Doppler Radar Observations and Ensemble-Based Data Assimilation for Cloud-Resolving Hurricane Prediction

点击翻译按钮获取中文摘要

This award is funded under the American Recovery and Reinvestment Act of 2009 (Public Law 111-5).Despite significant progress in short-range hurricane track forecasts, over the past decade there has been virtually no improvement in predicting hurricane intensity and amount of precipitation, and very limited skill exists in the numerical prediction of tropical cyclone formation, rapid intensification or decay. It remains unclear whether the lack of numerical forecast skill derives from imperfect initial conditions, inadequate data assimilation techniques, imperfect forecast models, or from intrinsic predictability limits due to chaotic dynamics. This research will use ensemble-based data assimilation (EnDA) to explore the initialization and prediction of tropical cyclones as well as the fundamental limits to the predictability of hurricane structure and intensity. The research will utilize special data sets collected on three 2005 hurricanes (Katrina, Ophelia and Rita) during the Rainband and Intensity Experiment (RAINEX). The RAINEX was a joint NSF/NOAA field project that employed multiple aircraft with Doppler radars and dropsonde capabilities to investigate the hurricane intensity changes that are associated with eyewall-rainband interactions. Other cases to be examined include Emily (2005), Wilma (2005), and Humberto (2007). All of these selected cases have extensive and unique Doppler radar observations that were collected over the lifecycles of the hurricanes. Major objectives of the research include: (1) assess the impacts on predictability of assimilating ground-based and airborne radar observations in addition to regular and experimental in-situ and remotely sensed observations; (2) explore the role of the large-scale environment and internal dynamics, including moist convection, on the predictability of hurricane structure and intensity change; and (3) evaluate the impact and effectiveness of different observing platforms and targeting strategies on hurricane prediction through observing system numerical experiments with both simulated and actual observations. Intellectual merit: The combination of the EnDA technique with field observations constitutes a potentially significant advance in data assimilation and understanding of dynamics, intensity, structure, prediction, and fundamental limits of predictability of tropical cyclones. The work will also may provide guidance for future observing network design and observing strategy. Broader impacts: The research will provide educational benefits through the extensive participation of a doctoral graduate student and a postdoctoral scientist. Research findings will be used in advanced educational courses and undergraduate research projects. Knowledge gained about the predictability of hurricanes potentially will benefit the operational communities and the general public. The EnDA techniques and computer codes will be made freely available online with ample documentation to benefit the larger academic community.