Satellite retrieval algorithm of high spatial resolution sea surface partial pressure of CO2: Application of machine learning in Xiangshan Bay in autumn

doi:10.3969-j.issn.1001-909X.2023.01.007

Abstract

Abstract:

Coastal bays are greatly affected by human activities and natural changes, and the influence mechanism of variation in seawater carbon source and sink patterns is extremely complex. Due to the small spatial scale of the bay, it is necessary to use wide-bands high-spatial resolution satellite remote sensing for monitoring the air-sea CO₂ flux. Compared with the traditional kilometer-level ocean color satellite data, the retrieval of the sea surface partial pressure of CO₂ (pCO₂), the key parameter to calculate air-sea CO₂ flux, is extremely challenging in small-scale bays. Taking Xiangshan Bay in Zhejiang Province in autumn as an example, a satellite retrieval algorithm for sea surface pCO₂ was proposed based on the in situ pCO₂ data and Sentinel-2 satellite images in the past five years, using the machine learning method of support vector machine (SVM). The algorithm validation results showed a good performance with R² of 0.92 and RMSE of 23.23 μatm, and the satellite-derived results were consistent with the in situ values. On this basis, the satellite products of pCO₂ in Xiangshan Bay in autumn from 2017 to 2021 (September to November) were produced. The results revealed that the pCO₂ of Xiangshan Bay showed a decreasing trend from the top of the bay to the mouth of the bay, with an average value of 514.56 μatm, of which the average pCO₂ in the inner bay was 551.94 μatm and the average pCO₂ in the outer bay was 477.19 μatm, which implied that Xiangshan Bay was a source of atmospheric CO₂ as a whole. There was no significant trend change of pCO₂ in autumn in the past five years. Combined with the analysis of in situ data of multiple parameters, it was found that the sea surface pCO₂ of autumn in Xiangshan Bay in 2021 was jointly regulated by physical mixing and biological activities. Sea surface temperature (SST) had a good positive correlation with pCO₂, which was mainly reflected by the thermodynamic equilibrium of carbonate system. In addition, the normalized pCO₂(NpCO₂) with average temperature had a good negative correlation with seawater salinity and dissolved oxygen saturation. The relationship between NpCO₂ and salinity resulted from the exchange of sea water inside the bay and offshore coastal water under tidal effect. Long-time series satellite data analysis also confirmed that sea surface pCO₂ had a relatively consistent trend with the average tide height inside and outside the bay, and this trend was stronger in the outside bay than that in the inner bay. In this study, a set of pCO₂ remote sensing retrieval methods in the small-scale bay was constructed, which laid a good foundation for the subsequent long-time series satellite monitoring of sea-air CO₂ fluxes.

Key words: coastal bay, sea surface pCO₂, support vector machines, high spatial resolution satellite remote sensing, Xiangshan Bay

CLC Number:

P734
TP79

LIU Tingyu, BAI Yan, ZHU Bozhong, LI Teng, GONG Fang. Satellite retrieval algorithm of high spatial resolution sea surface partial pressure of CO₂: Application of machine learning in Xiangshan Bay in autumn[J]. Journal of Marine Sciences, 2023, 41(1): 82-95.

Figures/Tables 9

Fig.1 The study area, cruise tracks and sampling stations (Red triangles represent the location of sampling stations; red starsrepresent the position of the tide gauging stations; black dots represent the location of the pCO2 extremums.)

Tab.1 Band informations of Sentinel-2 MSI and Landsat-8 OLI sensor

Landsat-8 OLI				Sentinel-2 MSI
波段序号	波段信息	波长/nm	空间分辨率/m	波段序号	波段信息	波长/nm	空间分辨率/m
Band1	气溶胶	430~450	30	Band1	气溶胶	433~453	60
Band2	蓝光	450~515	30	Band2	蓝光	458~523	10
Band3	绿光	525~600	30	Band3	绿光	543~578	10
Band4	红光	630~680	30	Band4	红光	650~680	10
				Band5	植被红边	698~713	20
				Band6	植被红边	733~748	20
				Band7	植被红边	773~793	20
Band5	近红外	845~885	30	Band8	近红外	785~900	10
				Band8a	植被红边	855~875	20
				Band9	水蒸气	935~955	60
Band6	短波红外	1 560~1 660	30	Band10	短波红外	1 360~1 390	20
Band7	短波红外	2 100~2 300	30	Band11	短波红外	1 565~1 655	20
Band8	全色波段	503~676	15	Band12	短波红外	2 100~2 280	20

Fig.2 Curve of in situ remote sensing reflectance of eleven stations in Xiangshan Bay (The pink rectangular areas represent the range of 8 bands of Sentinel-2 MSI, with band numbers at the top of the figure, and the solid dots represent the center wavelengths of each band.)

Fig.3 Correlation matrix of pCO2 and remote sensing reflectance at major bands and band ratio

Fig.4 Algorithm validation result (Fig.4a and fig.4b are modeling and validation results of the SVM training and validation dataset. The black straight line is the 1∶1 line; The orange line is the overall fitted curve. Fig.4c and fig.4d are pCO2 results of in situ measurement (Nov. 10, 2021) and satellite retrieval (Nov. 11, 2021), respectively. Fig.4e is the comparison between in situ pCO2 and satellite-derived pCO2 at corresponding longitude.)

Fig.5 Spatial distribution of Sentinel-2 MSI satellite-derived pCO2 (a, c, e, g, i, k) and Landsat-8 OLI SST (b, d, f, h, j, l) of adjacent time in Xiangshan Bay in autumn 2017-2021 (Red lightning represents the location of Datang power Plant.)

Fig.6 Spatial distribution of satellite-derived pCO2 in Xiangshan Bay in autumn 2017-2021 (Red lightning represents the location of Datang power Plant.)

Fig.7 Spatial distribution of underway measured parameters and their correlations with pCO2 in Xiangshan Bay in Nov. 10, 2021 (Fig.7a, fig.7c and fig.7e are the spatial distribution of SST, salinity and dissolved oxygen saturation, respectively. Fig.7b is the correlation between pCO2 and SST and the red and blue dots represent the data closest to the mouth and top of the bay. Fig.7d and fig.7f are the correlation between NpCO2 and salinity and dissolved oxygen saturation. The black straight lines are the fitting curves of total data, and the red dashed line is the fitting curve with the data in circles removed.)

Fig.8 Analysis of mechanism affecting satellite-derived pCO2in Xiangshan Bay in autumn

References 31

[1]	FRIEDLINGSTEIN P, O’SULLIVAN M, JONES M W, et al. Global carbon budget 2022[J]. Earth System Science Data, 2022, 14: 4811-4900. DOI URL
[2]	GRUBER N. Carbon at the coastal interface[J]. Nature, 2015, 517: 148-149. DOI
[3]	LOHRENZ S E, CAI W J, CHAKRABORTY S, et al. Satellite estimation of coastal pCO₂ and air-sea flux of carbon dioxide in the northern Gulf of Mexico[J]. Remote Sensing of Environment, 2018, 207: 71-83. DOI URL
[4]	ETCHETO J, DANDONNEAU Y, BOUTIN J. Estimating the interannual variability of the air-sea CO₂ flux in the east equatorial Pacific[J]. Physics and Chemistry of the Earth, Part B: Hydrology, Oceans and Atmosphere, 1999, 24(5): 405-410. DOI URL
[5]	ONO T, SAINO T, KURITA N, et al. Basin-scale extrapo-lation of shipboard pCO₂ data by using satellite SST and Chl a[J]. International Journal of Remote Sensing, 2004, 25(19): 3803-3815. DOI URL
[6]	LOHRENZ S E, CAI W J. Satellite ocean color assessment of air-sea fluxes of CO₂ in a river-dominated coastal margin[J]. Geophysical Research Letters, 2006, 33(1): L01601.
[7]	JOSHI I D, WARD N D, D’SA E J, et al. Seasonal trends in surface pCO₂ and air-sea CO₂ fluxes in Apalachicola Bay, Florida, from VIIRS ocean color[J]. Journal of Geophysical Research: Biogeosciences, 2018, 123(8): 2466-2484. DOI URL
[8]	TELSZEWSKI M, CHAZOTTES A, SCHUSTER U, et al. Estimating the monthly pCO₂ distribution in the North Atlantic using a self-organizing neural network[J]. Biogeosciences, 2009, 6(8): 1405-1421. DOI URL
[9]	BAI Y, CAI W J, HE X Q, et al. A mechanistic semi-analytical method for remotely sensing sea surface pCO₂ river-dominated coastal oceans: A case study from the East China Sea[J]. Journal of Geophysical Research: Oceans, 2015, 120(3): 2331-2349. DOI URL
[10]	吕航宇, 白雁, 李骞, 等. 夏季珠江口海域海水CO₂分压的卫星遥感反演[J]. 海洋学研究, 2018, 36(2):1-11. DOI
	LÜ H Y, BAI Y, LI Q, et al. Satellite remote sensing retrieval of aquatic pCO₂ in summer in the Pearl River Estuary[J]. Journal of Marine Sciences, 2018, 36(2): 1-11.
[11]	SONG X L, BAI Y, CAI W J, et al. Remote sensing of sea surface pCO₂ in the Bering Sea in summer based on a mechanistic semi-analytical algorithm (MeSAA)[J]. Remote Sensing, 2016, 8(7): 558. DOI URL
[12]	LE C F, GAO Y Y, CAI W J, et al. Estimating summer sea surface pCO₂ on a river-dominated continental shelf using a satellite-based semi-mechanistic model[J]. Remote Sensing of Environment, 2019, 225: 115-126. DOI URL
[13]	XU P, MAO X Y, JIANG W S. Mapping tidal residual circulations in the outer Xiangshan Bay using a numerical model[J]. Journal of Marine Systems, 2016, 154: 181-191. DOI URL
[14]	吴凡杰, 吴钢锋, 董平, 等. 基于三维水动力水质数值模型的象山港排污策略研究[J]. 海洋环境科学, 2021, 40(1):24-33.
	WU F J, WU G F, DONG P, et al. An investigation of the pollutants discharge control strategies in Xiangshan Bay based on 3D hydrodynamic and water quality numerical modelling[J]. Marine Environmental Science, 2021, 40(1): 24-33.
[15]	张健, 李佳芮, 刘书明, 等. 象山港生态系统脆弱性及其评价体系构建[J]. 海洋学研究, 2017, 35(2):74-81. DOI
	ZHANG J, LI J R, LIU S M, et al. The vulnerability of ecosystem and evaluation system construction of Xiangshangang Bay[J]. Journal of Marine Sciences, 2017, 35(2): 74-81. DOI
[16]	陈华伟, 吴卫飞. 象山港内新增网箱养殖污染物对海水水质的影响预测[J]. 渔业研究, 2021, 43(2):183-192. DOI
	CHEN H W, WU W F. Prediction of the influence of new cage culture pollutants on the seawater quality in Xiangshan Harbor[J]. Journal of Fisheries Research, 2021, 43(2): 183-192.
[17]	钱茹茹. 2001—2020年宁波近岸海域赤潮灾害特征分析[J]. 江苏海洋大学学报:自然科学版, 2022, 31(3):11-17.
	QIAN R R. Analysis on the characteristics of red tide disaster in Ningbo coastal waters from 2001 to 2020[J]. Journal of Jiangsu Ocean University: Natural Science Edition, 2022, 31(3): 11-17.
[18]	BAI Y, PAN D L, CAI W J, et al. Remote sensing of salinity from satellite-derived CDOM in the Changjiang River dominated East China Sea[J]. Journal of Geophysical Research: Oceans, 2013, 118(1): 227-243. DOI URL
[19]	WEISS R F. Carbon dioxide in water and seawater: The solubility of a non-ideal gas[J]. Marine Chemistry, 1974, 2(3): 203-215. DOI URL
[20]	TAKAHASHI T, OLAFSSON J, GODDARD J G, et al. Seasonal variation of CO₂ and nutrients in the high-latitude surface oceans: A comparative study[J]. Global Biogeo-chemical Cycles, 1993, 7(4): 843-878.
[21]	国家海洋局. 海洋调查规范第4部分:海水化学要素调查: GB/T 12763.4—2007[S]. 2007.
	State Oceanic Administration. Specifications for oceanographic survey—Part 4: Survey of chemical parameters in sea water: GB/T 12763.4—2007[S]. 2007.
[22]	陈楚群, 潘志林, 施平. 海水光谱模拟及其在黄色物质遥感反演中的应用[J]. 热带海洋学报, 2003, 22(5):33-39.
	CHEN C Q, PAN Z L, SHI P. Simulation of sea water reflectance and its application in retrieval of yellow substance by remote sensing data[J]. Journal of Tropical Oceanography, 2003, 22(5): 33-39.
[23]	LI T, ZHU B Z, CAO F, et al. Monitoring changes in the transparency of the largest reservoir in Eastern China in the past decade, 2013-2020[J]. Remote Sensing, 2021, 13(13): 2570. DOI URL
[24]	WANG X Y, ZHONG Y X. Statistical learning theory and state of the art in SVM[C]// Proceedings of the second IEEE International Conference on Cognitive Informatics, 2003: 55-59.
[25]	YANG K, YU Z Y, LUO Y, et al. Spatial and temporal variations in the relationship between lake water surface temperatures and water quality-A case study of Dianchi Lake[J]. Science of the Total Environment, 2018, 624: 859-871. DOI URL
[26]	KIM Y H, IM J, HA H K, et al. Machine learning approaches to coastal water quality monitoring using GOCI satellite data[J]. GIScience & Remote Sensing, 2014, 51(2): 158-174.
[27]	FRIEDRICH T, OSCHLIES A. Neural network-based estimates of North Atlantic surface pCO₂ from satellite data: A methodological study[J]. Journal of Geophysical Research, 2009, 114(C3): C03020.
[28]	黄菊, 高郭平, 程天宇, 等. 2015年冬季南黄海水文特征及海-气CO₂通量分析[J]. 上海海洋大学学报, 2017, 26(5):757-765.
	HUANG J, GAO G P, CHENG T Y, et al. Hydrological features and air-sea CO₂ fluxes of the Southern Yellow Sea in the winter of 2015[J]. Journal of Shanghai Ocean University, 2017, 26(5): 757-765.
[29]	SHIM J, KIM D, KANG Y C, et al. Seasonal variations in pCO₂ and its controlling factors in surface seawater of the northern East China Sea[J]. Continental Shelf Research, 2007, 27(20): 2623-2636. DOI URL
[30]	张龙军. 东海海-气界面CO₂通量研究[D]. 青岛: 中国海洋大学, 2003.
	ZHANG L J. The study on the CO₂ flux at the sea-air interface over the East China Sea[D]. Qingdao: Ocean University of China, 2003.
[31]	石晓勇, 陆茸, 张传松, 等. 长江口邻近海域溶解氧分布特征及主要影响因素[J]. 中国海洋大学学报:自然科学版, 2006, 36(2):287-290,294.
	SHI X Y, LU R, ZHANG C S, et al. Distribution and main influence factors process of dissolved oxygen in the adjacent area of the Changjiang Estuary in autumn[J]. Periodical of Ocean University of China, 2006, 36(2): 287-290, 294.

Satellite retrieval algorithm of high spatial resolution sea surface partial pressure of CO₂: Application of machine learning in Xiangshan Bay in autumn

RichHTML

PDF

Knowledge

Abstract

Cite this article

share this article

Figures/Tables 9

References 31

Related Articles 1

Recommended Articles

Metrics

Comments