| Sea surface salinity is a very important parameter in the marine system, it has a significant contribute to the study on marine environment, global water cycling, climate prediction and so on. In recent years, with the rapid development of the technology on sea surface salinity remote sensing by microwave satellite, monitoring salinity in a large range for a long period comes true. But SMOS and Aquarius, the two microwave satellite to remote sensing salinity of the sea surface, have not yet reached the expected accuracy. Therefore, it has great application value to carry out deep research on salinity inversion algorithm.Using the salinity inversion algorithm for aperture microwave radiometer as the main line, this article firstly introduced the basic theory of passive microwave remote sensing observation of sea surface salinity. In consideration of influencing factors coming from sea surface, atmosphere, ionosphere and space, the forward model of sea surface microwave radiation was built, and the effect of incident angle, sea surface temperature, wind speed and wind direction on the sea surface brightness temperature was analyzed. On the base of the forward model, research was carried out on the statistical regression inversion algorithm, neural network inversion algorithm and iterative inversion algorithm based on cost function; then simulation was done to analyze the error of brightness temperature, sea surface temperature and wind speed, which would influence the accuracy of salinity inversion. Finally, I compared the sea surface brightness temperature computed by the forward model with the Aquarius brightness temperature data to verify the accuracy of the forward model, then I compared the retrieval salinity obtained by the three algorithms with the Aquarius salinity data, result shows that the retrieval accuracy in open sea area reaches 0.36 psu, so we can come to the conclusion that the three inversion methods are effective. |
PDF Full Text Request