| |
| Keywords: | marine aquaculture chlorophyll-a concentration back analysis Sentinel-2 NDCI model |
|
| 点击此处可从《热带海洋学报》浏览原始摘要信息 |
| 点击此处可从《热带海洋学报》下载免费的PDF全文 |
|
