Abstract
Atmosphere–ocean coupled model predictions have been hindered by the imbalance of initial states between atmosphere and ocean obtained from independent data assimilation systems. This study tests an atmosphere–ocean coupled data assimilation (CDA) method applied to a state-of-the-art coupled global climate model, the Global Seasonal Forecasting System, version 5 (GloSea5), and investigates its impacts on forecast skills. Weakly coupled data assimilation (WCDA) combines preexisting atmosphere and ocean analysis fields with the coupled model background states, for which the incremental analysis update (IAU) is employed to gradually adjust from the background states to the analysis fields yet maintain balanced states between atmosphere and ocean. While the global analysis from WCDA maintains comparable quality in the spatial distribution of temperature and precipitation to existing reanalysis datasets, it improves the tropical precipitation variability due to the atmosphere–ocean coupling. In short-range forecasting from WCDA, the widespread bias of surface air temperature is reduced, which was originally induced by the differences between sea surface temperature (SST) in the atmospheric initial conditions and that in the oceanic initial conditions. The WCDA impact on the forecast skill is more pronounced in the subseasonal time-scale Madden–Julian oscillation (MJO) forecasts by reducing initialization shock in moisture; otherwise, atmospheric convection becomes much suppressed initially and then suddenly produces a large amount of precipitation in the forecasts from uncoupled initialization.
© 2025 American Meteorological Society. This published article is licensed under the terms of the default AMS reuse license. For information regarding reuse of this content and general copyright information, consult the AMS Copyright Policy (www.ametsoc.org/PUBSReuseLicenses).
