Intercomparison of high-resolution regional reanalysis with global products over the Red Sea

Abstract:

High-resolution surface atmospheric fields play a crucial role in modeling the ocean surface and subsurface processes. For narrow sea basins surrounded by complex topography, like the Red Sea, available global reanalyses may not supply accurate boundary information to ocean models because of their coarse resolution. To fulfill the respective needs of ocean models in the Red Sea, we have generated high-resolution regional reanalysis by assimilating available observation using the Weather Research and Forecasting (WRF) model for a period of one year (2009). The three-/four-dimensional variational data assimilation (3/4DVAR) methods were used for assimilation. We have evaluated the quality of these regional reanalysis products using observations from synoptic stations, scatterometers and buoy. We have also compare the outputs with three global reanalysis datasets of different resolutions; NCEP-GFS, ERA-Interim and Japanese 55 Year Reanalysis. The comparison suggests that the seasonal pattern of wind is reasonably reproduced in all the reanalysis products, however there is an increased sensitivity around mountain gaps for high-resolution reanalysis, where the channeling effect is well represented in 4DVAR with an accurate spatial structure. Among the ocean surface processes, waves are one of the quick and direct responders of changes in wind fields. We have considered a wave model set up for the Red Sea to evaluate the relevance of the different reanalysis products. Comparison with buoy observations indicates that wave simulations resulting from the regional reanalysis are more accurate than the global reanalysis products.