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 581

a b c

Fig. 24.11 TC Longwang (2005). VIE ( shaded ;10Jkg 1 ) and wind ( vector ;ms 1 ) component
of ( a ) FSV, ( b ) CNOP, ( c ) CSV over a 24-h optimization time interval initialized at 0000 UTC
30 Sep 2005. The big boxes ( dashed ) indicate the verification areas; the box ( solid ) indicates the
sensitive area determined by the ( a ) FSV, ( b ) CNOP, ( c ) CSV; the symbol

˚ indicates the initial

cyclone center (From Chen and Mu 2012 )

a b

and wind (vector; m s 1 ) component of
( a ) CNOP, ( b ) CSV over a 24-h optimization time interval initialized at 0000 UTC 10 Sep 2008.
The big boxes ( dashed ) indicate the verification areas; the box ( solid ) indicates the sensitive area
determined by the ( a ) CNOP, ( b ) CSV; the symbol

TC Sinlaku (2008). VIE ( shaded ,10Jkg 1 / Fig. 24.12

˚ indicates the initial cyclone center (From

Chen and Mu 2012 )

For simplicity, the CNOP-, CSV-, FSV-, sensitive areas and randomly selected
areas are marked by CNOP Sen, CSV Sen, FSV Sen, and Ran Area, respectively.
The random initial errors (Ran Err) are added to the four areas. Notably, the random
initial errors have the same size measured by the dry energy norm. The four
experiments noted by RA-CN, RA-CS, RA-FS, and RA-RA, respectively represente
that the initial errors are introduced into the CNOP Sen, CSV Sen, FSV Sen, and
Ran Area.
First, the distributions of CNOP Sen, CSV Sen, and FSV Sen are checked. Here,
the maximum of the vertically integrated energy (VIE) of CNOP, CSV, and FSV
are defined as the centers of CNOP Sen, CSV Sen, and FSV Sen (indicated by
the smaller squares in Figs. 24.11 and 24.12 , differing from the above studies).
It is seen that the considerable difference exists between the locations of FSV Sen
and CSV Sen (Fig. 24.11 a, c). Although the VIE maxima of CNOP and CSV are