Microsoft Word 张培军_news_.doc

Size: px

Start display at page:

Download "Microsoft Word 张培军_news_.doc"

缆伉随
7 years ago
Views:

1 模式参数的不确定性对日本南部黑潮大弯曲路径预报的影响 1, 张培军 1, 王强 (1., 66071;., ) 摘要 : 基于 1.5 层浅水方程模式, 利用条件非线性最优参数扰动 (CNOP-P) 方法, 研究模式参数的不确定性对黑潮大弯曲路径预报的影响研究表明, 单个模式参数误差如侧向摩擦系数误差界面摩擦系数误差以及在不同季节具有不同约束的风应力大小误差, 对黑潮大弯曲路径预报的影响较小, 并且对背景流场的选取具有一定的敏感性 ; 所有模式参数误差同时存在时对黑潮大弯曲路径预报具有一定的影响, 并且预报结果在 9 个月左右不能被接受因此, 要提高黑潮大弯曲路径的预报技巧, 模式中的参数需要给出更好的估计关键词 : 黑潮大弯曲路径预报 ; 条件非线性最优参数扰动 ; 模式参数误差中图分类号 : P731.3 文献标识码 : A 文章编号 : (015) doi: /hykx ,,,, [1],,,, [-4] [5] [6],, 3, Ishikawa [7] 1.5-,,, Fujii [8] (SV), Wang [9] (CNOP-I),,,,, 106 / 015 / 39 / 5,,,, Wang [9] (CNOP-P) 3,,, [10],,, 1 浅水方程模式和条件非线性最优参数扰动 (CNOP-P) 方法 1.1 浅水方程模式 1.5 Schmeits [11] : ; : : ( ) : (1984-),,,, ; :,,,,, : , wangqiang@ qdio.ac.cn

2 Pierini [1], 1.5,, (1): ædu ö F h uv tan q - ç - vsin dt - q= + çè ø cosq j æ u sinq vö a j E u - - mu ç çè cos q cos q j + - (1a) ø h ædv ö h u tan q ç + usin F dt + q=- + çè ø q æ v sinq uö a q E v - + mv ç çè cos q cos q jø + - (1b) h h 1 é ( hu) ( hvcos q) ù + + = 0 (1c) t cosq ê j q ú ë û, j q u v h D dt, : U r, gh AH 0 F, E,, 0 U r HU 0 RI, r 0 = m, = s 1, U =0.1 m/s H =500 m g =0.044 m/s, =103.5 kg/m 3, 0 =0.1 Pa, A H =450 m /s, R I = s 1 (15 o ~55 o N, 1 o E~158 o W) Crank-Nicholson, 10 d, 0. o 0. o,, 00 m Hellerman [10] 100 a, 90 a 1 50 m, 136 o ~140 o E [14],, 1, 90 a 17,, Wang [9] 1.5, 1975~1979 [13] [14-15], 1.5,,, Fig.1 1 The time series of the Kuroshio path index simulated by the model Marine Sciences / Vol. 39, No. 5 /

3 , 1. 条件非线性最优参数扰动方法 [16] T 0 X 0 ( u0, v0, h0), (1) P MT ( p ), T ( )( 0) (,, ) MT p X ut vt ht (1) T p *, : J ( p) C p ( p1, p,, pm ) p1 1, p,, p : * J( p ) max J( p ) () pc, J ( p) MT( p p)( X0) MT( p)( X0), p ( p1, p,, pm ) i, i 1,, m, xp( T) MT( p p)( X0) MT( p)( X 0 ) ( utp, vtp, htp), T 1 J ( p) xp ( T) ( ht htp )[( utp ) ( vtp ) ]dxdy KE. (3) ( o ~35 o N, 130 o ~140 o E),,, J ( p) C, J1( p) J( p) C (SPG) [17] 1.3 模式参数误差的约束设置 1.5 3, (A H )(R I ) ( 0 ),,,,, m, [18],,, Hellerman and Rosenstein (HR) [10] Comprehensive Ocean-Atmosphere Data Set(COADS05) [19], COADS05, 0.5, 1945~1989, HR 1870~ 1976 (1945~ 1976 ), COADS ~ 1977,,,, (root-mean-square error, RMSE) 1 1, 表 1 模式参数误差的约束设置 Tab.1 The constraint condition of the model parameter error i, i 1,,3,4 A H 45 m /s (A H 10%) R I s 1 (R I 10%) Pa(), Pa() Pa(), Pa() 不同背景场下的 CNOP-P 3,, 1 3, 1a, 3 35~3670 ~71857~ / 015 / 39 / 5

4 , 3,,,,, 7, : A H ( AH), R I ( RI), (3~5) 1 ( MAM), (6~8) ( JJA), (9~11 ) 3 ( SON), (1~) 4 ( DJF) ( ALL), i 0, i 1,,3,4, 34 SPG3 CNOP-P 3 4, CNOP-P,, 1, 表事件 1 的 7 种情况的 CNOP-P Tab. The seven CNOP-Ps for the Event 1 CNOP-P ( m 5 /s) AH - 45 m /s 37.1 RI s.66 MAM Pa 18.6 JJA Pa 18.1 SON Pa 1.47 DJF Pa 0.61 ALL -8-1 (45m /s, s, Pa, Pa, Pa, Pa) 146 表 3 事件的 7 种情况的 CNOP-P Tab.3 The seven CNOP-Ps for the Event CNOP-P ( m 5 /s) AH - 45 m /s RI s.53 MAM Pa 5. JJA Pa 17.9 SON Pa 1.7 DJF Pa 0.63 ALL -8-1 (45m /s, s, Pa, Pa, Pa, Pa) 130 表 4 事件 3 的 7 种情况的 CNOP-P Tab.4 The seven CNOP-Ps for the Event 3 CNOP-P ( m 5 /s) AH -45 m /s RI s.04 MAM Pa.19 JJA Pa 5.3 SON Pa 7.0 DJF Pa 6.17 ALL -8-1 (45m /s, s, Pa, Pa, Pa, Pa) 108, 3, 1 AH DJF CNOP-P, CNOP-P CNOP-P,, 3,,,, CNOP-P,,,,, 3,, 3 AH CNOP-P,, 3 CNOP-P,, CNOP-P CNOP-P, 1, 6 CNOP-P 8 1 ( 45m /s, s, Pa, Pa, Pa, Pa), CNOP-P (45m /s, s 1, pa, Marine Sciences / Vol. 39, No. 5 /

5 Pa, Pa, Pa) ALL CNOP-P, 1, ( m 5 /s) ALL CNOP-P ( m 5 /s), 3 模式参数不确定性对日本南部黑潮路径变异预报的影响 CNOP-P, CNOP-P,,, CNOP-P, 1 a, 3 7, MAM JJA SON DJF RI CNOP-P, CNOP-P, CNOP-P,, CNOP-P, 3 AH CNOP-P, RI, 1 CNOP-P, ; 3 CNOP-P,, CNOP-P( ALL),,,, CNOP-P,, 7, RMSE, E RMS, : M N 1 ERMS() t ( htf (, i j) htb (, i j)) MN (4) i j 1, htb h tf t, (5 ~35 N, 13 ~140 E), MN, E RMS,,, (5 ~35 N, 13 ~140 E) 90 a (root-mean-square variation, RMSV), V RMS Komori [0] Miyazawa [6] E RMS V RMS, 3, 3, CNOP-P E RMS, V RMS, CNOP- P E RMS 9 V RMS, CNOP-P E RMS CNOP-P,, 4 结论与讨论 1.5,,, 3,,,, CNOP-P,,,, 9 V RMS,,, :,, 110 / 015 / 39 / 5

6 CNOP-P Fig. The background field and the Kuroshio axes caused by CNOP-Ps Fig.3 3 CNOP-P (E RMS ) Time series of RMSE of the upper-layer thickness between background field and forecast field caused by CNOP-Ps Marine Sciences / Vol. 39, No. 5 /

7 ,,,,,,,, : [1] Kawabe M. Variations of current path, velocity, and volume transport of the Kuroshio in relation with the large meander [J]. J Phys Oceanogr, 1995, 5: [] Hurlburt H E, Smedstat O M, RhodesR C, et al. A feasibility demonstration of ocean model eddy-resolving nowcast/forecast skill using satellite altimeter data[r]. Stennis: Naval Research Laboratory, Stennis Space Center, Hancock, 000. [3] Kagimoto T, Miyazawa Y, Guo X, et al. High resolution Kuroshio forecast system -Description and its applications[c]//ohfuchi W, Hamilton K. High Resolution Numerical Modeling of the Atmosphere and Ocean. New York: Springer, 008: [4] Kamachi M, Kuragano T, Sugimoto S, et al. Short-range prediction experiments with operational data assimilation system for the Kuroshio south of Japan[J]. J Oceanogr, 004, 60: [5] Usui N, Tsujino H, Fujii Y, et al. Short-range prediction experiments of the Kuroshio path variabilities south of Japan[J]. Ocean Dynamics, 006, 56(5): [6] Miyazawa Y, Yamane S, Guo X, et al. Ensemble forecast of the Kuroshio meandering[j]. J Geophys Res, 005, 110: C1006. [7] Ishikawa Y, Awaji T, Komori N, et al. Application of sensitivity analysis using an adjoint model for short-range forecasts of the Kuroshio path south of Japan [J]. J Oceanogr, 004, 60: [8] Fujii Y, Tsujino H, Usui N, et al. Application of singular vector analysis to the Kuroshio large meander [J]. J Geophys Res, 008, 113: C0706. [9] Wang Q, Mu M, Dijkstra H A. Application of conditional nonlinear optimal perturbation method to the predictability study of the Kuroshio large meander [J]. Adv Atmos Sci, 01, 9: [10] Hellerman S, Rosenstein M. Normal monthly wind stress over the world ocean with error estimates [J]. J Phys Oceanogr, 1983, 13: [11] Schmeits M J, Dijkstra H A. Bimodal Behavior of the Kuroshio and the Gulf Stream [J]. J Phys Oceanogr, 001, 31: [1] Pierini S A. Kuroshio Extension system model study: decadal chaoticself-sustained oscillations [J]. J Phys Oceanogr, 006, 36: [13] Nishida H. Description of the Kuroshio meander in [j]. Report of Hydrographic Research, 198, 17: [14] Qiu B, Miao W. Kuroshio path variations south of Japan: bimodality as a self-sustained internal oscillation [J]. J Phys Oceanogr, 000, 30: [15] Miyazawa Y, Guo X, Yamagata T. Roles of mesoscale eddies in the Kuroshio paths [J]. J Phys Oceanogr, 004, 34: 03-. [16] Mu M, Duan W S, Wang Q, et al. An extension of conditional nonlinear optimal perturbation approach and its applications [J]. Nonlinear Processes in Geophysics, 010, 17: [17] Birgin E G, Martinez. J M, Raydan M. Nonmonoton espectral projected gradient methods on convex sets [J]. SIAM J Optimiz, 000, 10: [18] Josey S A, Kent E C and Taylor P K. Wind stress forcing of the ocean in the SOC climatology: comparisons with the NCEP NCAR, ECMWF, UWM/COADS, and Hellerman and Rosenstein datasets [J]. J Phys Oceanog, 00, 3: [19] Da Silva A M, Young C C, Levitus S. Atlas of surface marine data 1994: Algorithms and procedures[j]. US Department of Commerce, NOAA, NESDIS, 1994, 1: 83. [0] Komori N, Awaji T, Ishikawa Y, et al. Short-range forecast experiments of the Kuroshio path variabilities south of Japan using TOPEX/Poseidon altimetric data[j]. J Geophys Res, 003, 108: C1. 11 / 015 / 39 / 5

8 The impact of the model parameter uncertainties on the predictability of the Kuroshio large meander path south of Japan ZHANG Pei-jun 1,, WANG Qiang 1 (1. Key Laboratory of Ocean Circulation and Waves, Institute of Oceanology, Chinese Academy of Sciences, Qingdao 66071, China;. University of Chinese Academy of Sciences, Beijing , China) Received: Mar., 6, 014 Key words: prediction of the Kuroshio large meander path; conditional nonlinear optimal parameter perturbation; model parameter error Abstract: Based on a 1.5-layer shallow-water model, the impact of model parameter errors on the prediction of the Kuroshio large meander (LM) path in south of Japan was investigated by the conditional nonlinear optimal parameter perturbation (CNOP-P) method. The results showed that the impacts of the single model parameter errors such as the lateral friction coefficient, interfacial friction coefficient and the amplitude of wind stress with different constraints in different seasons, on the prediction of the Kuroshio LM path were relatively small. In addition, they were sensitive to the background field, which was the Kuroshio transition state from non-large meander (NLM) path to LM path. However, the optimal error mode of all the model parameter errors existing simultaneously had a large impact on the prediction of the Kuroshio LM path and the results of prediction could not be accepted. Therefore, in order to enhance the forecast skill of the Kuroshio LM path in this model, the model parameters should give the best possible estimates. ( 本文编辑 : 李晓燕 ) Marine Sciences / Vol. 39, No. 5 /

Microsoft Word - 1--齐继峰_new_.doc

Microsoft Word - 1--齐继峰_new_.doc 45 6 Vol.45, No.6 2014 11 OCEANOLOGIA ET LIMNOLOGIA SINICA Nov., 2014 * 1, 2, 3 1, 2 1, 2 1, 2 (1. 266071; 2. 266071; 3. 100049) 1956 2005 PN, NCEP,, (PDO),, 19 33Sv, 24.30Sv(1Sv=10 6 m 3 /s);, 1956 2005