Targeted observation and its impact in DOTSTAR and T-PARC

Well coordinated with the T-PARC program (Elsberry and Harr 2008) in 2008, targeted observation by DOTSTAR (Wu et al. 2007a) was successfully conducted for NW Pacific tropical cyclones near Taiwan with 10 flight missions, 52 flight hours, and 172 dropwindsondes deployed for 7 typhoons. Seven different targeting guidance products (Wu et al. 2009) were use to design the targeted strategy, including JMA AV (Yamaguchi et al. 2009), NOGAPS SV (Peng and Reynolds 2006), ECMWF SV (Buizza et al 2007), MM5 SV (Kim and Jung 2009), ADSSV (Wu et al. 2007b), ETKF (Majumdar et al. 2002), and NCEP ensemble variance (Aberson 2003). In this paper, the comparison of these targeted techniques used to design the flight paths and drop points during DOTSTAR and T-PARC will be shown. Preliminary results on the impact of the dropwindsonde data from the targeted observation of DOTSTAR in 2008 will also be evaluated based on the NCEP GFS denial experiments. A case in point, Typhoon Sinlaku (2008), which is well observed with four aircrafts during T-PARC, is selected to demonstrate the huge positive improvement in track prediction from the dropwindsonde data, in particular from the DOTSTAR data based on different targeting methods. Dynamic analysis on how the drowindsonde data lead to the track forecast improvement is performed and will be presented in the conference.