大气科学  2016, Vol. 40 Issue (6): 1273-1283 PDF

The Relationship Between Prior-winter SST around Austria and Summer Rainfall in the Yangtze River Valley of China
DONG Zhulei, REN Baohua, ZHENG Jianqiu, LU Guoyang, XU Di
University of Science and Technology of China, Hefei 230026
Abstract: The HadISST (Hadley Centre Global Sea Ice and Sea Surface Temperature) dataset, the NCEP/NCAR (National Centers for Environment Prediction/National Center for Atmospheric Research) monthly reanalysis data, and monthly precipitation data collected at 160 stations in mainland China from 1980 to 2012 are used to study the relationship between the summer rainfall over Yangtze River valley in China and the prior-winter Sea Surface Temperature Anomalies (SSTA) around Austria.Results show that the leading mode of the monthly SSTA around Austria revealed by the Empirical Orthogonal Function (EOF) analysis exhibits the characteristics of consistent SST variation trend over the entire region, and the SSTA around Austria has a seasonal continuity, which can persist from the prior-winter through the following summer.Analysis of the Niño 3.4 index with the ENSO linear signal removed in the SST field shows that the consistent mode may be a local phenomenon that is independent of ENSO events.Based on the consistent mode of prior-winter SSTA around Austria, the Consistent Mode Index (CMI) is defined to characterize the influence of SSTA around Austria on the summer rainfall in the Yangtze River valley.The CMI has certain implications for short-term forecast of summer precipitation in the Yangtze River valley.The prior-winter SSTA around Austria may affect East Asian summer atmospheric circulations in two ways.First, the prior-winter SSTA signal around Australia can persist through the following summer and trigger a teleconnection wave train between the Northern and Southern Hemispheres, leading to an abnormal West Pacific Subtropical High (WPSH) that subsequently affects the summer precipitation in the Yangtze River valley.Second, the SSTA around Australia can result in abnormal convections in the tropics in both hemispheres, especially abnormal convection activities over the Philippine Sea, and eventually lead to anomalous changes in summer rainfall in the Yangtze River valley.
Key words: SST around Austria      Yangtze River valley      Summer rainfall      CMI (consistent mode index)      Tropical convection
1 引言

2 资料和方法

 $\mathit{\boldsymbol{Q}} = \frac{1}{g}\int_{300}^{{p_{\rm{s}}}} {q\mathit{\boldsymbol{V}}} {\rm{d}}p,$ (1)

 $\left\{ \begin{array}{l} {Q_u} = \frac{1}{g}\int_{300}^{{p_{\rm{s}}}} {qu} {\rm{d}}p, \\ {Q_v} = \frac{1}{g}\int_{300}^{{p_{\rm{s}}}} {qv} {\rm{d}}p, \end{array} \right.$ (2)

 $\xi = \xi * - \frac{{Z \times {\mathop{\rm cov}} (\xi *, Z)}}{{{\mathop{\rm var}} (Z)}},$ (3)

3 澳大利亚周边海温主模态

 图 1 1979年1月至2012年12月澳大利亚周边逐月海表温度距平(SSTA)的EOF分析主模态(EOF1，左列)以及对应的主分量(PC1，右列)：(a)、(b)原始场；(c)、(d)剔除ENSO线性信号后 Figure 1 The first EOF mode (EOF1, left column) of SSTA around Austria and corresponding principal component (PC1, right column) for Jan 1979 to Dec 2012: (a), (b) The original fields; (c), (d) the linear signal of ENSO is removed

 图 2 去除ENSO线性信号(a)前、(b)后，冬季时间序列的PC1与中国160站夏季降水回归分布(单位：mm month-1)。虚线方框区域表示长江流域；打点区域超过90%信度检验 Figure 2 The regression (units: mm month-1) between the PC1 of prior-winter time series and summer rainfall at 160 stations in China: (a) Before removing the linear signal of ENSO; (b) after removing the linear signal of ENSO.The area shown in the box is the Yangtze River valley; the dotted areas denote regression above the 90% confidence level

 ${I_{{\rm{CM}}}} = 0.48{T_{\rm{A}}} + 0.37{T_{\rm{B}}},$ (4)

 图 3 1980~2012年长江流域夏季降水指数SRI与冬季澳大利亚周边海表温度相关系数分布。A、B方框区域表示海温关键区；深色阴影区通过95%信度检验 Figure 3 Correlation coefficient between the Summer Rainfall Index (SRI) in the Yangtze River valley and the prior-winter sea surface temperature around Austria during 1980-2012.Areas A and B shown in the box are the key regions of SST; the dark shaded areas denote correlation above the 95% confidence level

 图 4 1980~2012年长江流域夏季降水指数SRI与一致模指数CMI的标准化时间序列 Figure 4 Normalized time series of the summer rainfall index (SRI) and consistent mode index (CMI) over Yangtze River valley

 图 5 (a)一致模指数与中国160站降水量回归分布。一致模指数(b)高值年、(c)低值年降水距平(单位：mm month-1)合成场。虚线方框区域表示长江流域；打点区域超过90%信度检验 Figure 5 (a) The regression (units: mm month-1) of CMI and summer rainfall at 160 stations in China.The composite distributions of precipitation anomalies (shaded, units: mm month-1) for (b) high CMI years and (c) low CMI years.The area shown in the box is the Yangtze River valley; areas above 90% confidence level are dotted

4 海温一致模影响长江流域夏季降水可能的物理机制

 图 6 一致模指数与500 hPa位势高度场(上，单位：gpm)、850 hPa风场(下，单位：m s-1)回归分布：(a、d)冬季，(b、e)春季，(c、f)夏季。图a-c的阴影区表示超过90%信度检验 Figure 6 Regressions between CMI and 500-hPa geopotential height (top panels, units: gpm), 850-hPa wind (bottom panels, units: m s-1): (a, d) Winter, (b, e) spring, (c, f) summer.Areas in Figs.a-c above 90% confidence level are shaded

 图 7 一致模指数高值年对应的El Niño事件(左列)、La Niña事件(中列)、非ENSO年(右列)海表温度距平(单位：K)分布：(a)、(b)、(c)冬季；(d)、(e)、(f)春季；(g)、(h)、(i)夏季。打点区域通过90%信度检验 Figure 7 Composite fields of SSTA (units: K) around Austria for the El Niño events (left), La Niña events (middle), and Non-ENSO events (right) corresponding to high CMI years: (a), (b), (c) The prior-winter; (d), (e), (f) spring; (g), (h), (i) summer.Areas above 90% confidence level are dotted

 图 8 同图 7，但为一致模指数低值年 Figure 8 As in Fig. 7, but for low CMI years

 图 9 一致模指数(a)高值年和(b)低值年水汽输送通量(箭头，单位：kg m-1 s-1)合成及(c)合成差值场，阴影部分通过95%的信度检验 Figure 9 Composite water vapor transport fluxes (arrows, units: kg m-1 s-1) in (a) high CMI years, (b) low CMI years, and (c) the differences of composite water vapor transport fluxes between high and low CMI years over the middle and lower reaches of the Yangtze River in summer.Areas in Fig.c above 95% confidence level are shaded

 图 10 (a)100°~120°E平均的一致模指数高值年与低值年垂直速度差值场(单位：m s-1)。阴影区表示通过90%信度检验，负值(虚线)表示有上升运动，正值(实线)表示有下沉运动。(b)一致模指数高值年与低值年OLR差值场(单位：W m-2)。阴影区域表示绝对值大于6 W m-2，负值(虚线)表示对流加强，正值(实线)表示对流抑制 Figure 10 (a) The composite differences of vertical wind (contours, units: m s-1) between high and low CMI years averaged over 100°-120°E.Areas above 90% confidence level are shaded; the negative values (dashed lines) indicate ascending motion and the positive values (solid lines) represent descending motion.(b) The composite differences of outgoing longwave radiation (units: W m-2) between high and low CMI years.Shaded areas represent absolute values greater than 6 W m-2; the negative values (dashed lines) indicate the convection is strengthened and the positive values (solid lines) mean the convection is suppressed

5 小结和讨论