首页 | 本学科首页   官方微博 | 高级检索  
     检索      

香港近海海域叶绿素a定量反演及时空变化分析
引用本文:孙宏亮,何宏昌,付波霖,王诗文,马瑞雪,吴明富.香港近海海域叶绿素a定量反演及时空变化分析[J].中国环境科学,2020,40(5):2222-2229.
作者姓名:孙宏亮  何宏昌  付波霖  王诗文  马瑞雪  吴明富
作者单位:桂林理工大学, 测绘地理信息学院, 广西 桂林 541004
基金项目:国家自然科学基金项目(41801071);国家自然科学基金(21976043);广西自然科学基金项目(2018GXNSFBA281015);桂林理工大学科研启动基金资助项目(GUTQDJJ2017096);广西八桂学者团队项目联合资助
摘    要:以2009~2019年HJ-1A/B卫星多光谱数据和对应日期的实测数据为数据源,通过预处理提取出各波段组合反射率并与实测叶绿素a浓度数据进行统计相关性分析,选取相关性最高的波段组合作为特征变量与2/3的实测叶绿素a浓度数据进行建模,并用剩下的1/3实测叶绿素a浓度数据进行精度验证以确定最佳遥感反演模型,最后根据最佳反演模型对2009-2019年的香港近海海域叶绿素a浓度进行反演,明晰该海域近10年的叶绿素a浓度时空变化特征.结果表明:利用HJ-1A/B卫星多光谱数据反演香港近海海域叶绿素a浓度的最佳波段组合为第3波段和第2波段比值(B3/B2),相关系数(r)为0.893;最佳反演模型为利用B3/B2构建的e指数回归模型(Chl=0.004e6.693(B3/B2)),决定系数(R2)为0.934,均方根误差(RMSE)为0.255μg/L,平均相对误差(RPD)为25%;近10年香港近海海域的叶绿素a浓度时空变化特征:空间上整体呈现“东高西低,由东向西逐渐减小”的分布特征,西部海域比东部海域平均浓度低5μg/L左右;2017年内呈“春低秋高,夏升冬降”的随季节变化特点,其中秋季最高,夏春两季次之,冬季最低.

关 键 词:叶绿素a浓度  HJ-1A/B卫星  多光谱数据  定量反演  时空变化分析  
收稿时间:2019-10-23

Quantitative inversion and analysis of spatio-temporal changes of chlorophyll-a concentration in Hong Kong's coastal waters
SUN Hong-liang,HE Hong-chang,FU Bo-lin,WANG Shi-wen,MA Rui-xue,WU Ming-fu.Quantitative inversion and analysis of spatio-temporal changes of chlorophyll-a concentration in Hong Kong's coastal waters[J].China Environmental Science,2020,40(5):2222-2229.
Authors:SUN Hong-liang  HE Hong-chang  FU Bo-lin  WANG Shi-wen  MA Rui-xue  WU Ming-fu
Institution:College of Geomatics and Geoinformation, Guilin University of Technology, Guilin 541004, China
Abstract:This paper statistically analyzed the correlation between reflectance of different band combinations and chlorophyll-a concentration using HJ-1A/B satellite multispectral images and measured measurement from 2009 to 2019. The statistical regression analysis was performed between the band combination with the highest correlation selected as a characteristic variable and 2/3 of the total measured chlorophyll-a concentration data to build inversion models. Theses inversion models were evaluated using the rest of measured chlorophyll-a concentration data to determine the best remote sensing inversion model, which was utilized to calculate the chlorophyll-a concentration in coastal waters of Hong Kong from 2009 to 2019, and clarify the spatiotemporal characteristics of chlorophyll-a in the past 10 years. The results indicate that:The optimal band combination derived from HJ-1A/B multispectral data is the ratio of band 3and band 2 (B3/B2), which had the highest correlation with the chlorophyll-a concentration in Hong Kong's coastal waters. The correlation coefficient (r) is 0.893. The best inversion model is the e-exponential regression model (), the determination coefficient (R2) is 0.934, the root means square error (RMSE) is 0.255μg/L, and the relative average error (RPD) is 25%. The spatial distribution characteristics of chlorophyll-a concentration in Hong Kong's coastal waters is high in east and low in west, gradually decreasing from east to west in the past 10years. The average concentration of chlorophyll-a in the western waters are about 5μg/L lower than the in the eastern waters. The seasonal variation of chlorophyll-a concentration in 2017 was low in the spring, and high in the autumn, rising in the summer and dropping in the winter. The chlorophyll-a concentration is the highest in autumn, followed by summer and spring, lowest in winter.
Keywords:chlorophyll a concentration  HJ-1A/B satellite  multispectral image  quantitative inversion  temporal and spatial distribution characteristics  
本文献已被 CNKI 等数据库收录!
点击此处可从《中国环境科学》浏览原始摘要信息
点击此处可从《中国环境科学》下载免费的PDF全文
设为首页 | 免责声明 | 关于勤云 | 加入收藏

Copyright©北京勤云科技发展有限公司  京ICP备09084417号