| Sea surface salinity (SSS) is an important parameter in describing the basic properties of the seawater, and plays a vital role in the global water cycle and the ocean circulation, and is also an important indicator of global climate change. The South China Sea is the largest waters of China, and is also the largest marginal sea of the Northwest Pacific, the change of SSS in it will have a significant impact on the hydrodynamic environment and ocean circulation in the South China Sea, climate change in China, marine fisheries and so on.So far, satellite remote sensing as the only viable method in large-scale, continuous SSS observation, overcomes the difficulty that the in-situ SSS data cannot meet the needs of scientific research, becoming an effective means of obtaining SSS. However, the special geographical location and complex climate characteristics of the South China Sea seriously affecting the accuracy of satellite remote sensing SSS there. This paper combines SMOS satellite data and Aquarius/SAC-D satellite data with in-situ data, supplemented by SODA, HYCOM/NCODA data sets, analyzing the SSS characteristics and evaluating SMOS SSS accuracy in the South China Sea, in order to provide data and observation results for the prospective improvement of satellite remote sensing SSS retrieval accuracy. The main work and results are as follows:(1) Using 2011 SODA salinity data to analyze the spatial distribution characteristics of SSS in the South China Sea, finding that SSS decrease progressively from the northeast to the southwest overall. Salt water passes in and out from Luzon and Taiwan channel, the volume change from offshore estuary, all lead to SSS anomaly in partial South China Sea.(2) The seasonal SSS distribution maps of the South China Sea show that:The SSS distribution there is banded, and has an apparent seasonal evolution, indicates different seasonal variation in different waters. Using the least square method linear fitting to analyze the SODA monthly averaged SSS data for observing the SSS abnormal change trend. The results indicate that SSS in the South China Sea has a declining trend overall, from 1980 to 2011.(3) Using the Empirical Orthogonal Function (EOF) analysis method to undertake spatial and temporal decomposition on the SODA monthly averaged SSS data. The first EOF mode reveals SSS in the South China Sea do have a declining trend and accord with the result in the first part. Besides, the second and third EOF modes show that in different waters of South China Sea, the variability of SSSA is different. In the south waters and north waters, the SSSA changes largely, and has a negative correlation. While, the SSSA changes little near to the central.(4) Choosing the SMOS satellite CATDS V1.0 data set, matching with the in-situ Argo data, and evaluating the accuracy of CATDS V1.0 data set by using statistics methods. The results show that the matching data sets have no significant linear relationship, the RMSE of the CATDS V1.0 data set and Aquarius/S AC-D satellite L3 SMI data set in the South China Sea is 0.8 and 0.689 respectively, and the satellite SSS data values are relatively high comparing to the in-situ data in waters adjacent to the continental shelves. This is probably due to strong offshore wind field and radio frequency interference (RFI) from the north of the South China Sea.(5) Analyzing the SMOS CATDS V1.0 SSS1 variability on small scales, observing and validating the mismatch between in-situ and satellite measurements in spatial coverage. The results show:in one 1° × 1° box, different locations buoys SSS present certain gradient variation at the same time, and the satellite data get a large difference comparing with in-situ data. Meanwhile, comparing among the daily averaged eight buoys in-situ SSS measurements and the closest HYCOM simulated SSS shows the existence of big difference. Therefore, reducing the effects of mismatch in spatial coverage can largely improve the applicability and validity of satellite remote sensing accuracy assessment. |
PDF Full Text Request