大气科学  2016, Vol. 40 Issue (6): 1199-1214 PDF

1 南京信息工程大学江苏省气象灾害预报预警协同创新中心, 南京 210044
2 中国气象科学研究院灾害天气国家重点实验室, 北京 100081
3 南京信息工程大学气象灾害省部共建教育部重点实验室, 南京 210044
4 嘉兴市气象局, 嘉兴 314050

Interannual and Interdecadal Variations of Summer Rainfall Duration over the Middle and Lower Reaches of the Yangtze River in Association with Anomalous Circulation and Rossby Wave Activities
LI Minggang1,2, GUAN Zhaoyong1,3, MEI Shilong4
1 Jiangsu Province Collaborative Innovation Center for Meteorological Disasters Prediction and Evaluation, Nanjing University of Information Science & Technology, Nanjing 210044
2 State Key Laboratory of Severe Weather, Chinese Academy of Meteorological Sciences, Beijing 100081
3 Key Laboratory of China Education Ministry for Meteorological Disasters, Nanjing University of Information Science & Technology, Nanjing 210044
4 Meteorological Administration of Jiaxing, Jiaxing 314050
Abstract: Droughts and floods are not only related to abnormal rainfall frequency or intensity, but also associated with the duration of rainfall on multiple timescales.Based on daily rainfall data collected at 249 stations in eastern China and the ERA-interim reanalysis, the long term changes in rainfall duration over the middle and lower reaches of the Yangtze River (hereafter MLRYR) and associated large scale circulation patterns and Rossby wave energy dispersion have been investigated.In the recent 35 years, the average duration of summertime wet spells decreases while that of dry spells increases, indicating a decreasing (negative) trend of rainfall duration over the MLRYR.Further analysis shows that this trend is related to inter-decadal changes in the frequency of persistent rainfall extremes, which is obviously higher in the 1980s and 1990s and lower in the 2000s.The anomalous circulation patterns that describe the inter-decadal and inter-annual changes in rainfall duration over the MLRYR look similar in some regions including Southeast China and the South China Sea, but different between mid-to high-latitudes and lower-latitudes.On both inter-decadal and inter-annual timescales, southeastern China is under control of a notable anticyclonic circulation anomaly in the middle and upper troposphere.The airflow converges into the MLRYR in the middle and lower troposphere and diverges in the upper troposphere with air flowing away from the MLRYR to the ocean.However, associated with inter-decadal changes, an anomalous anticyclonic circulation is located to the east of the Urals while an anomalous cyclonic circulation is found over Mongolia in the lower and upper troposphere.Meanwhile, an anomalous cyclonic circulation can be found over the equatorial Indian Ocean in the middle and lower troposphere.In contrast, on the inter-annual timescale, anomalous anticyclonic circulation can be found on both the east and west sides of Lake Baikal in the lower and upper troposphere.Divergence occurs in the lower troposphere over northeastern Maritime Continent with the air moving towards the MLYRV, whereas convergence develops in the upper troposphere and the air flows away from the MLRYR to lower latitudes.Characteristic Rossby wave propagation and energy dispersion demonstrate significant differences between inter-decadal and inter-annual timescales.On the inter-decadal timescale, a Rossby wave train with alternatively positive-negative-positive-negative geopotential height anomaly can be found in the mid-latitude from the Atlantic to Mongolia.The eastward propagation of waver energy affects the MLRYR.In the mid-and lower troposphere, however, the wave energy dispersion from lower latitudes to MLRYR is relatively weak.On the inter-annual timescale, the Rossby wave activity flux demonstrates more distinct local features.In the lower troposphere, strong wave energy propagates from South China Sea in the lower latitudes to the MLRYR.In the upper troposphere, wave energy dispersion from regions to the west of Lake Baikal to MLRYR is more obvious.These results are helpful for our better understanding of the mechanism for the persistent rainfall anomaly and related droughts/floods over the MLRYR.
Key words: Middle and lower reaches of the Yangtze River      Rainfall duration      Rossby wave activity
1 引言

2 资料与方法 2.1 资料

(1)中国东部地区的249个站点的逐日降水资料，其在研究时段内均无缺测，范围包括了110°E以东除去黑龙江、吉林、内蒙古和台湾四个省份以外的全部省份(图 1)。该资料取自中国大陆743站点逐日资料集(http://www.escience.gov.cn/metdata/page/index.html [2015-10-25])。

 图 1 1979~2013年平均6~7月东部地区（a）降水总天数分布（阴影，单位：d）及TC雨日比例（点标记），（b）降水时段平均持续天数（阴影，单位：d）及其线性趋势（点标记）。图（c）同（b）但为无雨时段。图（d）和（e）均为1979~2013年6~7月总降水日数的变化趋势，但（d）给出了降水时段和无雨时段持续时间变化趋势相同的站点，而（e）给出了二者趋势相反的站点（在图b，c，d和e中，显著趋势表示通过了90%信度的显著性检验）。图（f）给出了东部地区各个站点90%分位上降水时段（剔除TC雨日后）持续天数阈值（单位：d）及长江中下游地区子区域划分 Figure 1 (a) Multi-year mean total rainy days (shaded, units: d) and the proportion of TC-related ones (scattered) in June-July over eastern China during the period of 1979 to 2013, and (b) mean duration of precipitation days (shaded, units: d) as well as the linear trends (scattered) of wet spell duration. (c) Same as in (b) but for dry spells. (d) and (e) present linear trends of total rain days in June-July during the period of 1979-2013, but those stations with similar trends of dry and wet spell durations are illustrated in (d) whereas those with opposite trends are shown in (e). In (b), (c), (d) and (e), open circles are for negative trends, closed circles for positive trends, half closed circles for those with no trends, and circles with a cross inside indicate the trends are statistically significant at/above 90% confidence level. The threshold values (units: d) of duration days of wet spells at 90th percentile (after TC-related rain days is removed) are shown in (f) along with the sub-region of middle and lower reaches of the Yangtze River

(2) ERA interim逐月再分析资料(http://apps.ecmwf.int/datasets/data/interim-full-moda [2015-10-25])，使用的变量包括经向风、纬向风、位势高度、比湿等，水平分辨率为1°×1°。

2.2 持续降水时段、无雨时段的定义

2.3 波作用通量

 $\mathit{\boldsymbol{W = }}\frac{p}{{2000\left| {\mathit{\boldsymbol{\bar U}}} \right|}} \times \left[ \begin{array}{l} u\left({{{v'}^2} - \psi '{{\mathit{v'}}_x}} \right) + v\left({ - u'v' + \psi '{{u'}_x}} \right)\\ u\left({ - u'v' + \psi '{{u'}_x}} \right) + v\left({{{u'}^2} + \psi '{{u'}_y}} \right) \end{array} \right]$

3 降水持续特征的气候平均态及长期变化 3.1 降水持续特征的气候平均态及变化趋势

1979~2013年多年平均6~7月降水总日数存在由南向北减少的分布。由图 1a(阴影)可见，辽宁东部、华东中南部、湖南西部及两广等地区降水总天数相对较多，均在20日以上，其中两广地区最多，绝大多数站点超过25日。而我国东部的其他区域相对较少，均不到20日。考虑到西北太平洋热带气旋(Tropical Cyclones，简写为TC)活动对我国东南部地区6~7月降水存在一定影响，我们使用500 km固定圆方法(Lau et al., 2008Nogueira and Keim, 2010Barlow，2011Dare et al., 2012)对每个站点的每个降水日做了检查。当降水日当日西北太平洋上存在TC，且站点与TC中心距离小于500 km时，则认为该降水日为TC雨日。6~7月份多年平均TC雨日占总雨日的比例上(图 1a，点标记)，东南沿海地区TC雨日的比例相对较大，达到10%，尤其在广东南部和海南岛地区，部分站点的TC雨日比例超过15%。相比之下，TC活动对长江中下游地区降水的影响并不大，TC雨日比例一般在1%~5%左右。

3.2 长江中下游地区降水持续特征的年际及年代际变化

 图 2 长江中下游地区区域平均的6~7月（a）降水平均持续天数的逐年变化（点实线，单位：d）及线性趋势，（b）以及持续性降水事件总降水天数的逐年变化（点虚线，单位：d）及线性趋势 Figure 2 (a) Inter-annual variability of area-averaged duration of wet spells (solid line with closed circles, units: d) averaged over June-July over the middle and lower reaches of the Yangtze River for the period of 1979 to 2013 and its linear trend, and (b) variability of total rain days (dashed line with open circles, units: d) for persistent rainfall events as well as the linear trend

 图 3 长江中下游地区区域平均6~7月持续性降水事件降水总天数的（a）年代际变化分量（空心点虚线）及年际变化分量（实心点实线）的标准化时间序列，以及长江中下游地区6~7月持续性降水事件的（b）累计降水量（实心柱，单位：mm）及其占总降水量的百分比（空心柱）的逐年变化 Figure 3 (a) Normalized time-series of total rain days for persistent rainfall events on inter-decadal (dashed line with open circles) and inter-annual (solid line with closed circles) timescales in June-July over the middle and lower reaches of the Yangtze River, and (b) the total rainfall (solid bars, units: mm) for persistent rainfall events in June-July and its percentage (open bars) in the total precipitation amount in June-July

6~7月长江中下游地区持续性降水事件累计降水量值的逐年变化(图 3b)与总天数的变化(图 2b)对应亦较好。持续性降水事件总天数多的年份累计降水量值及其占总雨量的比例亦较大，如1983、1991、1998和1999年，累计雨量及其占6~7月总雨量的比例均分别达到300 mm和45%以上，远高于1979~2013年的平均值(152 mm和31%)。尤其是在1991年，累计雨量为401 mm，占总雨量的比例达到64%，而这一年持续性降水事件总降水天数为12.2日，接近多年平均值(6.35日)的两倍。

4 与降水持续特征变化相联系的环流型

4.1 年代际变化

 图 4 长江中下游地区持续性降水事件总降水天数多、寡年代（a）850 hPa、（b）500 hPa、（c）200 hPa高度上差值合成的旋转风流场（阴影区表示通过95%信度t检验）和辐散风场（红色粗箭头表示通过95%信度t检验，单位：m s-1）以及（d）差值合成的整层（1000~200 hPa）积分水汽通量速度势（阴影，单位：106 kg s-1）及辐散分量（矢量场，绿色粗箭头表示通过95%信度t检验，单位：kg m-1 s-1)） Figure 4 Composite differences in anomalous circulations at (a) 850 hPa, (b) 500 hPa, and (c) 200 hPa and (d) the water vapor fluxes integrated from the earth surface up to 200 hPa between years with extremely high and low values of total rain days for persistent rainfall events on inter-decadal timescales. In (a), (b), and (c), streamlines indicate the rotational components and the shaded areas indicate the difference is statistically significant at/above 95% confidence level based on t test. The arrows represent divergent components (units: m s−1) and the red ones indicate they are statistically significant at/above 95% level of confidence. In (d), the shaded contours denote velocity potential (units: 106 kg s−1) of anomalous water vapor fluxes while vectors display the divergent components of the vapor fluxes (units: kg m−1 s−1) and the green vectors indicate they are statistically significant at/above 95% confidence level

4.2 年际变化

 图 5 同图 4，但为年际时间尺度 Figure 5 Same as in Fig. 4, but for inter-annual timescale

5 与降水持续性异常相联系的准定常Rossby波活动

 图 6 长江中下游地区持续性降水事件总降水天数多、寡年代（a）700 hPa、（b）500 hPa、（c）200 hPa高度上位势高度场的合成差值（阴影，单位：gpm）及相应的波作用通量（矢量，单位：m2 s-2） Figure 6 Composite differences in geopotential height (shades, in gpm) and wave activity fluxes (vectors, in m2 s-2) at (a) 700 hPa, (b) 500 hPa, and (c) 200 hPa between years of extremely high and low values of total rain days for persistent rainfall events on inter-decadal timescale

 图 7 同图 6，但针对年际时间尺度上的扰动 Figure 7 Same as in Fig.6, but for those on inter-annual timescales

6 结论与讨论