11:15 AM - 11:30 AM
[MGI29-09] Improving the salinity structure by AOEI in an EnKF-based ocean data assimilation system
Keywords:Ocean data assimilation, Ensemble Kalman filter, Adaptive observation error inflation, Salinity, Dynamical stability
With the enhanced satellite and in-situ observations such as Himawari-8 and Argo profiling float, the number and frequency of observations are dramatically increased since the 2000s. We have demonstrated that the combination of the incremental analysis updates (Bloom et al. 1996) and relaxation-to-prior methods (Zhang et al. 2004; Whitaker and Hamill 2012) significantly improved geostrophic balances and reproducibility in an Ensemble Kalman filter (EnKF)-based ocean data assimilation system assimilating observations as frequently as every day (Ohishi et al. in prep.). However, we have found that the low salinity structure in the ocean interior was degraded in the mid-latitude regions, and that the signals propagated toward subtropical regions. Therefore, the salinity field was not well reproduced in the subtropical and mid-latitude regions.
To investigate the cause of the salinity degradation, we perform salinity budget analyses using outputs from the ocean data assimilation system. Here, we conserve all terms in the temperature and salinity budget equations and analysis increments. Also, we estimate density increments from temperature and salinity increments. Cooler and warmer temperature increments near the surface and in the sub-surface, respectively, weakens density stratification in the mid-latitude region. These result in stronger vertical diffusions of salinity and eventually lead to the degradation.
Innovation statistics of Desroziers et al. (2005) estimated from observations and ensemble forecasts imply that representation errors exist in the mid-latitude region with an abundance of fronts and eddies. Minamide and Zhang (2017) developed the adaptive observation error inflation (AOEI) method for all-sky satellite radiance data assimilation in the atmosphere to account for large differences of satellite radiances between cloudy and clear-sky conditions. Here we apply the AOEI method to the ocean data assimilation system for the first time to consider the representation errors in the mid-latitude region. The results show that the AOEI successfully decreases the temperature increments and reduces strong vertical diffusion, so that the salinity structure is greatly improved.
To investigate the cause of the salinity degradation, we perform salinity budget analyses using outputs from the ocean data assimilation system. Here, we conserve all terms in the temperature and salinity budget equations and analysis increments. Also, we estimate density increments from temperature and salinity increments. Cooler and warmer temperature increments near the surface and in the sub-surface, respectively, weakens density stratification in the mid-latitude region. These result in stronger vertical diffusions of salinity and eventually lead to the degradation.
Innovation statistics of Desroziers et al. (2005) estimated from observations and ensemble forecasts imply that representation errors exist in the mid-latitude region with an abundance of fronts and eddies. Minamide and Zhang (2017) developed the adaptive observation error inflation (AOEI) method for all-sky satellite radiance data assimilation in the atmosphere to account for large differences of satellite radiances between cloudy and clear-sky conditions. Here we apply the AOEI method to the ocean data assimilation system for the first time to consider the representation errors in the mid-latitude region. The results show that the AOEI successfully decreases the temperature increments and reduces strong vertical diffusion, so that the salinity structure is greatly improved.