Remote Sensing Of The Eutrophic Water In The Bo Sea And The Yellow Sea

Among the ways to monitor the eutrophication along coastal waters,the traditional sail-sampling method fails to reflect the real pollution condition due to its limited searching range and weal periodicity.However,the remote sensing of ocean color can break through the limitation of space and time to flexibly collect the spectrum data of the research area for the purpose of analysis.This study,by taking the Bo Sea and the Yellow Sea as the research area,targeting the optical features of the coastal waters,and using for reference the basic principle of comprehensive trophic level index?TLI?,constructs the comprehensive trophic level index?TLIcr?applicable to the coastal waters.Besides,by analyzing the optic features of the coastal waters,the study establishes the remote inverse models of TLI_cr and offers quantitative assessment of the eutrophic degree in the research area by means of this model.The main topics of this thesis are as follows:I.Based on the TLI construction idea towards lakes,the annual average extreme value can be calculated with the help of the remote data,so the equations of the separate nutrition degrees regarding such water quality parameters as chlorophyll-a,total nitrogen,total phosphorus and transparency can be deduced.The study adopts actually measured 202-group data to define the weight coefficient of the research area's water quality parameters.With the equations of the separate nutrition degrees and the weight coefficient combined,the assessed index fitting to the eutrophic coastal waters can be worked out.II.Using NQAA way to calculate the inherent optical quantity of the offshore waters,this text,according to the research area's optical features,adopts the principal component analysis?PCA?to analyze the correlativity between the eutrophic nutrition index and the absorption coefficient,and constructs inverse models by means of the PCA and multiple linear regression analysis ways.Employing 1/3 of the actually measured date?67 groups?and 86 groups of satellite data matched to the actually measured station positions,the study verifies and compares the accuracy of inverse models,which turns out that the ones built by a?490?and a?551?parameters had the best effect and relatively high fitting accuracy?R²=0.87?.In addition,for these two models,the error made by the satellite image authentication is tinier?MAPD=13.63%?.Hence,the reliability assessment of the eutrophic satellite products is achieved.III.Based on the VIIRS data and the eutrophic inverse models,the study works out the hierarchical graphs of the regional eutrophication in the Bo Sea and the Yellow Sea from 2012 and2018,and further discusses the spatio-temporal features and driving factors of the seasonal eutrophic states in those sequential years,2018 included.Through the analysis of the eutrophic changing trends,the results shows that the nearshore eutrophication is higher than that of the offshore,the Bo Sea's is higher than the Yellow Sea's,three inner bays of the Bo Sea are higher than the Central part in terms of the eutrophic degree,and the seasonal change of the high-to low eutrophication is in the sequence of winter,fall,spring and summer.To sum up,the inverse results in this text present that in terms of eutrophication,the nearshore is higher than the offshore,the Bo Sea than the Yellow Sea and the fall than the water,in accordance with the eutrophic trends of the coastal waters,which further verifies the reasonability,feasibility and applicability of the eutrophic index's inverse models.