Analysis of the regional spectral properties in northwestern South China Sea based on an empirical orthogonal function
-
Abstract: This study presents an analysis of the spectral characteristics of remote sensing reflectance (Rrs) in northwestern South China Sea based on the in situ optical and water quality data for August 2018. Rrs was initially divided into four classes, classes A to D, using the max-classification algorithm, and the spectral properties of whole Rrs were characterized using the empirical orthogonal function (EOF) analysis. Subsequently, the dominant factors in each EOF mode were determined.The results indicated that more than 95% of the variances of Rrs are partly driven by the back-scattering characteristics of the suspended matter. The initial two EOF modes were well correlated with the total suspended matter and back–scattering coefficient. Furthermore, the first EOF modes of the four classes of Rrs (A–D Rrs–EOF1) significantly contributed to the total variances of each Rrs class. In addition, the correlation coefficients between the amplitude factors of class A–D Rrs–EOF1 and the variances of the relevant water quality and optical parameters were better than those of the unclassified ones. The spectral shape of class A Rrs–EOF1 was governed by the absorption characteristic of chlorophyll a and colored dissolved organic matter (CDOM). The spectral shape of class B Rrs–EOF1 was governed by the absorption characteristic of CDOM since it exhibited a high correlation with the absorption coefficient of CDOM (ag (λ)), whereas the spectral shape of class C Rrs–EOF1 was governed by the back-scattering characteristics but not affected by the suspended matter. The spectral shape of class D Rrs–EOF1 exhibited a relatively good correlation with all the water quality parameters, which played a significant role in deciding its spectral shape.
-
Figure 1. The study area and the location distribution of different sampling stations in the northwestern South China Sea, including Beibu Gulf, Qiongzhou Strait and the nearby sea of Leizhou Peninsula.
Figure 2. Flow chart of the maximum classification algorithm. The maximum value of Rrs can be used to confirm different classes. The maximum value of classes A, B, and D is at 400, 490, and 565 nm, respectively. However, class C Rrs had a wide and flat peak ranging from 480 to 560 nm.
Figure 4. The level of turbidity and water types in the study area, the color changing from blue to orange shows that the water is increasingly turbid. The value less than 1.5 represents case-I waters; otherwise, it represents case-II waters.
Figure 5. Spatial distribution of Chl-a (μg/L) and TSM (mg/L) in the Beibu Gulf, Qiongzhou Strait and the nearby sea of Leizhou Peninsula in summer. The maximum values of Chl-a concentration (46 μg/L) is located at outside the Zhanjiang Bay and the TSM (12 mg/L) is located at the central of the Qiongzhou Strait.
Figure 6. The whole remote sensing reflectance measured in the northwestern South China Sea in summer (N=45).
Figure 8. The first three EOF modes for the whole set of Rrs. The first EOF mode (blue curve) accounts for 91.57% of the total variances of Rrs, whereas the second and three EOF modes (red and yellow curve) only account for 4.71% and 2.99%, respectively.
Figure 9. Correlation coefficient calculated between EOFi and the discrete in situ water quality and optical parameters. a. EOF1; b. EOF2 and c. EOF3.
Figure 10. Spectral shape of the first EOF mode for each class of Rrs spectra. Class A EOF1 (blue curve) accounts for 96.82%, class B EOF1 (red curve) accounts for 97.53%, class C EOF1 (yellow curve) accounts for 82.75%, and class D EOF1 (purple curve) accounts for 92.30%.
Table 1. Notations used in this study
Parameter Value Lu(λ) upward radiance (W/m2·sr) Lsky downward radiance of sky (W/m2·sr) Ed(0+λ) above sea surface downward irradiance (W/m2·sr) ρ Fresnel reflectance of the air–sea interface λ wavelength (nm) Rrs remote sensing reflectance (sr–1) EOFi ith empirical orthogonal mode r Correlation coefficient Chl–a chlorophyll a concentration (µg/L) TSM total suspended matter (mg/L) CDOM colored dissolved organic matter ag (λ) absorption coefficient of CDOM bb(λ) back–scattering coefficient TUB turbidity (UTN) 
下载: 导出CSV
Table 2. Statistics of in-situ datasets for each class of Rrs
Class Statistics Chl-a TSM TUB ag(433) bb(530) Class A min 0.80 0.20 0 0.001 0.002 max 2.10 2.50 0.33 0.006 0.004 mean 1.34 1.03 0.11 0.004 0.003 standard deviation 0.47 0.90 0.35 0.001 0.000 coefficient variation/% 0.35 0.88 6.85 0.35 0.13 Class B min 0.60 0.20 0.07 0.002 0.002 max 2.50 1.40 0.15 0.006 0.004 mean 1.82 0.73 0.12 0.003 0.005 standard deviation 0.54 0.43 0.05 0.001 0.007 coefficient variation/% 0.30 0.59 0.45 0.40 1.40 Class C min 2.40 0.50 0.00 0.003 0.003 max 19.4 4.7 165.1 0.025 0.057 mean 6.78 1.98 39.1 0.006 0.019 standard deviation 5.75 1.16 68.7 0.006 0.018 coefficient variation/% 0.85 0.59 1.76 0.96 0.95 Class D min 3.00 0.80 0.09 0.002 0.012 max 46.0 12.2 169.5 0.012 0.523 mean 17.2 3.08 106.4 0.006 0.080 standard deviation 12.2 2.71 73.7 0.003 0.123 coefficient variation/% 0.71 0.88 0.69 0.50 15.4 Note: Min and Max represent the minimum and maximum values, respectively. The coefficient variation is the standard deviation divided by the mean.
下载: 导出CSV
Table 3. The correlation coefficient calculated between the classes A–D Rrs–EOF1 and the variance of optical and water quality parameters
Correlation coefficient Class A EOF1 Class B EOF1 Class C EOF1 Class D EOF1 TSM –0.94 0.47 –0.02 0.49 bb(530) –0.39 0.07 –0.53 0.44 bb(412) –0.45 0.07 –0.60 0.47 Chl-a 0.62 0.06 –0.12 –0.51 ag(250) –0.55 0.13 0.38 –0.30 ag(433) –0.49 –0.40 0.48 –0.44 TUB –0.18 0.29 0.33 –0.59
下载: 导出CSV
-
[1] Allan T D. 1985. Remote assessment of ocean color for interpretation of satellite visible imagery: In Gordon H R and Morel A Y, eds. Lecture Notes on Coastal and Estuarine Studies. Berlin: Springer-Verlag, 114. [2] Dickey T D. 2001. Inherent optical properties and irradiance. In: Steele J H, ed. Encyclopedia of Ocean Sciences. Oxford, UK: Academic Press, 244–253 [3] Fu Dongyang, Zhang Ying, Liu Dazhao, et al. 2015. Evaluation model of coastal water quality and application research based on principal component analysis—a case of Leizhou Peninsula waters. Journal of Marine Sciences (in Chinese), 33(1): 45–50 [4] Gordon H R, Brown O B, Jacobs M M. 1975. Computed relationships between the inherent and apparent optical properties of a flat homogeneous ocean. Applied Optics, 14(2): 417–427. doi: 10.1364/AO.14.000417 [5] Huang Yichen, Li Yan, Shao Hao, et al. 2008. Seasonal variations of sea surface temperature, chlorophyll-a and turbidity in Beibu gulf, MODIS imagery study. Journal of Xiamen University (Natural Science) (in Chinese), 47(6): 856–863 [6] IOCCG. 2000. Remote sensing of ocean colour in coastal, and other optically-complex, waters. Dartmouth, Canada: IOCCG [7] Kari E, Kratzer S, Beltrán-Abaunza J M, et al. 2017. Retrieval of suspended particulate matter from turbidity–model development, validation, and application to MERIS data over the Baltic Sea. International Journal of Remote Sensing, 38(7): 1983–2003. doi: 10.1080/01431161.2016.1230289 [8] Kim S W, Saitoh S I, Ishizaka J, et al. 2000. Temporal and spatial variability of phytoplankton pigment concentrations in the Japan Sea derived from CZCS images. Journal of Oceanography, 56(5): 527–538. doi: 10.1023/A:1011148910779 [9] Li Qinzheng, Chen Peng, Sun Langlang, et al. 2018. A global weighted mean temperature model based on empirical orthogonal function analysis. Advances in Space Research, 61(6): 1398–1411. doi: 10.1016/j.asr.2017.12.031 [10] Liu Zilin, Ning Xiuren, Cai Yuming. 1998. Distribution characteristics of size-fractionated chlorophyll a and productivity of phytoplankton in the Beibu Gulf. Haiyang Xuebao (in Chinese), 20(1): 50–57 [11] Lubac B, Loisel H. 2007. Variability and classification of remote sensing reflectance spectra in the eastern English Channel and southern North Sea. Remote Sensing of Environment, 110(1): 45–58. doi: 10.1016/j.rse.2007.02.012 [12] Ma Ronghua, Kong Weijuan, Duan Hongtao, et al. 2009. Quantitative estimation of phycocyanin concentration using MODIS imagery during the period of cyanobacterial blooming in Taihu Lake. China Environmental Science (in Chinese), 29(3): 254–260 [13] Mobley C D. 1999. Estimation of the remote-sensing reflectance from above-surface measurements. Applied Optics, 38(36): 7442–7455. doi: 10.1364/AO.38.007442 [14] North G R, Bell T L, Cahalan R F, et al. 1982. Sampling errors in the estimation of empirical orthogonal functions. Monthly Weather Review, 110(7): 699–706. doi: 10.1175/1520-0493(1982)110<0699:SEITEO>2.0.CO;2 [15] Ta Zhijie, Yu Ruide, Chen Xi, et al. 2018. Analysis of the spatio-temporal patterns of dry and wet conditions in central Asia. Atmosphere, 9(1): 7. doi: 10.3390/atmos9010007 [16] Tang Shilin, Larouche P, Niemi A, et al. 2013. Regional algorithms for remote-sensing estimates of total suspended matter in the Beaufort Sea. International Journal of Remote Sensing, 34(19): 6562–6576. doi: 10.1080/01431161.2013.804222 [17] Xia Huayong, Li Shuhua, Shi Maochong. 2001. Three-D numerical simulation of wind-driven current and density current in the Beibu Gulf. Acta Oceanologica Sinica, 20(4): 455–472 [18] Xie Fei, Guo Ziqi, Tian Ye, et al. 2014. The preliminary inquiry of Chlorophyll-a inversion algorihtms applicable to guanting reservoir. In: Proceedings of 2013 IEEE International Geoscience and Remote Sensing Symposium. Melbourne, VIC, Australia: IEEE, 3785–3788. doi: 10.1109/IGARSS.2013.6723655 [19] Ye Huping, Li Junsheng, Li Tongji, et al. 2016. Spectral classification of the Yellow Sea and implications for coastal ocean color remote sensing. Remote Sensing, 8(4): 321. doi: 10.3390/rs8040321 [20] Zhang Tinglu, Shi Yingni. 2005. A method to classify case I and case II waters. Periodical of Ocean University of China (in Chinese), 35(5): 849–853 -
点击查看大图
计量
- 文章访问数: 427
- HTML全文浏览量: 144
- PDF下载量: 5
- 被引次数: 0
