| Ocean salinity is a very important dynamic parameters, for the study of the ocean thermohaline circulation and the global water cycle. Although satellite remote sensing of sea surface salinity (SSS) has achieved rapidly development, the existing satellite remote sensing methods of the ocean sea surface salinity cann’t fulfill salinity retrieval needs in some regions.This paper focuses on improving the accuracy of SSS with satellite microwave remote sensing, involving error analysis of SSS products, the method of reversion, error correction in SSS product, and new detection methods for future satellite missions.The main research work includes the following aspects:(1) The commonly used L-band dielectric constant models were compared. The two-scale model, small slope approximation model and empirical model were also compared.(2) Analysis of the error sources of SSS of spaceborne microwave radiometer. On a large scale, we analyzed the error contributions of SSS inversion from sea surface wind speed and temperature using simulation methods.(3) The SSS regions error characteristics are studied, by comparing SSS of the SMOS, Aquarius and in-situ measurment salinity. The relationship formula of error of SSS and SSS change were deduced. Regional analysis and calculation of error characteristics of different salinity data products concluded that there is a large part of the Inherent bias in some regions; at the SSS stratification obvious offshore area, the differences of the SSS change of data products between the two radiometers less than the SSS change of data products between the radiometers and in-situ measurment.(4) The SMOS brightness temperature data with incidence angle of35～55°are used to invert SSS, in order to reduce the brightness temperature measurement error associated with the angles. The SMOS SSS in the Pacific Northwest (0N-30N,120E-150E) near the South China Sea are retrieved. The accuracy of retrieved SSS with the two methods was estimated by comparing it with the Argo SSS and ESA level2SSS. The results show that the methed improves the SMOS salinity retrieval accuracy. (5) A method is proposed for the correction of SSS regional bias of the monthly level3products. Based on the stable characteristics of the large SSS biases from month to month in some regions, corrected SMOS SSS maps can be obtained from the monthly mean values after removing the regional biases. SMOS and ISAS SSS were also compared with Aquarius SSS to show the advantage of SMOS SSS with the correction.The accuracy of the SMOS SSS measurements is greatly improved, especially near coastlines, high latitudes and some open ocean.(6) The joint SSS inversion method with L, S, C, X-band radiometer is proposed. When the error of radiometer brightness temperature is0.5K, SSS accuracy of the joint inversion by multi-frequency radiometer is better than0.2psu in middle and low latitudes regions. In the higher latitudes regions, the SSS accuracy from L, S, C, X-band radiometers is better than the ones from L, C, X-band radiometers.(7) Principle and simulation of retrieving SSS from GNSS_R signal was given for the frist time in attempt of evaluating its possibility of sensing SSS. When1psu is changed in SSS, the normalized scattering cross-section (NSRS) change is less than0.05dB according to the second-order small slope approximation (SSA2) simulation calculation. When1psu is changed in SSS, polarization coherent phase change is about0.4°according to Fresnel equations and geometrical optics (GO) approximation model. The results showed that the measurement accuracy of SSS can not be achieved directtly by GNSS_R at present. |
PDF Full Text Request