| 78 | 0 | 38 |
| 下载次数 | 被引频次 | 阅读次数 |
研究中国近海溶解无机氮(DIN)浓度的精确时空监测，为理解海洋生态系统生物地球化学过程及近海水质评价与治理提供科学依据。结合卫星遥感影像与机器学习算法，构建中国近海DIN浓度遥感反演模型，通过递归特征消除法与网格搜索交叉验证法优化模型参数，重点分析渤海、黄海、东海和南海沿岸50 n miles范围内的典型近岸海域，并采用Spearman秩相关系数分析DIN浓度与人类活动指标的相关性。研究结果显示，模型在测试集上的决定系数(R2)达0.836，平均绝对误差(MAE)为0.047 mg/L。2003—2021年中国近海DIN浓度空间异质性显著：近岸高于远岸、封闭性海湾高于开敞性海域，渤海近岸浓度最高，为(0.315±0.045)mg/L，高于东海(0.200±0.006)mg/L和黄海(0.193±0.007)mg/L，南海相对较低，为(0.118±0.017)mg/L。DIN浓度呈现明显的季节性变化，近岸50 n miles海域冬季浓度最高，为(0.221±0.007)mg/L，夏季最低，为(0.155±0.005)mg/L，全年呈“先降后升”的单峰趋势。不同地区人类活动对DIN的影响不同：华东地区DIN浓度与废水排放、GDP正相关，华南受农业扩张与城市污染共同影响，华北与GDP、海水养殖负相关，东北无显著相关性。
Abstract：Dissolved inorganic nitrogen(DIN) is a crucial nutrient source for marine primary production and plays a significant role in regulating biological geochemical processes in marine ecosystems. However, driven by climate change and anthropogenic activities such as rapid urbanization and industrialization,DIN concentrations in coastal waters of China have exhibited pronounced changes in spatiotemporal distribution. Therefore, systematically identifying the spatiotemporal variation and influencing factors of DIN is crucial for health assessment and scientific management of coastal ecosystems. In this study, a DIN concentration inversion model suitable for China's coastal waters was developed by integrating satellite remote sensing imagery with machine learning algorithms. The parameters of the random forest(RF)model were optimized using Recursive Feature Elimination(RFE) and Grid Search with Cross-Validation(GridSearchCV) to improve model performance. Based on the optimized model, we estimated DIN concentrations in typical coastal areas within 50 nautical miles of the shorelines of the Bohai Sea, Yellow Sea,East China Sea, and South China Sea using MODIS image data during the period 2003-2021. Additionally, the Spearman rank correlation coefficient was used to assess the correlation between DIN concentration and human activities. On the test set, the coefficient of determination(R2) of the model was as high as0.836, with a mean absolute error(MAE) of 0.047 mg/L. From 2003 to 2021, DIN in the coastal waters of China presented significant spatial heterogeneity, with higher concentrations observed nearshore and in semi-enclosed bays. During this period, the average annual DIN concentration in the coastal water of Bohai Sea was highest(0.315±0.045) mg/L, and with greater annual fluctuations than the East China Sea(0.200±0.006) mg/L and Yellow Sea(0.193±0.007) mg/L. The annual average DIN concentration in the South China Sea was the lowest and most stable(0.118±0.017) mg/L. DIN concentrations presented pronounced seasonal variations, with the highest levels observed in winter(0.221±0.007) mg/L and the lowest in summer(0.155±0.005) mg/L, and followed a unimodal annual pattern, initially decreasing and then increasing. The impacts of human activities on DIN concentration varied across regions: DIN concentrations were positively correlated with wastewater discharge and GDP in East China; In South China, DIN concentrations were influenced by agricultural expansion and urban pollution; In North China, DIN concentrations exhibited a negative correlation with GDP and marine aquaculture; while in Northeast China,no significant correlation was observed between DIN concentrations and human activities. This study provides an effective technical path for remotely sensing DIN concentration in China's coastal seas, and a scientific basis for evaluating and managing coastal water quality.
陈艳,刘绥华,王堃,等, 2020.基于Landsat卫星影像的草海水质遥感反演及营养状态评价[J].水生态学杂志, 41(3):24-31.CHEN Y, LIU S H, WANG K, et al, 2020. Remote sensing ofCaohai Lake water quality using landsat satellite images[J]. Journal of Hydroecology, 41(3):24-31.
贾琳, 2021.近岸海域溶解氧与溶解无机氮预测分析方法研究[D].济南:山东大学.
蒋昊,赵亮,张晶, 2020.黄海溶解无机氮及颗粒有机氮收支与转化模型[J].中国环境科学, 40(9):3981-3991.JIANG H, ZHAO L, ZHANG J, 2020. Budgets and transforma-tion of dissolved inorganic and particulate organic nitro-gen in the Yellow Sea:a model study[J]. China Environ-mental Science, 40(9):3981-3991.
郑辉, 2018.大型海藻对水产养殖废水中无机氮和活性磷酸盐吸收作用研究[J].科技通报, 34(4):258-261.
王颢,陈克亮,吴烨飞,等, 2023.闽江入海断面溶解无机氮长时间序列分析及入海通量估算[J].中国环境监测, 39(6):98-109.
BARNIEH B A, DAVIS C, 2025. Monitoring anthropogenicimpacts on water quality in Ghana with remote sensing[J].Nature Reviews Earth&Environment, 6:378.
BLATCHFORD M L, MANNAERTS C M, ZENG Y J, et al,2019. Status of accuracy in remotely sensed and in situ ag-ricultural water productivity estimates:a review[J]. Re-mote Sensing of Environment, 234:111413.
BREITBURG D, LEVIN L A, OSCHLIES A, et al, 2018. De-clining oxygen in the global ocean and coastal waters[J].Science, 359(6371):eaam7240.
DAI Y H, YANG S B, ZHAO D, et al, 2023. Coastal phyto-plankton blooms expand and intensify in the 21st century[J]. Nature, 615(7951):280-284.
DONG Y, YANG J L, ZHAO X R, et al, 2022. Nitrate leachingand N accumulation in a typical subtropical red soil withN fertilization[J]. Geoderma, 407:115559.
DUCE R A, LAROCHE J, ALTIERI K, et al, 2008. Impacts ofatmospheric anthropogenic nitrogen on the open ocean[J].Science, 320(5878):893-897.
FAN W J, XU Z H, DONG Q, et al, 2023. Remote sensing-based spatiotemporal variation and driving factor assess-ment of chlorophyll-a concentrations in China’s Pearl Riv-er Estuary[J]. Frontiers in Marine Science, 10:1226234.
FU Y Z, XU S G, LIU J W, 2016. Temporal-spatial variationsand developing trends of Chlorophyll-a in the Bohai Sea,China[J]. Estuarine, Coastal and Shelf Science, 173:49-56.
GUYON I, WESTON J, BARNHILL S, et al, 2002. Gene se-lection for cancer classification using support vector ma-chines[J]. Machine Learning, 46(1):389-422.
JIAO N Z, TANG K, CAI H Y, et al, 2011. Increasing the micro-bial carbon sink in the sea by reducing chemical fertiliza-tion on the land[J]. Nature Reviews Microbiology, 9:75.
HO M, KESSOURI F, FRIEDER C A, et al, 2023. Effect ofocean outfall discharge volume and dissolved inorganic ni-trogen load on urban eutrophication outcomes in the South-ern California Bight[J]. Scientific Reports, 13(1):22148.
HUANG Y H, PAN J Y, DEVLIN A T, 2023. Enhanced esti-mate of chromophoric dissolved organic matter using ma-chine learning algorithms from landsat-8 OLI data in thePearl River estuary[J]. Remote Sensing, 15(8):1963.
KASPAR H F, HALL G H, HOLLAND A J, 1988. Effects of seacage salmon farming on sediment nitrification and dissimi-latory nitrate reductions[J]. Aquaculture, 70(4):333-344.
KHANGAONKAR T, NUGRAHA A, XU W W, et al, 2019.Salish sea response to global climate change, sea levelrise, and future nutrient loads[J]. Journal of GeophysicalResearch:Oceans, 124(6):3876-3904.
LI G X, QIAO L L, DONG P, et al, 2016. Hydrodynamic condi-tion and suspended sediment diffusion in the Yellow Seaand East China Sea[J]. Journal of Geophysical Research:Oceans, 121(8):6204-6222.
LIN J Y, WANG P, WANG J Z, et al, 2024. An extensive spatio-temporal water quality dataset covering four decades(1980–2022) in China[J]. Earth System Science Data, 16(2):1137-1149.
LUBCZYNSKI M W, LEBLANC M, BATELAAN O, 2024.Remote sensing and hydrogeophysics give a new impetusto integrated hydrological models:a review[J]. Journal ofHydrology, 633:130901.
MAO Y P, ZHANG H, CHENG Y H, et al, 2023. The charac-teristics of nitrogen and phosphorus output in China’shighly urbanized Pearl River Delta region[J]. Journal ofEnvironmental Management, 325:116543.
MARITORENA S, SIEGEL D A, PETERSON A R, 2002. Opti-mization of a semianalytical ocean color model for global-scale applications[J]. Applied Optics, 41(15):2705-2714.
MOORE C M, MILLS M M, ARRIGO K R, et al, 2013. Pro-cesses and patterns of oceanic nutrient limitation[J]. Na-ture Geoscience, 6:701-710.
O’REILLY J E, MARITORENA S, MITCHELL B G, et al,1998. Ocean color chlorophyll algorithms for SeaWiFS[J].Journal of Geophysical Research:Oceans, 103(C11):24937-24953.
SUN X Y, DONG Z J, ZHANG W J, et al, 2022. Seasonal andspatial variations in nutrients under the influence of natu-ral and anthropogenic factors in coastal waters of thenorthern Yellow Sea, China[J]. Marine Pollution Bulletin,175:113171.
THOMSEN E, HERBECK L S, JENNERJAHN T C, 2020.The end of resilience:surpassed nitrogen thresholds incoastal waters led to severe seagrass loss after decades ofexposure to aquaculture effluents[J]. Marine Environmen-tal Research, 160:104986.
WANG J J, YU Z G, WEI Q S, et al, 2019. Long-term nutrientvariations in the Bohai Sea over the past 40 years[J]. Jour-nal of Geophysical Research:Oceans, 124(1):703-722.
WANG X, WEN Z D, LIU G, et al, 2022. Remote estimates of to-tal suspended matter in China’s main estuaries using Landsatimages and a weight random forest model[J]. ISPRS Journalof Photogrammetry and Remote Sensing, 183:94-110.
WANG Z X, ZHANG Z Q, LI H L, et al, 2024. Water qualityin the Ma’an archipelago marine special protected area:remote sensing inversion based on machine learning[J].Journal of Marine Science and Engineering, 12(10):1742.
ZHOU L B, YANG X, LI K Z, et al, 2024. Regime shift in a coast-al pelagic ecosystem with increasing human-induced nutri-ent inputs over decades[J]. Water Research, 263:122147.
基本信息:
DOI：10.15928/j.1674-3075.202506270005
中图分类号:X55;X87
引用信息:
[1]张昕玥,徐志豪,范文杰等.中国近海溶解无机氮遥感监测及时空变化研究[J].水生态学杂志,2025,46(05):1-9.DOI:10.15928/j.1674-3075.202506270005.
基金信息:
国家自然科学基金重点项目(52239005)