The Yin-He Global Ocean Data Assimilation and Forecast System
Yan Chen, Senliang Bao, Yu Cao, Weimin Zhang, Huizan Wang
Abstract
This study successfully developed a Yin-He Global Ocean Data Assimilation and Forecast System (YHGO), which commenced operational trial runs on 2024 August 15. The system innovatively employs a particle filter-based assimilation method—the localized weighted ensemble Kalman filter—overcoming the limitations of traditional approaches that assume linear or weakly nonlinear models and observation operators while effectively handling non-Gaussian background errors. It incorporates a mass conservation ocean model, which better represents real ocean physics than conventional volume-conserving models. Additionally, it integrates an intelligent reconstruction module that derives subsurface temperature and salinity profiles from satellite sea-surface observations, which are then assimilated as pseudo-observations. Validation based on daily 1- to 10-d forecasts from 2024 August 15 to 2024 November 30, following the Global Ocean Data Assimilation Experiment OceanView Intercomparison and Validation Task Team (GOV-IVTT) Class 4 framework, demonstrates that YHGO achieves higher accuracy in sea-surface temperature and sea-level anomaly predictions than the widely used GLO12V4 system. However, systematic biases remain in the temperature and salinity profile forecasts, which represent key areas for future improvement and optimization to enhance the overall performance of the system.