The Advances in Targeted Observations for Tropical Cyclone Prediction Based on Conditional Nonlinear Optimal Perturbation (CNOP) Method - Data Assimilation for Atmospheric, Oceanic and Hydrologic Applications - page 591

Fig. 24.18 DOTSTAR observations at 1200 UTC 17 May 2004 ( square points ) and best track of
Nida from 1200 UTC 17 to 0000 UTC 20 May 2004 ( solid line ). The numbers from 1 to 15 indicate
the sequence of dropsonde observations (From Chen 2011 )

b

a

Fig. 24.19 VIE ( shaded ;Jkg 1 ) and wind ( vector ;ms 1 ) component at the level 0.7 of ( a ) CNOP
and ( b ) FSV. The box defined by dashed lines indicates the verification area. Pentangle points
indicate the dropsonde observations used in ( a ) CNOPDROP and ( b ) FSVDROP. The symbol '˚'
indicates the position of Nida's center at 1200 UTC 17 May 2004 (From Chen 2011 )

dropsondes in the CNOP sensitive area improved the track forecasts significantly,
the dropsondes in the FSV sensitive areas also increased the accuracies of the
24-h and 36-h track forecasts, but the impact is comparatively small. However,
the dropsondes in the randomly selected region have negative effects on the track
forecasts, especially on the 24-h track forecast.
These results indicate that the CNOP method would be useful in decision making
about dropsonde deployments.