1 PREDICTION OF FORT WORTH TORNADIC THUNDERSTORMS USING 3DVAR AND CLOUD ANALYSIS WITH WSR-88D LEVEL-II DATA Ming Hu 1,2 , Ming Xue *1,2 , Keith Brewster 1 and Jidong Gao 1 1 Center for Analysis and Prediction of Storms 2 School of Meteorology University of Oklahoma, Norman, OK 73019 1. INTRODUCTION * The development of high-resolution nonhy- drostatic models and the rapid increase of com- puter power are making the explicit prediction of thunderstorms a reality (Droegemeier 1990; Lilly 1990; Droegemeier 1997; Xue et al. 2003, X03 hereafter). Data assimilation plays an important role in providing an accurate initial condition for the model forecast. The operational US WSR- 88D Doppler radar network (Crum and Alberty 1993) is a key source of data for initializing storm- scale numerical weather prediction (NWP) mod- els as it the only operational platform capable of providing observations of spatial and temporal resolutions sufficient for resolving convective storms. The analysis of radar data to arrive at a com- plete set of initial conditions for a NWP model is challenging, because radars only observe a very limited set of parameters, the most important be- ing the radial velocity and reflectivity. Their spatial coverage is often incomplete. To determine at- mospheric state variables that are not directly observed, certain retrieval or assimilation tech- niques have to be used. Four-dimensional variational (4DVAR) data assimilation, which obtains a full set of model ini- tial conditions that provides the best fit between the model solution and radar observations within a time (assimilation) window, is considered ideal for this purpose. Some encouraging 4DVAR re- sults with both simulated and real radar data have been obtained by, for example, Sun et al. (1991; 1997; 1998). On the other hand, the complexity of developing and maintaining the adjoint code needed by a 4DVAR system and the high compu- tational cost of 4DVAR technique for high- resolution applications are limiting its use in re- * Corresponding Author Address: Dr. Ming Xue, School of Meteorology, University of Oklahoma, 100 East Boyd, Norman, OK 73019. E-mail: mxue@ou.edu. search and operation. Another relatively new technique is the ensemble Kalman filter (EnKF) method, which has been shown recently to pro- duce single-Doppler radar analyses of thunder- storms that are of similar quality as the 4DVAR analysis (Snyder and Zhang 2003; Tong and Xue 2004; Zhang et al. 2004). While also expensive because of the need for running an analysis and forecast ensemble of significant sizes, EnKF method enjoys the simplicity in implementation and is much more flexible. Other simpler, yet faster, methods exist that attempt to retrieve unobserved variables from the radar data. The retrieved state variables can then be analyzed into the model initial conditions. The wind retrieval methods include the so-called sim- ple adjoint method (Qiu and Xu 1992; Qiu and Xu 1994; Xu et al. 1994; Gao et al. 2001) and two- scalar method of Shapiro et al. (1995), among others. The former employs a simple prognostic equation and its adjoint to determine the advec- tive winds that produce the best fit between the predicted and observed radial velocity and/or re- flectivity. The latter method is based on the con- servation of two scale quantities and is demon- strated by Weygandt et al. (2002a). Additionally, the retrieved three-dimensional wind fields at more than one time level can be used to retrieve additionally thermodynamic fields (Gal-Chen 1978). The retrieved fields can then be combined via an analysis procedure, as is done in Wey- gandt et al. (2002b). Such multi-step procedures have the advantages of being able to make use of multiple radar volume scans in an inexpensive way, but the involvement of multiple steps and the use of retrieved instead of direct observations make the optimality of analysis difficult to impose. Another alternative is to analyze the radial ve- locity data directly via a three-dimensional varia- tional (3DVAR) analysis procedure. Certain dy- namic or equation constraints can be built into the 3DVAR cost function with relative ease. Such a system has been developed within the ARPS model (Xue et al. 1995; 2000; 2001) framework and documented in Gao et al. (2002; 2004). It is used in this study to analyze radial velocity and J1.2 11th Conference on Aviation, Range, and Aerospace 22nd Conference on Severe Local Storms