Font Size: a A A

Research On Remote Sensing Retrieval Of Chlorophyll A Concentration In The Coast Of Fujian Province Based On Sentinel Satellite

Posted on:2022-04-03Degree:MasterType:Thesis
Country:ChinaCandidate:G L XiaoFull Text:PDF
GTID:2491306320483734Subject:Hydraulic engineering
Abstract/Summary:PDF Full Text Request
As one of the three elements of water color,chlorophyll is an important reference index for characterizing and evaluating the degree of eutrophication of water bodies.At the same time,chlorophyll-a has a wide range of practical applications in measuring phytoplankton biomass,estimating marine primary productivity,red tide warning,detecting and judging marine fishing grounds,etc.Therefore,from the perspectives of ecology,society,and economy,monitoring and retrieving the concentration of chlorophyll-a are of great value and significance.In this paper,based on the OLCI remote sensing image data and the chlorophyll-a concentration data measured by the buoy,the remote sensing inversion study of the chlorophyll-a concentration in the coastal waters of Fujian is carried out.The main research conclusions are as follows:(1)The applicability of FLAASH,QUAC,and C2RCC three atmospheric correction methods in OLCI image data is compared.Through correlation analysis,it is found that the C2RCC atmospheric correction has the highest correlation between the remote sensing reflection of each band and the concentration of chlorophyll-a.And the correlation coefficient is the largest at the b21 band,which is 0.601.The correlation between the remote sensing reflectance and chlorophyll-a concentration of each band obtained by FLAASH and QUAC atmospheric correction is consistent,and they are all negatively correlated.(2)Three machine learning methods,XGBoost,Cat Boost,and random forest,were used to construct the chlorophyll-a concentration inversion model.Compare the model inversion results based on the coefficient of determination(R2),root mean square error(RMSE)and average absolute percentage error(MAPE)between the measured value and the estimated valuez.The results show that the R2,MAPE,and RMSE of the XGBoost model perform best among the three models,respectively 0.93,18.78%,and0.67μg?L-1.At the same time,it was found that the performance of the same model in different chlorophyll-a concentration ranges was different.(3)Analyzing the temporal and spatial distribution of chlorophyll-a concentration,it is found that the overall chlorophyll-a concentration in July is lower than that in June.Spatially,the chlorophyll-a concentration in coastal waters where human activities are intensive and the estuary of rivers is high,and in the far-shore waters is low.The chlorophyll-a concentration in most sea areas is concentrated in the range of0.1~3μg?L-1.
Keywords/Search Tags:Chlorophyll-a, Remote Sensing Inversion, OLCI, Color Remote Rensing
PDF Full Text Request
Related items