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 586

a b

34N
32N
30N
28N
26N
24N
22N
20N
18N
16N

34N
32N
30N
28N
26N
24N
22N
20N
18N
16N

111E 114E 117E 120E 123E 126E 129E 132E 135E 138E 111E 114E 117E 120E 123E 126E 129E 132E 135E 138E

c

34N
32N
30N
28N
26N
24N
22N
20N
18N
16N

111E 114E 117E 120E 123E 126E 129E 132E 135E 138E

Fig. 24.15 TC Nock-Ten (2004). ( a ) the routine observations, ( b ) the CNOP sensitive areas and
( c ) the random selected areas denoted by number 1, 2, 3, 4 respectively (From Qin 2010)

Table 24.4 The RMSE of both routine and additional observation relative to only routine
observation for Nock-Ten. The first column represents the kind of additional data that is utilized.
The variables from the second to the eighth column stands for the vertical mean zonal wind,
meridional wind, temperature, surface pressure, specific humidity, vertical wind, and vertical
integrated total perturbation energy, respectively. Negative values represent that the RMSEs
produced by the corresponding additional data were reduced
U (%) V (%) T (%) PP (%) QV (%) W (%) TPE (%)

CNOP -2.58 -9.86 -4.01 0.62 0.06 -3.60 -12.74

Random1 -2.16 -0.96 -1.63 -0.10 -2.41 -5.23 -0.90

Random2 -0.28 -1.57 0.87 -4.78 0.00 1.18 -1.56

Random3 1.60 3.93 -0.54 -5.27 0.56 -0.59 5.74

Random4 -1.03 1.85 2.62 -4.02 0.35 3.43 2.86

The difference between the typhoon centre positions of the control run and of the
nature run is defined as the error in the typhoon track forecast without dropsonde
data. After identifying the sensitive regions for the optimization period (24-48 h),
simulated dropsondes are deployed at 24 h over these regions to obtain observational
data, which represent the sum of the forecasts of the nature run at that time and