Development of a Convection-Permitting Air-Sea-Coupled Ensemble Data Assimilation System for Tropical Cyclone Prediction

A regional-scale fully coupled data assimilation (DA) system based on the ensemble Kalman filter is developed for a high-resolution coupled atmosphere-ocean model. Through the flow-dependent covariance both within and across the oceanic and atmospheric domains, the fully coupled DA system is capable of updating both atmospheric and oceanic state variables simultaneously by assimilating either atmospheric and/or oceanic observations. The potential impacts of oceanic observations, including sea-surface temperature, sea-surface height anomaly, and sea-surface current, in addition to the observation of the minimum surface pressure at the storm center (HPI), on tropical cyclone analysis and prediction are examined through observing system simulation experiments of Hurricane Florence (2018). Results show that assimilation of oceanic observations not only resulted in better analysis and forecast of the oceanic variables but also considerably reduced analysis and forecast errors in the atmospheric fields, including the intensity and structure of Florence. Compared to weakly coupled DA in which the analysis update is performed separately for the atmospheric and oceanic domains, fully coupled DA reduces the forecast errors of tropical cyclone track and intensity. Results show promise in potential further improvement in tropical cyclone prediction through assimilation of both atmospheric and oceanic observations using the ensemble-based fully coupled DA system. Plain Language Summary Air-sea interactions are critical to tropical cyclone (TC) development. However, oceanic state variables are still poorly initialized and are inconsistent with atmospheric initial fields in most operational coupled TC forecast models. In this study, the ensemble Kalman filter is used to develop a regional-scale air-sea fully coupled data assimilation (DA) system for coupled atmosphere-ocean prediction. The potential impacts of oceanic observations, including sea surface temperature, sea surface height anomaly, and sea surface current on TC analysis and prediction, are examined through observing system simulation experiments of Hurricane Florence (2018). Results show that, using the flow-dependent covariance within and across the oceanic and atmospheric domains, the fully coupled DA system is capable of updating both atmospheric and oceanic state variables simultaneously by assimilating either atmospheric and/or oceanic observations. The track, intensity, and sea surface height forecasts of Florence are improved by assimilating oceanic observations. When compared to weakly coupled DA, in which the analysis update is performed separately for the atmospheric and oceanic domains, fully coupled DA reduces the forecast errors of TC track and intensity. It shows great promise in potential further improvements in TC prediction through using the ensemble-based fully coupled DA system. Key Points A convection-permitting air-sea-coupled ensemble data assimilation (EDA) system for improving tropical cyclone prediction is developed The coupled EDA system is tested through observing system simulation experiments (OSSEs) of Hurricane Florence (2018) Assimilation of oceanic observations using the coupled EDA system improves the analysis and prediction of Florence