Development of an “on-line” coupled data assimilation capability for Regional Community Earth System Model

Abstract:

A Regional Community Earth System Model (R-CESM) has been developed at the International Laboratory for High-Resolution Earth System Prediction (iHESP). The R-CESM couples ROMS, WRF and CLM within the framework of CIME (Common Infrastructure for Modeling the Earth) – the coupling infrastructure used in the latest version of CESM2. Developing a data assimilation (DA) capability for the R-CESM is an important step towards enhancing our predictive capability for climate extremes and their regional impact. In this talk, we introduce a new DA system within the R-CESM framework. This system utilizes an “on-line” ensemble coupled data assimilation (ECDA) procedure originally developed by Zhang et al (2005, 2007) that embeds an ensemble Kalman Filter algorithm into a forecast model as a set of subroutines, hence enables very high computational efficiency. The new DA system is configured in the Gulf of Mexico with spatial resolutions of 9km for WRF and 3km for ROMS, respectively. We tested the DA system by assimilating sea surface height (SSH) and SST. In a perfect model framework, SSH shows a stronger impact on the model analysis and forecast than SST. Ensemble forecasts initialized from an analysis of assimilating Aviso Sea level Anomaly only show similar skills to those initialized from Copernicus reanalysis. This new ECDA enabled R-CESM will serve as a platform for regional climate prediction studies in the Gulf of Mexico.