Impact of assimilating altimeter data on eddy characteristics in the South China Sea

Abstract

From satellite altimetry, it is known the benefit of assimilating sea level anomalies (SLA) has been shown in the context of operational ocean forecast systems. However, how much data assimilation (DA) of altimetry data improves the representation of mesoscale eddies has still not been investigated in previous studies. Especially in the South China Sea (SCS), no estimation for that has been done in a long time. In this study, a nested SCS model system uses the Ensemble Optimal Interpolation method to assimilate along-track SLA data from 1993 to 2011. We assess the representation of eddy characteristics in two hindcast simulations – one with DA and the other without – and compare them to satellite-derived eddy characteristics. In the whole SCS, the assimilation improves the number of simulated cyclonic (anticyclonic) eddies by 10.3% (13.6%). The corresponding improvement in the eddy-rich northern SCS is 17.9% (19.6%). Assimilation improves the seasonality of eddy occurrence, with cyclones and anticyclones showing an obviously asymmetric seasonality. However, diagnosed assimilation effects and associated residual errors show large spatial and temporal dependencies. The radii of anticyclonic eddies smaller than 70 km are not changed with DA. The results show that deficiencies of cyclonic eddies in winter and anticyclonic eddies in summer around 13∘N in the southeastern SCS are not well corrected by DA, where one of the shortcomings resulted from the used wind forcing. Although this study does not conduct one-to-one forecasting experiments for each eddy track, improved eddy reproduction is the first step towards detailed validation metrics for eddy forecasting systems.