1, : SVD 59 changes of the vertically integrated diabatic heating, sea level p ressure, and the geopotential height at 500 hpa, the variations of the

Size: px

Start display at page:

Download "1, : SVD 59 changes of the vertically integrated diabatic heating, sea level p ressure, and the geopotential height at 500 hpa, the variations of the"

丰应
5 years ago
Views:

1 33 1 Vol. 33 No Transactions of A tmospheric Sciences Feb. 2010,,. SVD [ J ]., 2010, 33 (1) : Zhang Peng2bo, Guan Zhao2yong, Cai Jia2xi. Impacts of interannual variations of Australian high on the summer temperature in China by SVD analysis[ J ]. Trans A tmos Sci, 2010, 33 (1) : SVD 1, 2, 3, 1, 2, 1, 2 ( 1., ; 2., ; 3., ) :, SVD ( ), SVD ( SVD1) 0197 SVD1 ( ), ( ) 500 hpa : ( ), ( ), ( ), P2J, ( ), ( ), ( ) ( ), ( ), ( ) : ; ; : P461 : A : (2010) Impacts of Interannual Variations of Australian High on the Summer Temperature in China by SVD Analysis ZHANG Peng2bo 1, 2, 3, GUAN Zhao2yong 1, 2, CA I J ia2xi 1, 2 (1. School of A tmospheric Sciences, NU IST, Nanjing , China; 2. Key Laboratory of Meteorological D isaster of M inistry of Education, NU IST, Nanjing , China; 3. Nanjing Meteorological Bureau, Nanjing , China) Abstract:B y using the sea level p ressure data from NCEP2NCAR reanalysis and monthly mean tempera2 ture data from 107 stations in China, the relationship s between interannual variations of the Australian high and summ er temperature in China have been studied using the SVD singular value decomposition and regression methods. The time series of coefficient of left singular vector of SVD1 was found to be highly correlated w ith the Australian high index as reported in many studies. The correlation coefficient is 0197, which shows that the SVD1 can well dep ict the variations of the Australian high. The first mode of SVD disp lays a close relationship between interannual variations of the Australian high and the summer temper2 ature in southern area of the Yangtze R iver. W hen the Australian high is stronger(weaker) than normal, summ er temperature in southern area of the Yangtze R iver is significantly lower( higher) than normal. The : ; : : (2007BAC29B02) ; ( KLME060101) : (1982 ),,,,, bobozpb@163. com; ( ),,,, guanzy@nuist. edu. cn.

2 1, : SVD 59 changes of the vertically integrated diabatic heating, sea level p ressure, and the geopotential height at 500 hpa, the variations of the p recip itation and the total cloud cover are factors related to the formation of the temperature anomalies during boreal summ er. W hen the Australian high is stronger(weaker) than normal, anomalous divergence ( convergence) in Australia causes anom alous convergence ( divergence) in the warm pool east of the Philipp ines and thus the anomalous positive ( negative) eddy source, which excites the P2J wave pattern. M eanwhile, the western Pacific subtrop ical high seem s to extend further westward ( eastward) than normal, which transports more ( less) water vapor northeastward to the southern area of the Yangtze R iver. A t the same time, the anomalous convergence ( divergence) of wind field and moisture over the southern area of the Yangtze R iver eventually facilitates more ( less) rainfall in southern China during boreal summ er and the lower( higher) temperature than normal in summ er season. Key words:australian high; interannual variability; summ er temperature in southern China 0, 20,,, :,, [ 123 ], [ 4 ],,,, [ 5 ], [ 6 ],, [ 7 ] [ 8 ] ENSO,, [ 9 ] ; [ 10 ],, [ 11 ],,,,, 20 70, [ 12 ] [ ] [ 15 ],, [ 16 ] ENSO :,, ;, [ 17 ],,,, (1) NCEP /NCAR , [ 18 ] ; (2) (NMC) ( E, N ) 107 ; (3) N ino3 CPC,

3 ; (4) ( ) 12 2, 3 5, 6 8, a [ 19 ] SVD ( 6 8 ) ( E, S), 107 ENSO, N ino3 ( y ), ENSO, ( y - I N ino3 ) 1 p s - p t pt 9T p s 9t dp = 1 p s - p t pt - gv g T + T p s p - 9T + Qg dp 9t c p : p s p t 100 hpa ; = dp d t ; = R d c p,, t, 011, EN SO 2, [ 20 ] EN SO, EN SO,, [ 21 ], 2003,,,, [ 16 ] EN SO, [ 22 ] [ 23 ] ( ), ( I A HS ) ( E, S ) (6 8 ) ( 1a ), [ 15 ], SVD SVD ( 1), SVD % ( 1a b) ,,,,,, SVD 1 [ 22 ] [ 23 ] 0197 ( 2), SVD 1 ( I SVD 1 ) 1 El N ino 107 SVD 4 Table 1 The SVD com ponents as decom posed from SL P w ith El N ino signal rem oved and the summ er tem perature anom alies in China / % / % / % / %

4 1, : SVD 61 1 SVD 1 ( a) ( b) ( a ; b 0125 ) Fig. 1 ( a) The left heterogeneous correlations of SVD 1 and ( b) the right heterogeneous correlations of SVD 1 ( The shaded area in ( a) denotes the clim atological m ean SL P larger than hpa; the values larger than 0125 in ( b) are shaded) 2 SVD 1 ( : ; : ; : ; R 1 : ; R 2 : ) Fig. 2 The series of tim e coefficient of left( solid line w ith closed circle) and right ( dotted, open circle) com po2 nents of singular vector SVD 1 and norm alized tim e series of A ustralian high index ( dashed, square) (R 1 is for the correlation coefficient betw een the coefficient series of left singular vector and norm alized tim e se2 ries of A ustralian high index w hereas R 2 is for the correlation betw een the tim e coefficient series of left sin2 gular vector and the tim e coefficient series of right singular vector) ( I A HS ) ( 2) I SVD 1 I SVD , ,

62 33 : 1959 1962 1966 1969 1970, 1976 1977 1979 1982 1988 1993 1994 1997 [ 25 ] Robert 1999 2006 ; : 1958 1963 1964 1968 1972 1974 1978 1981 1986 1990 1991 1996 2005 SVD 1 ( 3a) ( 3b),, ;,,, [ 17,

5 62 33 : , [ 25 ] Robert ; : SVD 1 ( 3a) ( 3b),, ;,,, [ 17, 24 ] 3a 3b,,,,, [ 26 ] 4,,,,,,,,,,, ;, 312 ( 5a) 5a,,, ( 5b),,,,, ( 5c), 3 SVD 1 ( t 90% ) a1 (mm /m on) ; b1 ( ) Fig. 3 Com posite differences of JJA m ean ( a) p recip itation and ( b) tem perature betw een the strong A ustralian high years and the w eak A ustralian high years ( The shaded areas denote the com posite differences significant at the 90% confidence level)

6 1, : SVD 63 4 SVD 1 ( % ; t 90% ) Fig. 4 C om posite differences of the JJA total cloud cover anom alies ( % ) betw een the strong and w eak A us2 tralian high years for period ( The sha2 ded areas denote the com posite differences signifi2 cant at the 90% confidence level),, G ill [ 28 ], 850 hpa,, SPC Z,,,,, 6b,,,, ( ), ( 6a),,, 500 hpa ( 6c),,, [ 29 ],,,,,,,,, EN SO SVD,,, SVD ,, SVD 1, SVD : (1) ( 6a),,, ;,, (2),,, 850 hpa ( 6a),, 500 hpa, ( 1a, 6a),,,, ( 5a), (3) : SPCZ ( 4),, ( 5b) G ill, P2J, [ 27 ], SPC Z,, P2J,,,,,,

64 33 5 SVD 1 ( t 90% ) a1 ; b1 ; c1 ( : d - 1 ; 100 hpa) Fig.

7 SVD 1 ( t 90% ) a1 ; b1 ; c1 ( : d - 1 ; 100 hpa) Fig. 5 Com posite differences of the JJA m ean ( a) vertically integrated tem perature advection, ( b) vertically inte2 grated diabatic heating, and ( c) vertically integrated horizontal tem perature advection betw een the strong and w eak A ustralian high years for period ( The shaded areas denote the com posite differences significant at the 90% confidence level) : N C EP /N CA R ; (NN C ) ; N ino3 NOAA 2C IR ES C lim ate Pred iction C en ter ( h t2 tp: / /www. cpc. ncep. noaa. gov) ; NOAA 2C IRES C li2 m ate D iagnostics Center ( http: / /www. cdc. noaa. gov) N C EP /N CA R ; G rad s ;! : [ 1 ],,, [ J ]., 2004, 19 (3) : [ 2 ],, [ J ]., 2001, 17 (4) : [ 3 ],, [ J ]., 2000, 21 (1) :

8 1, : SVD 65 6 SVD 1 a1 ( hpa; 011 ) 850 hpa (m / s) ; b1 ( kg / (m s) ) (10-4 kg / ( cm 2 s) ; ) ; c. 500 hpa ( gpm; 011 ) Fig. 6 Com posite differences of ( a) the JJA m ean SL P ( The shaded areas denote the com posite differences significant at the 90% confidence level) and w ind anom alies at 850 hpa, ( b) vertically integrated m oisture fluxes and their divergence ( The shaded areas denote the regions w here the divergence is less than or equals to 0), and ( c) 500 hpa geopotential heights ( The shaded areas denote the com posite differences significant at the 90% confidence level) betw een the strong and w eak A ustralian high years for period [ 4 ],,,. [ J ]., 2001, 24 (1) : [ 5 ],,. 500 hpa [ J ]., 1992, 12 ( 4 ) : [ 6 ],,. [ J ]., 2001, 24 (4) : [ 7 ],. [ J ]., 2005, 25 (1) : [ 8 ],,,. [ J ]., 2006, 29 (4) : [ 9 ],,. ENSO [ J ]., 1990, 6 (1) : [ 10 ],. [ J ]., 1982, 6 (1) : [ 11 ]. [ J ].

9 66 33, 2005, 10 (3) : [ 12 ] W ang B. Interdecadal changes in E l N ino onset in the last four decades[ J ]. J C lim ate, 1995, 8 (2) : [ 13 ] Xue F. Interannual to interdecadal variation of East A sian Sum 2 m erm onsoon and its association w ith the global atm ospheric cir2 culation and sea surface temperature [ J ]. A dv A tm os Sci, 2001, 18 (4) : [ 14 ] W ang H J. The w eakening of the A sian m onsoon circulation af2 ter the end of 1970s [ J ]. A dv A tm os Sci, 2001, 18 ( 3 ) : [ 15 ] Trenberth K E. Recent observed interdecadal clim ate changes in the N orthern Hem isphere [ J ]. B ull Am er M eteor Sci, 1990, 71 (7) : [ 16 ],,,. SVD [ J ].,. [ 17 ]. [ J ]., 1991, 15 (3) : [ 18 ] Kalnay E, Kanam itsu M. Kistler R, et a. l. The NCEP /NCAR 402year reanalysis p roject[ J ]. B ull Am er M eteor Soc, 1996, 77 (3) : [ 19 ],,. ENSO [ J ]., 1997, 21 (6) : [ 20 ],. ENSO [ J ]., 1995, 19 (2) : [ 21 ], [ J ]., 2005, 10 (1) : [ 22 ],,. [ J ]., 2003, 48 ( 3 ) : [ 23 ],. [ J ]., 1995, 15 ( 2 ) : [ 24 ],,,. 80 [ J ]., 2004, 62 (5) : [ 25 ],,. [M ]. :, 2000: [ 26 ] Robert D C, Zhang M ing2hua, W ielicki B A, et al. The influence of the1998 El N ino upon cloud2radiative forcing over the Pacif2 ic2w arm pool[ J ]. J C lim ate, 2001, 14 (9) : [ 27 ] N itta T. Convective activities in the trop ical w estern Pacific and their impact on the N orthern Hem isphere summ er circulation [ J ]. J M eteor Soc Japan, 1987, 65: [ 28 ] G ill A E. Som e simp le solution for heat induced trop ical circula2 tions[ J ]. Q uart J Roy M erteor Soc, 1980, 106 (5) : [ 29 ],. [ J ]., 1998, 22 (5) : ( : )

10 30 IOD [ ], IOD, [ 14 ] El N ino 8,, [ 15 ],, ENSO IOD, IOD N ino GISST2. 3b [ 16 ] CMAP [ 17 ] NCEP / NCAR [

10 30 IOD [ ], IOD, [ 14 ] El N ino 8,, [ 15 ],, ENSO IOD, IOD N ino GISST2. 3b [ 16 ] CMAP [ 17 ] NCEP / NCAR [ 30 1 Vol. 30 No. 1 2007 2 Journal of Nanjing Institute of Meteorology Feb. 2007, (, 210044) : 1979 1999 GISST CMAP, ( IOD ) ENSO, IOD N ino3, 3, : 6 8,, ; 5 60 %,( ) IOD, SS2 TA ( ), ( ), ( ) : ; ; ; :