There are ongoing efforts to establish an ensemble data assimilation and prediction system for tropical cyclones based on the finite-volume cubed-sphere (FV3) dynamic core with the capability to assimilate satellite all-sky infrared andmicrowave observations. To complement the system developments and improve our understanding of the assimilation of all-sky infrared and microwave observations, this study assesses their potential impacts on the analysis of Hurricane Harvey (2017) through examinations of the structure and dynamics of the ensemble-based correlations as well as single observation data assimilation experiments, using an ensemble forecast generated by a global-Toregional nested FV3-based model. It is found that different infrared and microwave channels are sensitive to different types of hydrometeors within different layers of the atmosphere, and the correlations vanish beyond 200 km in the region covered by cloud or abundant hydrometeors. The spatial correlations between brightness temperatures and model states will adjust the structure and intensity of the hurricane in the model so that the simulated hurricane will better fit the observed brightness temperatures. In general, these results show how assimilating infrared and microwave together can improve the analyses of tropical cyclone intensity and structure, which may lead to improved intensity forecasts. SIGNIFICANCE STATEMENT: Accurate knowledge of the structure of temperature, wind, moisture, and different types of hydrometeors is important for accurate predictions of hurricanes and associated hazards, and, therefore, essential for hazard preparedness and public safety. When hurricanes are far away from land, satellites provide the most, if not the only, observations. Despite this, satellite observations that are affected by clouds are mostly not yet ingested into operational forecastmodels.Using a group of 60 forecasts ofHurricane Harvey (2017), we found that the correlations between states of the numerical models and satellites infrared and microwave observations that are affected by clouds are consistent with the dynamical and thermodynamical processes within the hurricanes. Additionally, observations from different infrared or microwave spectral bands show different characteristics in terms of their most strongly correlated layer within the atmosphere and the most strongly correlated model states. Therefore, combining cloud-Affected observations from different spectral bands and ingesting them into numerical models has great potential to improve predictions of hurricanes and associated hazards.
All Science Journal Classification (ASJC) codes
- Atmospheric Science