珠江口河网水体非光学活性水质参数高光谱反演
蔡建楠(1982—), 男, 广东省潮州市人, 硕士, 高级工程师, 主要从事环境监测遥感应用研究。email: |
Copy editor: 姚衍桃
收稿日期: 2020-02-03
要求修回日期: 2020-04-18
网络出版日期: 2020-04-24
基金资助
中国科学院科研装备研制项目(YJKYYQ20170048)
国家自然科学基金项目(41676171)
广东省科技创新战略专项(PDJH2020b1091)
版权
Retrieval of non-optically active water quality parameters by hyperspectra for river network waters in the Pearl River estuary
Received date: 2020-02-03
Request revised date: 2020-04-18
Online published: 2020-04-24
Supported by
Instrument Developing Project of the Chinese Academy of Sciences(YJKYYQ20170048)
National Natural Science Foundation of China(41676171)
Science and Technology Innovation Strategy Project of Guangdong Province(PDJH2020b1091)
Copyright
以粤港澳大湾区中山市及其邻近水域河网水体为试验区, 同步采集现场光谱及水质数据, 研究受测水体的高光谱反射率特征, 并分析非光学活性参数中化学需氧量(CODCr)、总磷(TP)浓度与高光谱反射率的相关性。结果显示, 各河流水体光谱反射率主要受悬浮颗粒物和叶绿素a的影响; 在500~680nm波段范围内, 水体光谱反射率大小与CODCr、TP浓度呈负相关关系; 与单波段相比, 特定波段的反射率比值与CODCr、TP浓度值的相关性较高, 与CODCr、TP浓度值相关性最高的反射率比值波段组合分别为R675/R794、R690/R815。选择上述波段组合建立的水质反演模型具有良好的估算精度, 模型估算平均相对误差分别为27.2%、32.1%, 表明高光谱技术在珠江口河网水体非光学活性参数CODCr、TP浓度反演上具有较大的应用潜景。
蔡建楠 , 刘海龙 , 姜波 , 陈吟晖 , 李杰鸿 , 吴思晓 , 梁建霞 , 黄华 , 邢前国 . 珠江口河网水体非光学活性水质参数高光谱反演[J]. 热带海洋学报, 2021 , 40(1) : 58 -64 . DOI: 10.11978/2020013
Field spectra and water quality measurements were performed in Zhongshan City and its adjacent river network in the Guangdong-Hong Kong-Macao Greater Bay Area. The hyperspectral reflectance characteristic of the river network and its correlation with the concentration of two non-optical activity parameters Chemical Oxygen Demand (CODCr) and Total Phosphorus (TP) were analyzed. The results showed that the spectral reflectance of the river network was mainly affected by suspended particulate matter and chlorophyll-a. In the range of 500~680 nm, the spectral reflectance of the water had negative correlations with the concentrations of CODCr and TP. Compared with the single band, the reflectance ratio of the specific band had higher correlations with CODCr and TP concentrations. The reflectance ratio band combinations with the highest correlations with CODCr and TP concentrations are R675/R794 and R690/R815, respectively. The water quality retrieval models that depend on the band combination show good estimation accuracy, with 27.2% and 32.1% of the average relative errors, indicating that the hyperspectral technology has good potential application for non-optical activity parameters (CODCr and TP) retrieval in the river network of the Pearl River estuary.
表1 样品水质特征描述Tab. 1 Description of characteristics of water quality |
水质指标 | 最小浓度值/(mg·L-1) | 最大浓度值/(mg·L-1) | 平均浓度值/(mg·L-1) | 标准差/(mg·L-1) |
---|---|---|---|---|
CODCr | 5 | 72 | 15 | 9 |
TP | 0.04 | 0.95 | 0.18 | 0.12 |
表2 CODCr和TP浓度反演模型拟合结果比较Tab. 2 Fitting results of CODCr and TP concentration retrieval models |
水质指标 | 波段组合 | 拟合方式 | 回归模型方程 | R2 | RMSE |
---|---|---|---|---|---|
CODCr | R675/ R794 | 线性拟合 | y=-10.4x+33.85 | 0.332 | 8.641 |
指数拟合 | y=104.6e-1.239x | 0.563 | 6.991 | ||
幂函数拟合 | y=23.59x-1.135 | 0.661 | 6.159 | ||
TP | R690/ R815 | 线性拟合 | y=-0.143x+0.4331 | 0.301 | 0.109 |
指数拟合 | y=1.353e-1.227x | 0.467 | 0.096 | ||
幂函数拟合 | y=0.3629x-1.596 | 0.569 | 0.086 |
表3 CODCr和TP浓度估算结果评价Tab. 3 Evaluation of estimation results for CODCr and TP concentrations |
水质指标 | 绝对误差最大值/(mg·L-1) | 绝对误差最小值/(mg·L-1) | 平均相对误差/% |
---|---|---|---|
CODCr | 8.607 | 0.435 | 27.2 |
TP | 0.121 | 0.002 | 32.1 |
[1] |
陈楚群, 施平, 毛庆文, 2001. 南海海域叶绿素浓度分布特征的卫星遥感分析[J]. 热带海洋学报, 20(2):66-70.
|
[2] |
陈雯扬, 许惠平, 2010. 基于高光谱数据的东海近海悬浮泥沙浓度估测研究木[J]. 水运工程, (2):31-35.
|
[3] |
林剑远, 张长兴, 2019. 航空高光谱遥感反演城市河网水质参数[J]. 遥感信息, 34(2):23-29.
|
[4] |
刘志国, 周云轩, 沈芳, 2007. 河口水体泥沙浓度的水面光谱统计模式分析[J]. 水利学报, 38(7):799-805.
|
[5] |
彭建, 徐飞雄, 邓凯, 等, 2018. 琅琊山景区不同指标浓度下水质光谱差异分析[J]. 光谱学与光谱分析, 38(5):1499-1507.
|
[6] |
王桂芬, 曹文熙, 杨跃忠, 等, 2010. 珠江口藻华水体总吸收系数的变化特性及高光谱反演模式[J]. 热带海洋学报, 29(2):52-58.
|
[7] |
王晓, 逄勇, 沈春其, 等, 2017. 太湖流域跨界水环境动态监测点位优化布局[J]. 水资源保护, 33(5):130-137.
|
[8] |
邢前国, 2007. 珠江口水质高光谱反演[D]. 广州: 中国科学院研究生院(南海海洋研究所).
|
[9] |
杨锦坤, 陈楚群, 唐世林, 等, 2007. 珠江口水体叶绿素荧光特性研究[J]. 热带海洋学报, 26(4):15-20.
|
[10] |
张海威, 张飞, 李哲, 等, 2017. 艾比湖流域地表水水体悬浮物、总氮与总磷光谱诊断及空间分布特征[J]. 生态环境学报, 26(6):1042-1050.
|
[11] |
张雪, 赖积保, 李家国, 等, 2019. 基于高分一号影像的深圳市黑臭水体遥感识别[J]. 科学技术与工程, 19(4):268-274.
|
[12] |
|
[13] |
|
[14] |
|
[15] |
|
[16] |
|
[17] |
|
[18] |
|
/
〈 | 〉 |