Two kinds of Bayesian-based cost functions (i.e., the unconstrained cost function and parameter-constrained cost function) are investigated for retrieving the sea surface salinity (SSS). In low SSS regions, we have an...Two kinds of Bayesian-based cost functions (i.e., the unconstrained cost function and parameter-constrained cost function) are investigated for retrieving the sea surface salinity (SSS). In low SSS regions, we have analyzed the sensitivity of the two cost functions to geophysical parameters. The results show that the unconstrained cost function is valid for retrieving several parameters (including SSS, wind speed and significant wave height), and the constrained cost function, which largely depends on the accuracy of reference values, may lead to large retrieval biases. Furthermore, as a retrieval parameter, the sea surface temperature (SST) can re-sult in the divergence of other geophysical parameters in an unconstrained cost function due to the strong sensitivity of brightness temperature to SST. By using the unconstrained cost function and the simulated brightness temperature TB with white noises, the retrieval biases of SSS are discussed with the following two procedures. Procedure a): the simulated TB values are first averaged, and then SSS is retrieved. Procedure b): the SSS is directly retrieved from the simulated TB , and then the retrieved SSS values are aver-aged. The results indicate that, for low SSS and SST distributions, the SSS retrieval by procedure a) has less biases compared with that by procedure b), while the two procedures give almost the same retrieval results for high SSS and SST sea regions.展开更多
基金supported by the National Natural Science Foundation of China (Grant No. 40876094)the National 863 Project of China (Grant Nos. 2009AA09Z102 and 2008AA09A403)
文摘Two kinds of Bayesian-based cost functions (i.e., the unconstrained cost function and parameter-constrained cost function) are investigated for retrieving the sea surface salinity (SSS). In low SSS regions, we have analyzed the sensitivity of the two cost functions to geophysical parameters. The results show that the unconstrained cost function is valid for retrieving several parameters (including SSS, wind speed and significant wave height), and the constrained cost function, which largely depends on the accuracy of reference values, may lead to large retrieval biases. Furthermore, as a retrieval parameter, the sea surface temperature (SST) can re-sult in the divergence of other geophysical parameters in an unconstrained cost function due to the strong sensitivity of brightness temperature to SST. By using the unconstrained cost function and the simulated brightness temperature TB with white noises, the retrieval biases of SSS are discussed with the following two procedures. Procedure a): the simulated TB values are first averaged, and then SSS is retrieved. Procedure b): the SSS is directly retrieved from the simulated TB , and then the retrieved SSS values are aver-aged. The results indicate that, for low SSS and SST distributions, the SSS retrieval by procedure a) has less biases compared with that by procedure b), while the two procedures give almost the same retrieval results for high SSS and SST sea regions.
