The Study Of Sea Surface Salinity Model Based On Multi-source Remote Sensing Coordinated Retrieval

Sea Surface Salinity is one of the important parameter to describe the state of theocean. It is significant to study the variation and distribution of sea surface salinity, sothat the characteristics of the ocean itself and the knowledge about the role it plays inthe complex ocean-atmosphere system could be better understood. Meanwhile,meteorology, ecology, hydrology, fishery and other disciplines also focus on theextraction of sea surface salinity. The sea surface salinity retrieved from optical ormicrowave remotely sensed data is of advantage and disadvantage respectively. Bycoordinating above two remote sensing data effectively, the retrieval of sea surfacesalinity could be improved in accuracy. In this article, the research area is located inthe Hong Kong waters, near the eastern Pearl River of the northern South China Sea;the sample data is remote sensing data and measured sea surface data which werecollected from the year of2009-2011; the experiment is carried out to study theretrieval of sea surface salinity by coordinating optical and microwave remote sensingdata. Ultimately, the coordinated retrieval model of sea surface salinity based onETM+and SAR data was proposed. The following main research work is conducted:(1) From the perspective of single remote sensing data, study the mechanism ofsea surface salinity retrieved from optical and microwave data, and propose the seasurface salinity retrieval algorithm based on single remote sensing data. First, yellowsubstance concentration is obtained from the improved empirical spectral index andthe half empirical and half physical sea surface radiation transfer model; with yellowsubstance concentration as mediator, sea surface salinity is indirectly retrieved fromhyperion data algorithm. Secondly, sea surface brightness temperature is obtainedfrom the scattering coefficient of SAR data; sea surface salinity is collected fromretrieved sea surface brightness temperature on the basis of K-S model.(2) From the perspective of coordinating optical and microwave remote sensingdata fusion, put forward a method in which SAR data interpolates ETM+missing data,and remove the cloud and its shadow in ETM+images with SAR data, which has theinherent characteristic, that is, SAR data could penetrate through the cloud and fog. According to pixel position matching algorithm, SAR images can be converted tocloudless and shadow-free images based on ETM+pixels; replace the original ETM+cloud pixels by the corresponding cloudless pixels, then, the new ETM+images aregenerated after SAR interpolates ETM+missing data. In view of optical imagesaffected by cloud, the missing data interpolation method coordinates with the seasurface salinity retrieved from hyperion data algorithm, which could obtain seasurface salinity in some accuracy.(3) From the perspective of coordinating optical and microwave remote sensingmodel coupling, propose a method in which sea surface salinity is retrieved fromcoordinated ETM+and SAR data. The sea surface brightness temperature and seasurface emissivity are extracted from ETM+and SAR data respectively, the seasurface temperature is obtained from sea surface temperature retrieval algorithm, andthe sea water dielectric constant is extracted from SAR data; then, all above values areput into the Debye equation. Ultimately, the sea surface salinity coordinated retrievalmodel based on optical and microwave remote sensing model coupling is established.Verification analysis with the measured sea surface salinity shows that the coordinatedmodel displays higher multiple correlation coefficient (R2=0.8672) and smaller rootmean square error (RMSE=0.6253) than the other methods. The retrieved result ofcoordinated model is superior to any retrieved results using singe remote sensing data.The innovations of this article are:(1) the proposal of the retrieval methodcoordinating optical and microwave remote sensing data fusion. The method can beapplied to sea surface salinity retrieval research under the circumstance of shortage ofcloudless optical images;(2) the establishment of sea surface salinity coordinatedretrieval model based on ETM+and SAR data, according to coordinated retrievalmethod based on optical and microwave remote sensing model. It is not only increasethe retrieval accuracy of sea surface salinity, but also improve the adaptability of thesea surface salinity retrieval method.