Estimating Sea Surface Salinity by Using Moderate Resolution Imaging Spectroradiometer (MODIS) Data- A Case Study Taiwan Strait

• Main Authors: Basmah Mohammad Mufleh Alabbadi, 芭絲瑪
• Other Authors: LiChen
• Format: Others
• Language: en_US
• Published: 2015
• Online Access: http://ndltd.ncl.edu.tw/handle/g4vdq6

博士 === 中華大學 === 土木工程學系博士班 === 103 === The largest amount of the water is stored in the oceans. However, due to global warming, changes in the ocean water quality parameters such as salinity have become clearer. Continuous estimation of water quality parameter in the ocean is important because ocean plays a key role in water cycle between land, atmosphere and ocean, also serve as a source of human food. New technology of remote sensing in water quality parameters is to predict data of ocean parameters by using remote sensing images that reflect a wide range of spatial and temporal changes in ocean water and provides data for wide area, at short time and less cost. Nevertheless, some remote sensing satellites are calibrated for land use; therefore, their signal-to-noise ratio for a low-reflectance ocean-water surface is not suitable for obtaining substantial data. This thesis discussed estimates of sea surface water quality parameter from a Moderate-Resolution Imaging Spectroradiometer (MODIS) images and selected Taiwan Strait as the study area. MODIS/Terra satellites which provide 500 m resolution image was used to study sea surface salinity (SSS). The regression analysis (RA1 and RA2), and genetic algorithm combining operation tree (GAOT) were used to establish the predictive models by using image data and in situ salinity data. The genetic algorithm combining operation tree (GAOT) is a data mining method used to automatically discover relationships among nonlinear systems. Based on genetic algorithms, the relationships between input and output can be expressed as parse trees. The GAOT method typically has the disadvantages of premature convergence, which means it cannot produce satisfying solutions and performs satisfactorily when applied to only low-dimensional problems. Therefore, the GOAT method was enhanced using an automatic-incremental procedure to improve the search ability of the method and avoid trap in a local optimum. The GAOT model was improved by run the two independent GAOT, GAOT1 and GAOT2. The second run used the results of the two independent GAOT as inputs to obtain optimal results. These accurate results indicate that the IGAOT model performs more favorably than do the GAOT, and regression analysis models (RA1 and RA2), exhibits higher correlation coefficients (CCs), and involves fewer estimating errors. The results of this research indicated that the proposed technique is useful for estimating the strait salinity.