An Improved Coupled Data Assimilation System with a CGCM Using Multi-Timescale High Efficiency EnOI-Like Filtering

Coupled data assimilation (CDA) combining coupled models and observations plays a critical role in climate studies by producing a four-dimensional estimation of Earth system states. However, traditional CDA algorithms while being expensive lack sufficient representation of multi-scale background flows. Here, a Multi-timeScale High-Efficiency Approximate EnKF (MSHea-EnKF) has been implemented in the global fully coupled climate model of the Geophysical Fluid Dynamics Laboratory. It consists of stationary, low-frequency, and high-frequency filters constructed from the timeseries of a single model solution, with improved representation for low-frequency background error statistics and enhanced computational efficiency. The MSHea-EnKF is evaluated in a biased twin experiment framework with synthetic “observations” produced by the other coupled model, Community Earth System Model, and a three-decade coupled analysis experiment with real observations. Results show that while computationally costing only a small fraction of traditional ensemble CDA, the MSHea-EnKF significantly improves the assimilation quality due to better representation of the slow-varying background flows in the filtering. The coupled analysis of MSHea-EnKF also improves the estimation of the Atlantic meridional overturning circulation, including better standard deviation distribution and mass transport at the Rapid section. These results of MSHea-EnKF on prevalent resolution coupled model with high computational efficiency promises its further applications to high-resolution coupled model data assimilation and reanalysis which will greatly advance our understanding for seamless weather-climate analysis and predictions.Coupled data assimilation (CDA) combining coupled models and observations plays a critical role in climate studies by producing a four-dimensional estimation of Earth system states. However, traditional CDA algorithms while being expensive lack sufficient representation of multi-scale background flows. Here, a Multi-timeScale High-Efficiency Approximate EnKF (MSHea-EnKF) has been implemented in the global fully coupled climate model of the Geophysical Fluid Dynamics Laboratory. It consists of stationary, low-frequency, and high-frequency filters constructed from the timeseries of a single model solution, with improved representation for low-frequency background error statistics and enhanced computational efficiency. The MSHea-EnKF is evaluated in a biased twin experiment framework with synthetic “observations” produced by the other coupled model, Community Earth System Model, and a three-decade coupled analysis experiment with real observations. Results show that while computationally costing only a small fraction of traditional ensemble CDA, the MSHea-EnKF significantly improves the assimilation quality due to better representation of the slow-varying background flows in the filtering. The coupled analysis of MSHea-EnKF also improves the estimation of the Atlantic meridional overturning circulation, including better standard deviation distribution and mass transport at the Rapid section. These results of MSHea-EnKF on prevalent resolution coupled model with high computational efficiency promises its further applications to high-resolution coupled model data assimilation and reanalysis which will greatly advance our understanding for seamless weather-climate analysis and predictions.