Satellite Salinity Meter Sea Surface Salinity Inversion Technique

Sea surface salinity(SSS) is one of the most important parameters in study of climate forecasting, global water cycle, sea ice observation, marine disasters monitoring, marine ecosystems and military field. Obtaining spaceborne microwave radiometer data on inversion of sea surface salinity can realize large-scale and continuous observation. Our coming ocean salinity satellite will carry three or more radiometers(including C, K and L band) and a L-band scatterometer. In order to take advantage of satellite data for each band and reduce error on inversion of sea surface salinity, this thesis based on simulation data gets progressively results of sea surface salinity by using stepwise retrieval, coordinated retrieval or linear regression method.This thesis mainly includes:1. This thesis introduces detailedly the Two-scale model, theSmall-Slope-Approximation and some empirical models, which are the popularmodels for computing the emissivity. While, these effecting factors, consist ofspace factors, ionosphere, atmosphere, foam and RFI are introduced in thisthesis.2. This thesis makes discussion of effects on backscattering coefficient by differentpolarization, incident angle and relative wind direction when inversing the windfield using L-band data. This thesis also analyzes the L-band scatterometermeasurement accuracy of wind speed and direction.3. This thesis adopts stepwise retrieval, coordinated retrieval and linear regressionmethod on inversion of sea surface temperature and wind speed with C/K banddata. On the basis of retrieval and accuracy calculation, this thesis confirmscoordinated retrieval as the best methods of sea surface temperature and windspeed retrieval of the three methods.4. This thesis adopts stepwise retrieval, coordinated retrieval and linear regressionmethod on inversion of sea surface salinity with L-band data. The result showsthat stepwise retrieval can find and update parameter in time. Coordinatedretrieval is the best in accuracy while Linear regression method is the fastest.We can make choice according to our requirements in practice.