Assessment of the JMA Serial Observation Lines in the Northwestern Pacific in OSSE Studies with the GFDL Ensemble Coupled Data Assimilation System
Jae Ho Lee, You‐Soon Chang, Shaoqing Zhang
Abstract
Abstract Two major serial observational positions (137°E and 165°E, operated by the Japan Meteorological Agency [JMA] in the Northwestern Pacific) were assessed using observation system simulation experiments (OSSEs) with the Geophysical Fluid Dynamics Laboratory (GFDL) ensemble coupled data assimilation system. In the OSSEs, the “True Ocean” is a simulation time series of the fully coupled GFDL climate model (CM2.1), and the “control ocean” is defined as the same CM2.1 simulation result using different initial conditions and ensemble members. The sea surface temperature, subsurface temperature, and salinity data that were drawn from the “truth” were assimilated into the “control.” The degree by which the assimilation recovers the “truth” is an assessment of the representation of the serial observation lines for regional circulations and their variability in the Northwestern Pacific. Results show that the assimilation effect of sea surface temperature is maintained to the deeper ocean by adjusting the interannual variability of the ocean temperature in the Northwestern Pacific. The root mean square error in the western area of each observational system was much lower than that in the eastern area. This is closely related to the anomalous geostrophic clockwise flow pattern strengthened throughout the assimilation process of the JMA data. We confirmed this result using a reverse bias experiment. Therefore, the assimilation effect of 165°E appearing larger western areas is also stronger than 137°E. However, the assimilation performances of the two observation systems were not significantly different in the long‐term simulations of more than 25 years in the Northwestern Pacific.