Reconstruction Of The Remote Sensing Chlorophyll-a In Water And Applications

In recent years,eutrophication has been widely acknowledged to be a serious problem of hydroenvironment.Analysis of eutrophication is important for hydroenvironment protection.Chlorophyll-a(Chl-a)plays an important role in the assessement of eutrophycation.It is the mean pigment of phytoplankton for the photosynthesis and is a parameter for the measurement of phytoplankton biomass.Remote senseing technology has developed quickly recently,and provies a new access for Chl-a data.It has the advantages of wide spatial coverage,high time efficiency,good time consistency and low economic cost.Remote sensing Chl-a products are reliable data source for research on the spatial and temporal variability of the highly dynamic surface environment for its long time consistency and real time observation.However,influenced by the weather condition and failure of sensor operation,massive and irregular missing gaps can not be prevented in the remote sensing Chl-a products.In addition,the spatial resolutions of the remote sensing products are relatively lower and are hard to provide information in small scale for regional application.These facts are two obstacles in the application of remote sensing data and in the development of hydroenvironment remote sensing.Therefore,researching on remote sensing data reconstruction and dowscaling to obtain complete high resolution Chl-a dataset are benefitial for broadening and deepening the application of remote sensing products.To solve these problems,an EOF-based method(DINEOF)was first modified and improved to build a more efficient and accurate geophysical dataset filling method(VF-DINEOF).Secondly,a data fusion model based on the statistical dowscaling was developed to improve the resolution of the remote sensing Chl-a products.At last,a complete Chl-a and its releted climate variables dataset was built.Temporal and spatial variability of the Chl-a concentration in the Bohai Sea was studied and the mechanism of the patterns were identified.The main content of this dissertation are listed below:(1)The DINEOF method was studied and the remote sensing Chl-a data of the Bohai Sea from 2003?2014 was reconstructed.The accuracy of this method was also analysed.Then,the Variable Fuzzy theory was introduced to the original method.The whole Bohai Sea was partitioned into 3 subregions with the Variable Fuzzy Chlustering Iteratve Mode and an improved DINEOF method(VF-DINEOF)was developed.The new method is effective in reconstruction of remote sensing data and successfully obtained complete Chl-a concentration dataset.The results show that the new method improved the accuracy and efficiency.(2)A data fusion model based on statistical downscaling was developed to obtain finer resolution dataset of Chl-a.The MODIS L3 Chl-a product with resolution of 4 km and the Landsat 8 OLI spectral band data with resolution of 30 m were fused to combine both the advantages of the two satellite sensors.30 m resolution Chl-a concentration data in the study area was obtained and in-situ samples was used for validation.Results show that the method successfully improved the resolution of the original Chl-a data.The results were reliable for the accuracy of this method was within the 35%limit of the goal of remote sensing Chl-a retrieval.The downscaling data fusion results illustrated the details of the coastal environment and were benefitial for the research on the coastal eutrophication.(3)The remote sensing MODIS L3 Chl-a product and releted climate variables were reconstructed by the VF-DINEOF method and a temporal and spatial complete long term dataset of the Bohai Sea was obtained.The temporal and spatial variability of the Chl-a concentration was analysed by EOF.The Chl-a concentration varied greatly in spatial scale.It was higher in coastal area and shallow waters where higer risk of red tide exsits.The seasonal characteristics of the Chl-a were also identified.There were two peaks during the year which were in spring and early autumn in the Laizhou Bay,Bohai Bay,and the Liaodong Bay.Mann-Kindle non-parametric test and Hurst index methods were used.It revealed that the Chl-a concentration in the whole Bohai Sea would continue to increase in the future except for the east coast of the Laizhou Bay and the coast of Dalian.The spatial distribution of the Chl-a is basically driven by the environmental factors.Mariculture activity,river discharge and suspended sediment also influence Chl-a.