| Summary: | We suggest to transfer the empirical downscaling methodology, which was developed mostly for atmospheric dynamics and impacts, to regional ocean problems. The major problem for doing so is the availability of decades-long and homogeneous and spatially detailed data sets. We have examined the performance of the STORM multidecadal simulation, which was run on a 0.1° grid and forced with 1950–2010 NCEP re-analyses, in the South China Sea and found the data suitable. For demonstration we build with this STORM-data downscaling model for the regional throughflow.
The STORM data is compared with AVISO satellite observations and the ocean re-analysis dataset C-GLORS. We find the seasonal patterns and the inter-annual variability of sea surface height anomaly in both the C-GLORS data and the STORM simulation consistent with the AVISO-satellite data. Also the strong westward intensification and the seasonal patterns of South China Sea circulation steered by the monsoon have been presented well. As an important indicator of vertical movement, the sea surface temperature distribution maps are also very close, especially for the narrow upwelling region in summer. We conclude that the output of the STORM simulation is realistically capturing both the large-scale as well as the small-scale dynamical features in the South China Sea.
|
|---|