Extended Abstract (412K)5D.1
Observing-systems simulation experiments for tropical cyclone initialization based on four-dimensional variational data assimilation
Chun-Chieh Wu, National Taiwan Univ., Taipei, Taiwan; and K. H. Chou, Y. H. Kuo, and W. P. Huang
As the conventional observations usually have much less degree of freedom than the model does, the 4-dimensional variational (4D- VAR) data assimilation has become one of the most advanced processes to combine the observational data with the model in a way so that the initial conditions are consistent to the model dynamics and physics. Based on 4D-VAR, a bogus data assimilation (BDA) method has been developed recently by Zou and Xiao (2000) to improve the initial condition for tropical cyclone simulation. Given a specified sea level pressure (SLP) distribution, the BDA process can lead to a better initial typhoon structure. Zou and Xiao (2000) showed that with the better initialization procedure, the simulation of the track and intensity of Hurricane Felix (1995) are also improved. Xiao et al. (2000) expanded their work by assimilating the wind field data into the model. Using the observed data, such as the minimum SLP, and/or the maximum wind speed and radius of maximum wind speed, and based on the presumed idealized distribution of the SLP and/or 3-dimensional wind field structure, a dynamically consistent initial structure is produced. By comparing the simulation with different data used for BDA, their result indicates that the assimilation of only the pressure field is more effective than the assimilation of only the wind field. However, using a similar approach, Pu and Braun (2001) showed that the assimilation of wind field would be more useful than the assimilation of the pressure field, while assimilating both the wind and pressure fields would provide the best results. These studies present a useful new direction for tropical cyclone initialization using BDA. Nevertheless, important questions remain to be answered: (1) what are the most critical parameters for BDA; and (2) what can be done to improve the existing BDA procedures. The primary motivation of this study is to clarify and investigate the related questions of the BDA method mentioned above. In this study, we performed a set of Observing-Systems Simulation Experiments (OSSE’s) to address the above questions on BDA. Highlights of our findings are: 1. The assimilation of wind field is more effective than the assimilation of pressure field. 2. The assimilation of an axisymmetric vortex tends to misrepresent the actual storm and leads to a large track prediction error. This suggests that the inclusion of the asymmetric component of the tropical cyclone vortexis important for vortex initialization. 3. Direct replacement of BIC data to FEC may result a vertically incoherent iintial vortex, while BDA is able to recover a more balanced initial vortex structure. Ÿ Assimilation of wind at individual level(s) only slightly improves the initial vortex structure and its track simulation. This suggests that for a better representation of the tropical cyclone and a better storm simulation, comprehensive and more vertical levels of observed data are needed. 4. Both the improvement on the initial vortex structure and the environmental wind and mass information affects the track evolution, suggesting the need for near storm, as well as storm environment, observations in order to improve the track simulation.
Session 5D, Tropical Cyclone Prediction III (Parallel with Sessions 5A, 5B, & 5C)
Tuesday, 30 April 2002, 11:00 AM-12:30 PM
Next paper