1 研究区概况,,,,, [14],,,,,,,, 20 90, 7 m [15],, 2 数据获取与处理 2.1 光谱数据的获取与处理 : 00~14: : 00~14: 00, (Avantes), 180~1 170 nm, 2.4 nm

Size: px

Start display at page:

Download "1 研究区概况,,,,, [14],,,,,,,, 20 90, 7 m [15],, 2 数据获取与处理 2.1 光谱数据的获取与处理 : 00~14: : 00~14: 00, (Avantes), 180~1 170 nm, 2.4 nm"

童臧
5 years ago
Views:

1 青岛灵山湾潮间带沉积物粒度参数高光谱定量反演陈晓琳 1, 孙根云 1, 徐方建 1, 赵永芳 2 1, 王振杰 (1. ( ), ; 2., ) 摘要 : 潮间带表层沉积物粒度参数不仅是研究现代海岸动态和变迁的参照, 也是评价人类活动可持续发展的依据本文以灵山湾潮间带野外实地采集潮滩表层沉积物及其表面实时光谱信息为研究对象, 分析高光谱用于粒度参数识别的能力针对高光谱信息提取的困难, 对光谱数据进行数学变换, 利用多层感知器神经网络建立沉积物粒度参数与光谱反射率之间的关系模型, 探索了利用表层沉积物反射光谱特性进行沉积物特征定量反演的可行性结果表明 : 基于高光谱对沉积物粒度参数的定量预测是可行的, 建立的预测模型可以较准确地反映潮间带沉积物粒度特征, 其快速准确的优点为测定潮间带表层沉积物粒度参数提供了一种新方法该模型的建立为进一步提高潮间带变化监测和预测的研究提供了参考借鉴关键词 : 高光谱 ; 潮间带 ; 沉积物 ; 粒度 ; 定量反演中图分类号 : P76; P237 文献标识码 : A 文章编号 : (2015) doi: /hykx ,,,,,, [1-3], [4], [5-6],,, [7-9],,,, [10],, [10-12],,,,,,, [10-12],,,,, [13],,,,,,, : ; : : (No , , ); (No jch); (No. 14CX02038A, 14CX02039A, 12CX02003A) : (1991-),,,,, cxlin0929@163.com;,, genyunsun@163.com;,, xufangjiangg@163.com Marine Sciences / Vol. 39, No. 8 /

2 1 研究区概况,,,,, [14],,,,,,,, 20 90, 7 m [15],, 2 数据获取与处理 2.1 光谱数据的获取与处理 : 00~14: : 00~14: 00, (Avantes), 180~1 170 nm, 2.4 nm, 0.6 nm 2, 3~4, 40 cm 56cm ~ E, ~ N, 50 m, 5, Magellan explorist 500 GPS, <7 m 180~350 nm 1 000~1 170 nm,,, FWHM, 10 nm [16] 2.2 潮间带沉积物样品采集与分析, 0~2 mm,, 83 30% 1 mol/l, 3, ( ) BT-2002, 1~2 600 μm, <3% [17] 4 (,, ) 3 建模和验证方法 3.1 光谱数据的数学变换, (1)~ (3),, ;, [18-19] dr( ) R( m 1) R( ) (1) d m 1 2 d R( ) dr( m 1) dr( ) (2) 2 d m 1 1 lg lg R( ) (3) R( ), m, λ m λ m+1 m m+1 ; R(λ m ) R(λ m+1 ) m m 模型的建立与精度验证,,, [20] (multi-layer perceptron, MLP) MLP MLP,,, ( ),, ; 96 / 2015 / 39 / 8

3 , 4 ;, 2, 11, 72 (R 2 ), (root-mean-square error, RMSE) [21] 4 结果与讨论 4.1 沉积物粒度参数特征 1.10~2.51Φ 0.69~0.92 Shepard, ( - ), [22] 1,, 15%~45%,, 4.2 不同数学变换方法的光谱特征 5 表 1 粒度参数的特征统计值 Tab. 1 Statistical eigenvalues of grain size parameters (Φ) (%) , 1, 1a,, (VIS) (NIR),,, 1b, 450 nm, 550 nm, 470 nm,, 1c,, 470 nm 1d Fig.1 1 Reflectance curves of different mathematical transformed intertidal samples Marine Sciences / Vol. 39, No. 8 /

4 ,, 4.3 粒度参数与不同数学变换光谱相关分析,, 2, 4,,, 0.5 2, 0.9 Fig. 2 2 Correlation coefficients between granularity parameters and different mathematical transformed spectral,,, 2,, (770~3 800 nm) (440~580 nm), (620~770 nm) (390~440 nm) ;,, ;,, 4.4 粒度参数预测分析表 2 神经网络模型波段重要性 Tab.2 Important bands for ANN (Artificial Neural Network) model (nm) , 1 20, , 4 4, 98 / 2015 / 39 / 8

5 表 3 验证样本实测值与预测值的对比 Tab.3 Comparison between measured values and predicted values R 2 RMSE (Φ) (R ),, 4,, 3, RMSE ( 3),, 3 3 Pearson, 空间制图分析,,, 3 Fig.3 Comparison between measured and predicted values of validation samples a. (Φ ); b. ; c. ; d. a. average particle size(φ value) ;b. standard deviation; c. skewness; d. kurtosis Marine Sciences / Vol. 39, No. 8 /

6 ,,,, 4,, Fig.4 4 Comparison between measured and predicted values in spatial distribution 5 结论与存在的问题,, : 1), 4, 2),, (R ), Pearson 0.01,, 3),,,,, 100 / 2015 / 39 / 8

7 ,,,,, -,,, : [1] Liu W D, Baret F, Gu X F, et al. Relating soil surface moisture to reflectance[j]. Remote Sensing of Environment, 2002, 81(2-3): [2] Michael L W, Lin L, Susan L U. Predicting water content using Gaussian model on soil spectra[j]. Remote Sensing of Environment, 2004, 89(4): [3] Dehaan R L, Taylor G R. Field-derived spectra of salinized soils and vegetation as indicators of irrigation-induced soil salinization[j]. Remote Sensing of Environment, 2002, 80(3): [4],. [J]., 2004, 23(2): [5]. 40 [D]. :, [6]. [D]. :, [7],,,. [J]., 2009, 28(7): [8],,,. [J]., 2006, 25(1): [9],,,. [J]., 2011, 31(4): [10],,,. [J]., 2008, 30(3): [11],,. [J]., 2010, 28(3): [12],,,. [J]., 2009, 31(3): [13],,,. [J]., 2007, 27(8): [14],,,. [J]., 2014, 33(2): [15],,,. [J]., 2005, 23(3): [16] Kemper T, Sommer S. Estimate of heavy metal contamination in soils after a mining accident using reflectance spectroscopy[j].environment Science and Technology, 2002, 36(12): [17] McManus J. Grain size determination and interpretation [M]. London: Blackwell Scientific Publication, [18],,. [J]., 2011, 30(3): [19],,,. [J]., 2006, 31(11): 975. [20]. [M]. :, [21] Chang C W, Laird D A, Mausbach M, et al. Nearinfrared reflectance spectroscopy-principal components regression analyses of soil properties[j]. Soil Science Society of America Journal, 2001, 652: [22],,. [J]., 2002, 33(6): Marine Sciences / Vol. 39, No. 8 /

8 Quantitative inversion of sandy intertidal sediment grain size parameters content with hyperspectral in Lingshan Bay of Qingdao CHEN Xiao-lin 1, SUN Gen-yun 1, XU Fang-jian 1, ZHAO Yong-fang 2, WANG Zhen-jie 1 (1. School of Geosciences, China University of Petroleum, Qingdao , China; 2. Institute of Oceanology, Chinese Academy of Sciences, Qingdao , China) Received: Aug., 19, 2014 Key words: hyperspectral; intertidal zone; sediment; grain size; quantitative inversion Abstract: The sediment grain-size parameters of intertidal zone is not only a reference object for the research of modern coastal dynamics but also a basis for the evaluation of the sustainable development of human activities. This paper takes the surface sediment and the hyperspectral information of the intertidal zone in Lingshan Bay as study objects to analyze the ability in grain size parameters identifying. Mathematical transformations were taken aiming at solving the extraction difficulties in hyperspectral data. Multi-layer perceptron ANN (Artificial Neural Network) was established to build a relation model between the parameters and spectral reflectance to explore the feasibility of quantitative inversion of sediment characteristics. The final results show that using hyperspectral data to determine sediment grain-size parameters of intertidal zone is feasible. The advantages, i.e. fast and accurate, provide a new method for determining the intertidal zone sediment parameters. This would provide a reference for accurately monitoring and researching on the changes of intertidal zone. ( 本文编辑 : 李晓燕 ) 102 / 2015 / 39 / 8

[1] Nielsen [2]. Richardson [3] Baldock [4] 0.22 mm 0.32 mm Richardson Zaki. [5-6] mm [7] 1 mm. [8] [9] 5 mm 50 mm [10] [11] [12] -- 40% 50%

[1] Nielsen [2]. Richardson [3] Baldock [4] 0.22 mm 0.32 mm Richardson Zaki. [5-6] mm [7] 1 mm. [8] [9] 5 mm 50 mm [10] [11] [12] -- 40% 50% 38 2 2016 4 -- 1,2, 100190, 100083 065007 -- 0.25 mm 2.0 mm d 10 = 0.044 mm 640 3 300. Richardson--Zaki,,, O359 A doi 10.6052/1000-0879-15-230 EXPERIMENTAL STUDY OF FLUID-SOLID TWO-PHASE FLOW IN A VERTICAL