| Oceancolor remote sensing refers to the long-distance real-time observation of ocean color by sensors mounted on aqua satellites.It has become an important means of global marine environment monitoring as the unique advantages of wide coverage and all-day time.The key of oceancolor remote sensing lies in atmospheric correction and oceancolor inversion.Since the complex optical properties of water bodies in the coastal areas of the Bohai Sea in China,the use of atmospheric correction algorithms and chlorophyll inversion algorithms based on case I water bodies is unqualified to monitor coastal environment through oceancolor remote sensing technology.Determining opportune atmospheric correction algorithms and oceancolor inversion algorithms plays a significant part in implementing real-time environmental monitoring in coastal waters of the Bohai Sea by oceancolor remote sensing technology.Based on the NPP-VIIRS satellite products,MUMM,STD and NIR-SWIR three atmospheric correction algorithms will be selected in this study to correct the satellite imagery of Qinhuangdao and Liaodong Bay.Correction accuracy of each algorithm will be evaluated,and determine which atmospheric correction algorithm is more applicable to the offshore areas of the Bohai Sea.Meanwhile,regional chlorophyll-a concentration inversion algorithms for the coastal area of the Bohai Sea will be established based on the in situ chlorophyll-a concentration,and remote sensing reflectance data.The final research results show:(1)Compared with the other two atmospheric correction algorithms,the remote sensing reflectance obtained by the NIR-SWIR algorithm correction was positively correlated with in situ data from 410 nm to 671 nm,reflecting good correction results at the overall level precision.The highest accuracy appeared at 551 nm,with the relative error of 15%,and the R2 of 0.90.The MUMM correction algorithm had large correction errors at 410 nm and 443 nm,and showed a negative correlation with the in situ measurements,but accuracy of the corrected results at 486 nm to 671 nm had improved,and the highest accuracy appeared at 551 nm,even higher than NIR-SWIR algorithm,the relative error was 12%,R2 was higher to 0.96.The STD iterative algorithm only had a negative correlation with the in situ data at 410 nm,and the accuracy at 443 nm to 671 nm was similar to the NIR-SWIR algorithm.(2)The study for applicability of the three atmospheric correction algorithms in different water bodies shows that the assumption that the water radiance in the near-infrared and short-wave infrared bands is 0,which is close to the characteristics of the in situ watermass,due to the low turbidity of Qinhuangdao's water body.Thus,the STD iterative algorithm with NIR-SWIR algorithm shows higher correction accuracy in Qinhuangdao.However,correction accuracy decreases in Liaodong Bay with high turbidity,as the assumption mentioned above is in vain in this area;although the MUMM algorithm performs more stably in Liaodong Bay,the correction accuracy of this algorithm reduces a lot due to the large deviation of the parameter ?(745,862)in this algorithm compared with in situ ?(745,862)in Qinhuangdao.(3)According to the in situ Rrs(?)ratio and Chl-a,three inversion algorithms based on OC3V,RGCI and Tassan models were constructed.The inversion accuracy of new algorithms have been improved compared to satellite algorithms,and the RE values have been reduced to 65%,62%and 31%,respectively.Judging from the evaluation results,the Chl-a inversion algorithm based on the Tassan model is more suitable for the extraction of Chl-a along the Bohai Sea coast.However,there still exists some uncertainty due to the lack of few synchronous in situ data in the inversion accuracy assessment.Thus,synchronous data with wider spatiotemporal coverage is needed to be involved into the verification in the future. |
PDF Full Text Request