1.State Key Laboratory of Severe Weather, Chinese Academy of Meteorological Sciences, Beijing 100081, China 2.Department of Atmospheric and Oceanic Science, University of Maryland, College Park, Maryland 20742-2425, USA 3.LASG, Institute of Atmospheric Physics, Chinese Academy of Sciences, Beijing 100029, China Manuscript received: 2019-05-08 Manuscript revised: 2019-08-27 Manuscript accepted: 2019-09-25 Abstract:In this study, power spectral analysis and bandpass filtering of daily meteorological fields are performed to explore the roles of synoptic to quasi-monthly disturbances in influencing the generation of pre-summer heavy rainfall over South China. Two heavy rainfall episodes are selected during the months of April?June 2008?15, which represent the collaboration between the synoptic and quasi-biweekly disturbances and the synoptic and quasi-monthly disturbances, respectively. Results show that the first heavy rainfall episode takes place in a southwesterly anomalous flow associated with an anticyclonic anomaly over the South China Sea (SCS) at the quasi-biweekly scale with 15.1% variance contributions, and at the synoptic scale in a convergence zone between southwesterly and northeasterly anomalous flows associated with a southeastward-moving anticyclonic anomaly on the leeside of the Yungui Plateau and an eastward-propagating anticyclonic anomaly from higher latitudes with 35.2% variance contribution. In contrast, the second heavy rainfall episode takes place in southwest-to-westerly anomalies converging with northwest-to-westerly anomalies at the quasi-monthly scale with 23.2% variance contributions to the total rainfall variance, which are associated with an anticyclonic anomaly over the SCS and an eastward-propagating cyclonic anomaly over North China, respectively. At the synoptic scale, it occurs in south-to-southwesterly anomalies converging with a cyclonic anomaly on the downstream of the Yungui Plateau with 49.3% variance contributions. In both cases, the lower-tropospheric mean south-to-southwesterly flows provide ample moisture supply and potentially unstable conditions; it is the above synoptic, quasi-biweekly or quasi-monthly disturbances that determine the general period and distribution of persistent heavy rainfall over South China. Keywords: synoptic scale, pre-summer rainfall, quasi-biweekly scale, quasi-monthly disturbances 摘要:本研究基于对2008-15年4-6月逐日气象要素场的能量谱分析以及带通滤波,探讨了天气尺度至准月尺度扰动在华南前汛期强降水产生过程中的作用,进而具体分析了天气尺度和准两周尺度扰动,天气尺度和准月尺度扰动在两次强降水过程中的作用。研究结果表明,对第一次强降水过程而言,准两周尺度扰动的贡献率为15.1%,降水主要发生在与南海反气旋异常相关的异常西南风中。天气尺度扰动的贡献率为35.2%,降水主要处于异常西南风和东北风的辐合区,其分别与位于云贵高原下风坡的异常反气旋向东南移动以及来自更高纬的异常反气旋东移有关。与之不同的是,对第二次强降水过程而言,准月尺度扰动的贡献率为23.2%,降水主要发生在西西南异常与西西北异常的辐合区,该异常分别与南海反气旋以及华北气旋东移有关。在天气尺度上,降水发生在西南偏南风异常与云贵高原下游的气旋性异常辐合区,贡献率为49.3%。对两个个例而言,对流层下层平均西南偏南气流提供了充足的水汽和潜在的不稳定条件;但天气尺度、准两周尺度以及准月尺度扰动综合决定了华南持续性强降水的大概持续时间和空间分布。 关键词:天气尺度, 前汛期降水, 准两周尺度, 准月尺度扰动
HTML
--> --> --> -->
4.1. The heavy rainfall episode that peaked on 8 May 2014
Figure 4 shows the daily mean flows at 850 hPa, and the daily mean geopotential height field and precipitable water during 6?8 May 2014. It is evident that on the heavy rainfall day of 8 May, South China is dominated by warm and moist southwesterly to southerly flows of tropical origin (even from the Indian Ocean) in the lower troposphere (Fig. 4c). The southwesterly flow is associated with the WNPSH, whose axis extends from the central North Pacific to midwest Indian Ocean. This southwesterly flow turns cyclonically in the north of the target domain to a more southerly flow at about 25°N and southeasterly flow at higher latitudes, as an anticyclone (denoted by “A”) moves to the middle-east of China, which strengthens the supply of warm and moist air by southwesterlies to the target domain (cf. Figs. 4b and c). At 500 hPa, South China is located ahead of a weak trough axis with prevailing west-southwesterly winds extending from the southeast of the Tibetan Plateau (Fig. 4f). Clearly, the juxtaposition of midlevel quasi-geostrophic lifting, albeit weak, with the lower-level southwesterly warm and moist air from the SCS and Indian Ocean helps pre-condition a favorable environment for heavy rainfall production. The favorable environment, though relatively weak, appears to account for the generation of a near-zonally oriented rainfall belt, which is consistent with the mean flow at 500 hPa (cf. Figs. 4a, b and 4d, e). The rainfall pattern occurring on 8 May looks more like a warm-sector rainfall episode. In particular, precipitable water in the upstream regions reaches more than 50 mm, implying the presence of considerable water vapor available for the generation of heavy rainfall. Note that although the precipitable water pattern in the vicinity looks similar during the period of May 7?8, the moisture flux, the product of water vapor and horizontal winds, increases with increasing southwesterly flows (Figs. 4a-c), with its peak flux on 8 May. Ultimately, it is the moisture flux convergence that determines the amount of rainfall over the region. Figure4. Temporal evolution of (a?c) the daily mean 850-hPa wind vectors (units: m s?1), superimposed with daily mean rainfall (shaded; units: mm), and (d?f) the daily mean 500-hPa geopotential height (solid lines; units: gpm) and precipitable water (shaded; units: mm) during the period 6?8 May 2014. The thick solid line refers to the 5880-m geopotential isoline at 500 hPa. The gray shading represents topography greater than 1600 m. (a, d) 6 May; (b, e) 7 May; (c, f) 8 May. The inner box dashed in red denotes the target area, similarly for the rest of figures. The capital letter “A” (“C”) represents anticyclone (cyclone), similarly for the rest of figures.
However, the above favorable mean flow conditions, resulting from the interaction of several atmospheric weather systems, are significantly different when they are traced back two days earlier, i.e., 6 May (Fig. 4a). That is, the anticyclone, which accounts for the cyclonic turning of the southwesterly flow, originates from Mongolia, west of an extratropical cyclone (denoted by “C”) (Fig. 4b). It grows in amplitude as it intrudes southeastward (Figs. 4b and c). Meanwhile, the WNPSH retreats equatorward, as indicated by the 5880-m geopotential isoline at 500 hPa (Figs. 4d-f), but strengthens, as it extends more into the Bay of Bengal—similarly for the strengthened southwesterly flow, as indicated by the horizontal wind vectors. At 500 hPa, South China is governed by westerly flow, with the southeastward intrusion of colder and drier air on the northern side of the Tibetan Plateau on 6 May (Fig. 4d). This southeastward intrusion, as indicated by a dry pocket (shown in blue shades) of precipitable water, helps hydrostatically strengthen the midlevel trough (cf. Figs. 4e and f). After seeing the evolution of large-scale mean flows in relation to the generation of the heavy rainfall episode that peaked on 8 May 2014, we can now examine what circulation anomalies at the quasi-biweekly and synoptic scales are involved. For this purpose, Figs. 5 and 6 show the filtered quasi-biweekly (10?20-day) and synoptic scale (3?8-day) disturbances, respectively, in different phases at 850 hPa and 200 hPa in association with the filtered rainfall. In the driest phase (i.e., phase 1, or P1) of the quasi-biweekly scale (Fig. 5, left column), South China is dominated by northeasterly anomalies, with an anticyclonic anomaly (A1) on the southeast side of the Tibetan Plateau, and a cyclonic anomaly (C) and an anticyclonic ridge (A2) over the SCS. Although the anticyclonic (A1) and cyclonic (C) anomalies gradually move eastward to eastern-central China and the Bashi Strait, respectively, from P2 to P3, the northeasterly anomaly still dominates South China. Only after the midlatitude anticyclonic anomaly (A1) migrates into the Sea of Japan and the anticyclonic ridge (A2) strengthens into an anticyclonic anomaly in P4 does a large-scale southwesterly anomaly associated with A2 begin to prevail on the southeast side of the Tibetan Plateau. As a result, South China is influenced by the weak northeasterly and weak southwesterly anomalies, which are associated with the above-mentioned anticyclonic anomaly (A1) and summer monsoon, respectively. The anticyclonic anomaly (A2) over the SCS helps strengthen the southwesterly monsoonal flow across South China and generate a confluence zone along the coastline. Figure5. (a?e) Phase evolution of the quasi-biweekly bandpass-filtered daily mean 850-hPa wind vector (units: m s?1) superimposed with quasi-biweekly rainfall rate anomalies (shaded; units: mm d?1). Black-dashed lines refer to the 5880-m geopotential isoline at 500 hPa. (f?j) As in (a?e) except for the quasi-biweekly bandpass-filtered daily mean 200-hPa wind vector (units: m s?1) and the corresponding divergence (shaded; units: s?1). (a, f) P1 on 3 May; (b, g) P2 on 4 May; (c, h) P3 on 5 May; (d, i) P4 on 7 May and (e, j) P5 on 8 May 2014. A1, A2 represent anticyclone, C represents cyclone.
Figure6. (a?c) Phase evolution of the 3?8-day bandpass-filtered (a) daily mean 850-hPa wind vector (units: m s?1) superimposed with 3?8-day bandpass-filtered rainfall rate anomalies (shaded; units: mm d?1). Black-dashed lines refer to the 5880-m geopotential isoline at 500 hPa. (d?f) Temporal evolution of the 3?8-day bandpass-filtered daily mean 200-hPa wind vector (units: m s?1), the total wind speeds (dashed line; units: s?1) and divergence (shaded; units: s?1). (a, d) P1 on 6 May; (b, e) P3 on 7 May; (c, f) P5 on 8 May 2014.
In the wettest phase (i.e., P5), the southwesterly anomaly reaches a larger intensity than before, but gradually weakens from the southwestern to northeastern part in the target domain. Furthermore, this anomalous flow together with the trailing anomalous easterly flow of the anticyclonic anomaly (A1) leads to an intense southeast?northwest-oriented convergence zone where heavy rainfall is distributed on its southern side (see P5 in Fig. 5). It follows that the converging warm and moist air masses at the quasi-biweekly scale contribute positively to the generation of this heavy rainfall episode. The quasi-biweekly filtered fields in the upper troposphere show the migration of a cyclonic anomaly into the study area from the west during P1?P5 (Fig. 5, right column), which gives rise to the presence of south-to-southwesterly anomalies over South China. In P4, a pronounced divergence region is observed between Hainan and Taiwan, and in P5 it expands toward the southern coastal region, where diffluent horizontal wind vectors are evident. This upper-level diffluence is closely collocated with the lower-tropospheric convergence area. Thus, such a tropospheric juxtaposition tends to generate a favorable upward motion for the generation of the present heavy rainfall episode. On the synoptic time scale, the lower troposphere from P1 to P3 exhibits a southeastward propagating anticyclonic anomaly (i.e., A), which originates from the lee side of the Yungui Plateau at P1 and dominates South China with two centers in P3 (Fig. 6, left column). This scenario is different from that of a pre-summer heavy rainfall mode found by Huang et al. (2018). In P5, however, this anticyclonic anomaly moves southeastward into the SCS. This displacement allows a cyclonical turning of northeasterly anomalous flows (with likely cold air), associated with a higher-latitude anticyclonic anomaly, over the target domain, thereby assisting in lifting the southwesterly warm and moist air for the generation of the heavy rainfall episode that peaked on 8 May 2014. The upper troposphere at the synoptic time scale shows the southeastward propagation of a cyclonic anomaly, which is similar to but much better organized than that at the quasi-biweekly scale, with the presence of an upper-level jet stream around 35°N. South China is located over the southern entrance region of the jet stream, where favorable divergence can be seen as expected (Uccellini and Johnson, 1979). In P5, the cyclonic anomaly moves close to the heavy rainfall region, where significant diffluence is present, like that on the quasi-biweekly time scale (cf. Figs. 6 and 5). To further see the contribution of quasi-biweekly and synoptic disturbances to the heavy rainfall production, the latitude?height cross sections of divergence along 115°E in P5 for the two different scales of disturbances are given in Fig. 7. One can see evident convergence in the 850?400-hPa layer with two convergence centers in the vertical direction, and divergence aloft, implying the presence of favorable deep local upward motion for the rainfall generation. This should also be expected from Figs. 5 and 6, showing isentropic lifting as a warm and moist air parcel moves northward in the southwesterly flow (Raymond and Jiang, 1990; Zhang and Zhang, 2012). In general, the amplitudes of convergence and divergence of the 3?8-day filtered disturbances are twice more than those of the 10?20-day filtered disturbances. This suggests the more important dynamical forcing of the synoptic scale disturbances than the quasi-biweekly ones in influencing the persistent heavy rainfall event. Figure7. Meridional?height (y, z) cross sections of divergence (shaded; units: s?1) and in-plane flow vectors (units: m s?1) with the vertical motion magnified by 500, taken along 115°E in P5 for (a) the 10?20-day bandpass-filtered disturbances and (b) the 3?8-day bandpass-filtered disturbances.
Figures 8a and b show the temporal evolution of the respective 10?20-day and 3?8-day bandpass-filtered area-averaged moisture and potential temperature tendencies, superimposed with the area-averaged filtered vertical motion disturbances (w). The above area-averaged tendencies are calculated from their corresponding flux convergence across the target area: Figure8. (a) Time?pressure diagram of the 10?20-day bandpass-filtered anomalies of moisture flux convergence (shaded; units: g kg?1 s?1), thermal flux convergence (contours; units: K s?1) and meridional?vertical (y, z) flow vectors (units: m s?1) with vertical motion magnified by 500 averaged over the area (21°?25°N, 112°?117°E) on 3 (P1), 5 (P3), 8 (P5), 12 (P7) and 15 (P9) May 2014. (b) As in (a) but for the 3?8-day bandpass-filtered fields on 6 (P1), 7 (P3), 8 (P5), 9 (P7) and 10 (P9) May 2014.
where V denotes the horizontal wind vector, ω is the vertical velocity in P-coordinates; s is the target domain, and $\pi $ denotes a conserved variable (i.e., ${{{\rm{d}}\pi }}/{{{\rm{d}}t}} = 0$), which is the specific humidity or potential temperature herein. This equation is derived by combining ${{{\rm{d}}\pi }}/{{{\rm{d}}t}} = 0$ and the mass continuity equation. At the quasi-biweekly scale (Fig. 8a), we can see favorable upward motion from P4 to P6. From P1 to P3, increased warm and moist air prevails below 800 hPa, and extends upward to about 700 (500) hPa during P4 (P5), which provides sufficient moisture supply and potential instability (i.e., with warm air below cold air) to be favorable for the convective development associated with the heavy rainfall episode. The cold and dry air above 800 hPa from P1 to P3 is consistent with the presence of dominant northerly flow shown in Fig. 5. After P5, dry and cold air tendency appears, as can be expected. By comparison, the synoptic-scale fields show the presence of warm and moist disturbances in the lower troposphere, mostly below 850 hPa prior to P3, and they gradually weaken with time (Fig. 8b). Clearly, the quasi-biweekly and synoptic-scale disturbances both provide the moisture and energy needed for the generation of heavy rainfall. Certainly, the quasi-biweekly disturbances appear to generate more significant thermal and moisture transport than the synoptic disturbances to the heavy rainfall production, especially during P4 to P5.
2 4.2. The heavy rainfall episode that peaked on 21 May 2013 -->
4.2. The heavy rainfall episode that peaked on 21 May 2013
Figure 9 shows the temporal evolution of large-scale mean flows at 850 and 500 hPa associated with a heavy rainfall episode that peaked on 21 May 2013 (see Fig. 3c). Results show that this episode occurs ahead of a mesoscale cyclone (denoted by “C”) to the north-to-northwest under the influences of a southwesterly monsoonal flow originating from the Indian Ocean in the lower troposphere (Figs. 9a-c). Of relevance here is that the southwesterly flow strengthens during the period of 19?21 May, in association with the generation of an elongated southwest?northeast-oriented rainband. Corresponding to the lower-tropospheric mesocyclone is an intensifying trough at 500 hPa that extends from Northeast China to North Bay from 19 to 21 May (Figs. 9d-f). Unlike the 8 May 2014 rainfall episode, we see greater amplitude and spatial extent of the midlevel trough with a surface cold front, indicating more important roles of the higher-latitude cold and dry air intrusion behind, especially in the northwest of the target area, and favorable lifting ahead of the trough axis. In addition, the WNPSH retreats more eastward to the Indochina Peninsula during this rainfall episode. Precipitable water of more than 60 mm is distributed along the coastal land and water regions, where the heavy rainfall event takes place. On average, this precipitable water amount is at least 10 mm higher than that associated with the episode that peaked on 8 May 2014 (cf. Figs. 9 and 4). Figure9. As in Fig. 4 but during the period 19?21 May 2013: (a, d) 19 May; (b, e) 20 May; (c, f) 21 May.
Figure 10 shows the roles of the band-filtered quasi-monthly disturbances in determining the heavy rainfall episode that peaked on 21 May 2013. It is evident that this episode occurs in southwest-to-westerly anomalies converging with northwest-to-westerly anomalies located on the south of a cyclonic anomaly at 850 hPa in P5 (Fig. 10, left column). This cyclonic anomaly can be traced back to North China (denoted by “C”) in P1. The southeastward movement of this cyclonic perturbation from P1 to P5 facilitates the influences of westerly to northwesterly anomalies over South China. The associated southwesterly anomaly accounts for the advection of warm and moist air into the heavy rainfall region in P5. In particular, the heavy rainfall zone coincides well with the confluence of an anomalous northwesterly flow associated with the eastward-moving cyclonic anomaly and anomalous west-to-southwesterly flows associated with a quasi-stationary anticyclonic anomaly over the SCS during P4 and P5. This suggests the presence of cold frontal lifting at this time scale, which facilitates the generation of favorable conditions for the heavy rainfall production. Again, the heavy rainfall generation occurs during the eastward retreat of the WNPSH from the Bay of Bengal to the east side of the Indochina Peninsula (cf. Figs. 10 and 5). In the upper troposphere, a Rossby wave train with an anticyclonic anomaly is seen propagating northeastward into the western North Pacific during P1 to P5 (Fig. 10, right column). A distinct divergence zone associated with the propagating anticyclonic anomaly is observed to move southeastward across the target area, which facilitates a favorable vertical circulation for the occurrence of heavy rainfall. Figure10. As in Fig. 5 but with 15?40-day bandpass-filtered fields: (a, f) P1 on 11 May; (b, g) P2 on 13 May; (c, h) P3 on 15 May; (d, i) P4 on 18 May; (e, j) P5 on 21 May 2013.
To further illustrate the origins of quasi-monthly disturbances, Fig. 11 shows the evolution of the 15?40-day filtered OLR field. In the driest phase (P1), a weak positive (with depressed convection) OLR anomaly over South China is accompanied to its southeast by a strong negative OLR anomaly around the Philippine islands. This means that the quasi-monthly oscillation of heavy rainfall over South China may have a connection with the anomalous convective system in the western North Pacific and the SCS. This seesaw structure weakens in intensity as it migrates northwestward from P1 to P3. Its northwestward migration is more evident from P3 to P5, during which period a negative OLR anomaly (i.e., with active convective development) approaches the edge of South China in P4 and covers the rainfall center in P5. At the same time, another positive OLR anomaly is seen on its southeast, which represents an eastward shift of the WNPSH, as also shown in Fig. 10. This implies that the 15?40-day oscillation of pre-summer rainfall over South China may also correspond to variations in the WNPSH, in which the wave-like anomalous convective system originates from the equatorial western Pacific. In fact, this kind of opposite-phase variation in convective anomalies between South China and the SCS?Philippine islands is very similar to that found between the Yangtze River basin and the SCS?Philippine islands, as found by Mao and Wu (2006) and Mao et al. (2010). Figure11. As in Fig. 10 but for the 15?40-day filtered OLR (units: W m?2).
On the synoptic time scale (Fig. 12, left column), the filtered disturbances at 850 hPa show the presence of a cyclonic anomaly (i.e., “C”) in P5 that corresponds to a cyclone in the northwest (cf. Figs. 12 and 9c). Like that shown in Figs. 9a-c, this anomaly results from the southwestward extension of a cyclonic disturbance in P1. This disturbance grows in both amplitude and coverage from P1 to P5, due partly to the continued latent heat release associated with the heavy rainfall event. The southwestward extension of this cyclonic anomaly is facilitated by the eastward movement of an anticyclonic anomaly (i.e., “A”) out of the target area from P3 to P5. As the cyclonic anomaly approaches the target area, South China is under the influence of south-to-southwesterly anomalous warm and moist currents. In the upper troposphere (Fig. 12, right column), a higher-latitude Rossby wave train with a cyclonic anomaly is seen propagating southeastward. South China is located over the southern entrance region of an upper-level jet stream, where favorable diffluence facilitates the generation of favorable vertical motion for pre-conditioning a convectively unstable environment. Figure12. As in Fig. 6 but in (a, d) P1 on 19 May; (b, e) P3 on 20 May; and (c, f) P5 on 21 May 2013.
Figure 13 summarizes the temporal evolution of the area-averaged quasi-monthly and synoptic bandpass-filtered moisture flux convergence, thermal flux convergence and vertical motion fields, which differs somewhat from those associated with the episode that peaked on 8 May 2014. Specifically, at the quasi-monthly scale (Fig. 13a), upward motion with cold and dry anomalous air appears from P3 to P6, which corresponds to the northwest-to-westerly anomalous advection from higher latitudes (cf. Figs. 13a and 10). The opposite is true both prior to P2 and after P6. However, large moisture and heat flux divergence (convergence) prevails above (below) 850 hPa prior to P3. Subsequently, both the moisture flux divergence and the heat flux divergence decrease in magnitude first and then switch sign prior to P5, which is consistent with the southerly warm and moist flows into the target region (cf. Figs. 13b and 12f). Figure13. As in Fig. 8 but with (a) 15?40-day on 11 (P1), 15 (P3), 21 (P5), 27 (P7) and 31 (P9) May 2013 and (b) 3?8-day on 19 (P1), 20 (P3), 21 (P5), 22 (P7) and 23 (P9) May 2014.
PLOS Biology, 16 Apr, 2019, DOI: http://dx.doi.org/10.1371/journal.pbio.3000211Distinct roles of two myosins in C. elegans spermatid differentiationJu ...