Diurnal Variations of Precipitation over North China Regulated by the Mountain-plains Solenoid and B

HTML

--> --> -->

2. Data and methods

2.1. Satellite-derived rainfall and atmospheric reanalysis data

-->

2.1. Satellite-derived rainfall and atmospheric reanalysis data

To study precipitation activities, we use the Climate Prediction Center morphing technique (CMORPH) global precipitation analyses from NOAA for July and August in 1998-2016. CMORPH incorporates precipitation derived from the merged data of passive microwave sensors on various low-orbit spacecraft (Joyce et al., 2004; Janowiak et al., 2005). It provides rainfall data at a high resolution of 8 km every 30 minutes, and thus contributes fine-scale spatial details of rainfall. CMORPH is also widely used to show the diurnal cycle of rainfall and its regional characteristics over East Asia (He and Zhang, 2010; Jiang et al., 2017; Cai et al., 2018; Chen et al., 2019). Note, however, that CMORPH tends to overestimate afternoon precipitation and underestimate morning precipitation, probably because of the residual cloud anvil from afternoon local convection on land (Chen et al., 2018). After removing such systematic errors, CMORPH can capture well the spatial distributions of the DVP related to complex terrains. To relieve the possible effect of systematic bias, in this study we focus on the differences of precipitation under different conditions.
To present the atmospheric conditions, we use the Japanese 55-year Reanalysis from the Japan Meteorological Agency (JRA-55; Kobayashi et al., 2015). The JRA-55 model has a horizontal resolution of ~55 km. It provides pressure-level data products at a spatial resolution of 1.25^° longitude × 1.25^° latitude and a temporal resolution of 6 h. With a vertical spacing of 25 hPa below 750 hPa, it resolves well the atmospheric conditions in the lower troposphere. The new-generation reanalyses, such as JRA-55, with their improved model physical processes and data assimilation schemes, outperform their predecessors in almost all aspects, and they are increasingly being used to study atmospheric circulation and hydrological processes (Harada et al., 2016). (Chen et al., 2014a) comprehensively evaluated the performance of four new reanalysis datasets for representing the diurnal cycles over East Asia. JRA-55 is the best at capturing the diurnal variations in winds and temperature, suggesting a reliable representation of the diurnally varying dynamic processes. Moreover, JRA-55 also reproduces the regional differences and eastward propagation of the DVP over East Asia, indicating that it captures well the response of precipitation to wind diurnal variations.
In this study, we focus on North China, with mountains in the west and plains in the east, which shows an obvious difference in topography (Fig. 1a). The elevation of 700 m is chosen to divide mountains and plains. The period of interest is July-August, when the rainfall amount reaches its peak in North China (Fig. 1b). As we know, a diurnal cycle includes an afternoon phase (with strong solar heating and boundary-layer mixing) and a subsequent nocturnal phase. In general, the MPS develops from afternoon to evening (PM hours), while the low-level winds tend to veer under inertial oscillation from midnight to morning (AM hours). Local solar time ( LST= UTC+8) is applied throughout this study. The PM hours refer to the hours from 1200 to 2300 LST, while the AM hours refers to 0000 to 1100 LST of the next day. For the 6-h reanalysis data, 1400 and 2000 LST denote afternoon and early evening, while 0200 and 0800 LST denote late night and morning.

2

2.2. Definition of the intensities of the MPS, BLO and monsoon affecting North China

-->

2.2. Definition of the intensities of the MPS, BLO and monsoon affecting North China

Here, we briefly describe the MPS in North China and then propose an index to describe its intensity. It is recognized that the MPS includes daytime thermal-driven circulation and nighttime reversed circulation. Figures 2a and b show that the rising motion develops over the mountains at 1400 LST, and a circulation loop is formed between the mountains and plains at 2000 LST. The rising motion develops in the plains at 0200 LST and weakens at 0800 LST (Figs. 2c and d). Thus, we may estimate the MPS intensity over North China using four parameters of vertical-motion deviation at 800 hPa: rising (sinking) motion over mountains and sinking (rising) motion over plains at 2000 (0200) LST. The grids of mountains and plains are shown in Fig. 1a. A strong (weak) MPS event is defined when all four parameters are above (below) their climate mean, which is referred to as MPS(+) (MPS(-)). During July-August of 1998-2016, the daytime numbers of MPS(+) and MPS(-) are 123 and 127, with frequencies of 10.4% and 10.8%, respectively (Table 1).

Figure2. Longitude-vertical sections of 6-h zonal circulation (vectors: zonal wind, units: m s^-1; omega multiplied by 20, units: Pa s^-1) and the vertical velocity anomaly (shaded; units: 10^-2 Pa s^-1) at 37.5^°N averaged from July to August during 1998-2016. The contours denote the original value of vertical velocity (units: 10^-2 Pa s^-1).

Another forcing is the diurnal variation of low-level winds caused by the boundary-layer friction and inertial oscillation (the BLO). It is recognized that the top of the atmospheric boundary layer becomes highest in the afternoon due to turbulent vertical mixing and decreases at night with weakened mixing (Chen et al., 2013; Allabakash et al., 2017). Thus, we focus on the diurnal variation of wind at 925 hPa, as it is strongly affected by boundary-layer processes. Figure 3 shows that the diurnal deviation at 925 hPa is characterized as northerly winds at 1400 LST and southerly winds at 0200 LST, with a diurnal range of ~2 m s^-1 over the whole plains of eastern China. The southerly deviation at 0200 LST in the Yangtze River valley (28.5^°-32^°N, 112^°-120^°E), away from the mountains, may be used to denote the intensity of the BLO affecting North China. In this study, the days with southerly deviation at 0200 LST exceed its summer mean (1.04 m s^-1) plus (minus) one standard deviation (0.72 m s-1) are categorized as the strong (weak) BLO events, which are referred to as BLO(+) (BLO(-)). There are 186 days of BLO(+) and 188 days of BLO(-), with frequencies of 15.8% and 16.0%, respectively (Table 1). Thus, the BLO (MPS) may represent the regional forcings from the southern (western) boundary of the North China plains. Note that the days with the occurrence of both BLO(+) and MPS(+) are few (27), accounting for 14.5% of BLO(+) and 21.9% of MPS(+). Thus, the two regional forcings seem somewhat independent.

Figure3. Diurnal component of 925-hPa horizontal winds (vectors; units: m s^-1) at four synoptic hours averaged from July to August during 1998-2016. The daily mean at each grid point has been removed. The shaded areas denote the topography, with the dashed contour for the elevation of 700 m.

We also focus on the influence of different monsoon conditions. It is recognized that the southerly winds represent well the intensity of the East Asian summer monsoon (Fig. 1b). The enhancement of southerly wind in the Yangtze River valley is closely related to the beginning of the rainy season in North China (Ding, 1992; Zeng et al., 2012). In this study, the daily-mean southerly wind in the Yangtze River valley is used to denote the summer monsoon affecting North China from the southern boundary. The strong monsoon days are identified if more than 50% of the grid points in the Yangtze River valley have a daily-mean southerly wind component greater than 3 m s^-1 at 900 hPa, as in (Chen et al., 2013). There are 403 monsoon days (referred to as Monsoon(+)), with a frequency of 34.2% in the summer of 1998-2016 (Table 1). The remaining summer days are categorized as weak monsoon days and referred to as Monsoon(-). There are 667 days of Monsoon(-), with a frequency of 56.6%. Some days of synoptic disturbances with an obvious northerly wind are excluded.
We note that, under Monsoon(+), the conditional probability of BLO(+) is 23.3% (94/403), which is much higher than that of 13.8% under Monsoon(-) (92/667), as shown in Table 1. Thus, the large diurnal variations of low-level winds tend to occur on the strong monsoon days (Chen et al., 2013). In contrast, the conditional probability of MPS(+) under Monsoon(+) is 10.7% (43/403), which is comparable to that of 10.5% under Monsoon(-) (70/667), indicating that the occurrence of MPS(+) is insensitive to monsoon intensity. Thus, the large-scale monsoon condition may regulate the BLO intensity more than the MPS intensity. In section 4, we will compare these categories to clarify the change of regional forcings in regulating the DVP under different monsoon conditions.

3. Regional atmospheric circulation and DVP associated with the two regional forcings

3.1. Differences in atmospheric circulation under a strong MPS and BLO

-->

3.1. Differences in atmospheric circulation under a strong MPS and BLO

To depict the diurnal cycles of atmospheric circulations under MPS(+) and BLO(+), we construct composite 6-h deviations by subtracting the daily mean. Figure 4a shows that, at 1400 LST, the vertical circulation of the MPS is limited to near the foothills with sinking motion at ~115^°E, showing a relatively shallow structure. The rising motion is still weak over the mountains and indicates that the thermally driven local MPS is developing. At 2000 LST, the rising motion strengthens markedly and maximizes at 800-750 hPa (Fig. 4b). Meanwhile, the sinking motion extends to the entire plains, and an anomalous easterly wind is evident in the lower troposphere. It indicates that the MPS is fully developed due to the thermal contrast of large-scale topography at 2000 LST (Huang et al., 2010; Chen et al., 2013; Li et al., 2018). A meridional component of the MPS between the Yanshan Mountains and nearby plains also occurs at 2000 LST (not shown) and is superimposed on the latitudinal one, causing a strong sinking motion in the northern part of the North China plains.

Figure4. Longitude-vertical sections of anomalous zonal circulation (vectors: zonal wind, units: m s^-1; omega multiplied by 20, units: Pa s^-1) and vertical velocity with the daily mean removed (shaded; units: 10^-2 Pa s^-1) at 37.5^°N. The contours denote the original value of vertical velocity (units: 10^-2 Pa s^-1). (a-d) MPS(+); (e-h) BLO(+).

Figure 4c shows that the MPS is reversed at 0200 LST with downslope wind (westerly component and sinking motion) on the eastern slope (~114^°E). The upward branch of the MPS is well-established in the plains adjacent to mountain foothills (115^°-116^°E). At 0800 LST, the rising motion in the plains begins to weaken (Fig. 4d). In all four 6-h plots, the diurnal amplitudes (shaded) have a magnitude comparable to the actual vertical velocity (contours), indicating that the MPS plays a key role in regional circulation at the sub-daily time scale.
Under BLO(+), both the rising motion at mountains and the sinking motion at foothills at 2000 LST are weaker than those under MPS(+) (c.f., Figs. 4f and b). At 0200 LST, the rising motion at foothills due to the MPS reversal also weakens (c.f., Figs. 4g and c). However, the rising motion is extensive through the plains, with a maximum of more than -6× 10^-2 Pa s^-1 at ~116.5^°E (contours in Fig. 4g). It is much stronger than the climate mean of about -4× 10^-2 Pa s^-1 (contours in Fig. 2c). Therefore, we see a large diurnal variation of vertical motion in the plains associated with BLO(+), despite a relatively weak MPS.
To clarify the cause of regional differences in vertical motion, Fig. 5 shows the diurnal variations of horizontal winds and associated divergence. At 1400 LST, the deviations of northerly wind appear in plains, probably due to the frictional effect of turbulent mixing, and the upslope southeasterly winds develop near the Yanshan and Taihangshan Mountains (Fig. 5a). These deviations of wind induce an obvious horizontal divergence in the plains along mountain foothills. The upslope southeasterly winds continue to develop at 2000 LST and reach a magnitude of up to 3 m s^-1 (Fig. 5b). They induce strong rising motion of MPS(+) over mountains (Fig. 4b). The fully developed upslope winds also cause evident horizontal divergence in the plains, with the maximum divergence of ~200 km away from the mountains (Fig. 5b). Thus, such a divergence due to the well-established MPS causes the extensive sinking motion in the plains (Fig. 4b).

Figure5. Diurnal variations of the low-level winds (vectors; units: m s^-1) and horizontal divergence (shaded; units: s^-1) at 925 hPa with the daily mean removed. The contours denote the rising motion at 700 hPa (units: 10^-2 Pa s^-1). (a-d) MPS(+); (e-h) BLO(+).

The deviation of winds rotate to westerly at 0200 LST along-slope under MPS(+) (Fig. 5c). They converge with the southerly or southwesterly wind deviations in the plains, resulting in an along-slope convergence at the foothills. The strongest convergence appears in the northernmost part of the plains (36^°-38^°N, 114^°-118^°E), where the deviations of northerly winds from the Yanshan mountains, the westerly winds from the Taihangshan mountains and the southerly winds from southern plains are converging. At 0800 LST, the deviations of wind sustain horizontal convergence in the plains (Fig. 5d), which accounts for the rising motion, particularly along the mountain foothills (Fig. 4d). The low-level vorticity is also enhanced in the plains, likely favoring the development of convective vortices in the AM hours.
Under BLO(+), the deviation of winds exhibits a diurnal phase similar to those under MPS(+) (Figs. 5e and f). However, at 2000 LST they are characterized by the relatively strong easterly anomaly from ocean to land (Fig. 5f), resulting in the convergence over the inland plains. It indicates an establishment of the upward branch of anomalous land-sea breeze circulation in the plains (Huang et al., 2010). Thus, the BLO seems to induce rising motion early at 2000 LST, in contrast to the MPS, which induces sinking motion in the plains. The competing effect of different regional forcings is also seen over Central China plains (Yuan et al., 2012). At 0200 LST, the southwesterly winds in the plains are enhanced to ~2.5 m s^-1 (Fig. 5g). The induced convergence is evident over the North China plains, with a maximum in the northern part, which explains the extensive rising motion in the plains. Anomalous westerly winds occur at 0800 LST, with the largest amplitude in the southern part of North China and in the Yangtze River valley (Fig. 5h). They lead to strong convergence shifting southeastward to the coastal area near 120^°E. This suggests that the clockwise-rotating deviation of low-level winds may induce the change of horizontal convergence and rising motion at its downstream, which is shifted from the western mountains in the PM hours, to the North China plains late at night, and to the southeastern plains/coasts in the morning. In particular, the BLO induces an extensive rising motion in the plains, in contrast to that mainly confined to the foothills by the MPS. The deviation of low-level winds also exhibits similar clockwise rotation over Southeast China and induces the changing location of rising motion (Chen et al., 2013, Fig. 8).

2

3.2. DVP under MPS and BLO conditions

-->

3.2. DVP under MPS and BLO conditions

To clarify the impacts of the two regional forcings, we examine the spatial distributions of daily-mean rainfall and the diurnal variations of rainfall. On average, summer rainfall is mainly concentrated in the southeastern plains of North China (Fig. 6a). Figure 6d shows the west-east sections of hourly rain rate, considering that the steering-level winds in the mid-troposphere are mostly westerly. It shows that the rainfall with an afternoon peak at 1700 LST is extensive in the whole of North China, while the nocturnal rainfall at 2000-0800 LST appears in the plains more than 200 km from the mountains. The nocturnal rainfall exhibits an eastward-delayed feature, likely due to propagating rain systems, as also noted in previous studies (He and Zhang, 2010; Chen et al., 2014b, Chen et al., 2016; Sun et al., 2018). As a result, double rainfall peaks are evident over most areas of the North China plains (Yu et al., 2007; Zhou et al., 2008).

Figure6. Spatial distributions of daily-mean precipitation (shaded; units: mm) and 925-hPa horizontal winds (vectors; units: m s^-1) under the (a) climate mean, (b) MPS(+) and (c) BLO(+). The red dots (black hatching) in (a-c) denote the cumulative precipitation in the PM (AM) hours exceeding 55% of the daily mean. (d-f) Hovm?ller diagrams of rain rate averaged over 32.5^°-40^°N (shaded; units: mm h^-1) under (d) the climate mean (e) MPS(+) and (f) BLO(+). The sections are made based on the west-east distance relative to the mountains, which are denoted by the 700-m elevation in Fig. 1a.

Under MPS(+), daily precipitation has little change over the North China plains, except for an increase at its southern boundary (Fig. 6b). Such a pattern is probably due to the strong MPS tending to occur with strong thermal contrast under relatively dry conditions over North China, as shown later, while water vapor is mainly concentrated in the southern areas. Despite the lesser change in daily amount, the topographic differences in rainfall peaks between mountains and plains become more obvious. Figure 6e shows that, during 1700-2000 LST, the rainfall mainly occurs over the mountains and at the slope within 200 km from the mountains, where the upward branch of the daytime MPS is established. In contrast, the rainfall in the open plains (>400 km from mountains) is weaker than that of the climate mean, as the downward branch may inhibit convective growth. The rain systems redevelop from late night at foothills (~200 km from mountains) and propagate eastward, as a result of the reversed MPS. They produce morning rainfall over the plains within ~550 km from the mountains, as encircled in Fig. 6e. The MPS thus seems to regulate the local features of the DVP mostly adjacent to terrain (Zhang et al., 2019b). These characteristics correspond well to the nighttime upward branch of the reversed MPS and convergence zone that mainly occur over the foothill plains (Figs. 4c and 5c). Therefore, although the MPS does not increase daily rainfall obviously, it seems to effectively intensify the mountain-plain contrast of the DVP through redistributing the rainfall.
Figure 6c shows that, under BLO(+), the daily rainfall amount increases remarkably over the North China plains. The rainfall amount is largest in the central area of the plains (~37^°N), with a rain rate of ~10 mm d^-1, which is nearly double that of the climate mean. The rainfall has a broader area over the North China plains and a stronger intensity than that under MPS(+). BLO(+) thus plays a key role in the summer rainfall budget over North China. Figure 6f further shows that the rainfall systems develop markedly at the plains (200-400 km from mountains) from 2000 LST and then propagate eastward. They produce substantial rainfall during 0200-0800 LST, widespread over the open plains (300-800 km away from the mountains). The nocturnal rain rate in the plains is also much stronger than the afternoon one over the mountains, in contrast to that under the MPS with comparable rain rate over plains and mountains. Such extensive rainfall in the plains corresponds well to the regional horizontal convergence induced by the diurnally varying low-level winds (Fig. 5g).
Previous studies have shown that the rapid eastward shifting of rainfall phase in the Yangtze River valley is related to the inertial rotation of low-level winds (Chen et al., 2010, 2012a). Recent studies have pointed out that the BLO may be the main mechanism producing the nighttime mei-yu rainfall over the Yangtze River valley, while the MPS only makes a small contribution, at least in several cases of rainstorms (Chen et al., 2017; Xue et al., 2018). The nocturnal LLJ due to the BLO even contributes the most to the moisture convergence and precipitation over the Sichuan Basin, where the effect of the local MPS is secondary (Zhang et al., 2019b). This study indicates that the BLO also plays a leading role in regulating the DVP of North China, more so than the MPS.

2

3.3. Diurnal variations of the water vapor transport and moisture content

-->

3.3. Diurnal variations of the water vapor transport and moisture content

The water vapor transport and convergence are closely associated with precipitation activities (e.g., Trenberth, 1999; Pu and Dickinson, 2014). In this subsection, we examine the roles of their diurnal variations in regulating the DVP under the two regional forcings. Because the diurnal variations of both water vapor and winds are most evident in the lower troposphere, we calculate the vertically integrated flux of water vapor below 700 hPa (Q) and its horizontal divergence (div) using the following formula (Bluestein, 1993): \begin{equation*} {Q}=\int_{700}^{\rm surface}{v}q{\rm d}p,\quad {\rm div}=\frac{\partial {Q}_{v}}{\partial y}+\frac{\partial {Q}_{u}}{\partial x}-\frac{{Q}_{v}}{a}\tan(\varphi) . \end{equation*} in which the v, Q_v, Q_u, p, a, and φ are wind vector, meridional water vapor transport, zonal water vapor transport, air pressure, the radius of the earth and latitude in units of radian.
Figure 7 shows that the diurnal variations of water vapor transport and convergence are highly analogous to those of wind vector and convergence in Fig. 5. Under MPS(+), the water vapor flux converges to the mountains and diverges over the plains during daytime, due to the strong upslope winds (Figs. 7a and b). Such an enhanced convergence of water vapor flux may support the rainfall systems developing over the mountains and adjacent areas in the PM hours. As the MPS reverses at late night, the water vapor flux becomes convergent near foothills (Fig. 7c). The intensity of flux convergence increases by 100%-200% under MPS(+) relative to MPS(-) (not shown). It corresponds well to the rainfall systems that develop and intensify at the foothills at night (Fig. 6e). It seems that the diurnal variations of low-level winds are the key process controlling the water vapor transportation, convergence and even precipitation, possibly because the wind diurnal amplitude is much greater than that of water vapor content (Chen et al., 2013).

Figure7. Diurnal variations of vertically integrated (700 hPa to surface) water vapor flux with the daily mean removed (vectors, units: kg m^-1 s^-1), water vaper flux convergence (contours; units: 10^-3 kg m² s^-1), and subsequent 6-h rainfall (shaded; units: mm, e.g., 1400 LST shows the cumulative precipitation from 1400 to 1900 LST). (a-d) MPS(+); (e-h) BLO(+).

Under BLO(+), the convergence of water vapor flux to the mountains of North China is weaker than that under MPS(+) at 2000 LST (Fig. 7f). Instead, some features of moisture convergence appear in the plains, favoring the development of rain systems at 2000-2300 LST. This can be attributed to the anomalous easterly flow from the ocean that is convergent to the inland plains (Fig. 5f). At 0200 LST, a large amount of warm and humid air is advected from the Yangtze River valley to the North China plains by the southwesterly monsoon (Fig. 7g). It leads to the enhancement of flux convergence by approximately -0.5× 10^-3 kg m² s^-1 in the plains. As a result, a large amount of rainfall is widespread in the plains, with the maxima exceeding 4 mm in 6 h. These values under BLO(+) are around double those under MPS(+), suggesting that the BLO may play a leading role in the regulation of moisture conditions and precipitation at night.
To clarify the regulation of the regional forcings on humidity, we estimate the diurnal variation of region-averaged specific humidity. Figure 8 shows that the lower-tropospheric humidity over the mountains increases by ~0.2 g kg^-1 from 1400 LST to 2000 LST under MPS(+), compared to ~0.1 g kg^-1 under MPS(-). The strong upslope winds under MPS(+) seem to increase the humidity over mountains. As for the plains, the humidity decreases slightly by 0.1-0.2 g kg^-1 from 2000 LST to 0200-0800 LST under MPS(+), probably due to the condensation with radiative cooling at night. In contrast, it decreases by ~0.3 g kg^-1 through the night under MPS(-), likely due to the weak convergence of water vapor flux. Nevertheless, the daily humidity under MPS(+) is smaller than the climate mean (Fig. 8a). The moisture deficit is not good for rainfall systems, and thus the daily rainfall amount does not increase much under MPS(+), as shown in Fig. 6b.

Figure8. The 6-h variations of specific humidity (units: g kg^-1) averaged from surface to 700 hPa under (a) strong and (b) weak regional forcings. Bars and numbers in red (black) represent the values over the mountains (plains).

Figure 8 also shows that the daily-mean specific humidity under BLO(+) is ~1.2 g kg^-1 higher than that under BLO(-) over both mountains and plains. This is because BLO(+) tends to occur on Monsoon(+) days (Table 1). Under BLO(+), the high humidity of ~10.8 g kg^-1 is sustained in the plains throughout day and night. The moist condition along with enhanced convergence at night is favorable for the growth of organized convection in the plains (He and Zhang, 2010; Chen et al., 2013; Yuan et al., 2015). In stark contrast, the humidity decreases from ~10.0 g kg^-1 at 2000 LST to ~9.4 g kg^-1 at 0200-0800 LST under BLO(-). Such suppressed conditions of moisture and convergence are thus thought to be unfavorable for nocturnal convection in the plains.

4. Impacts of regional forcings on precipitation under different monsoon conditions

4.1. Characteristics of regional forcings on strong monsoon days

-->

4.1. Characteristics of regional forcings on strong monsoon days

In this section, we examine the influence of the regional forcings under strong monsoon conditions. Here, we measure their change using the increase in rainfall and moisture convergence under BLO(+)$\vert$Monsoon(+) and MPS(+)$\vert$Monsoon(+), compared to BLO(+) and MPS(+). Figures 9a-d show that, under a strong MPS with a strong monsoon (i.e., MPS(+)$\vert$Monsoon(+)), the diurnal variation of vertical motion (shaded) in the vicinity of the mountain slopes (112^°-116^°E) is analogous to that in Figs. 4a-d, suggesting little change in local MPS. In the plains east of 116^°E, the rising motion during 0200-0800 LST is enhanced in both diurnal deviation (shaded) and actual values (contours), as shown in Figs. 9c and d. The rising motion of -0.06 Pa s^-1 or stronger is widespread in the plains and even extends to the mid-troposphere under MPS(+)$\vert$Monsoon(+), compared to that in Figs. 4c and d. This strong rising motion at 0200-0800 LST is attributable to the enhanced convergence in the plains, although the convergence near the mountain slopes changes less (Figs. 10a-d). The rising motion is strongest at the southern part of the North China plains, i.e., at the terminus of enhanced nocturnal southerly monsoon flow from the Yangtze River valley (Figs. 10c and d). Figure 11a shows that the diurnal wind variations in the plains are southeasterly at 2000 LST and southwesterly at 0200 LST, with a large amplitude under MPS(+)$\vert$Monsoon(+). Therefore, wind diurnal variations under a combined influence of the MPS and monsoon flow result in the enhanced low-level convergence and ascent over the plains at night.

Figure9. As in Fig. 4 but under (a-d) MPS(+)$\vert$Monsoon(+) and (e-h) BLO(+)$\vert$Monsoon(+).

Figures 9e-h show that, under a strong BLO with a strong monsoon (i.e., BLO(+)$\vert$Monsoon(+)), the vertical sections of regional circulation exhibit highly similar patterns to those in Figs. 4e-h. The magnitude of rising motion (contours) over the plains at 0200-0800 LST becomes double that under BLO(+), partly due to the enhanced diurnal amplitude (shaded), as shown in Figs. 9g and h. Figures 10e-h show that the 700-hPa rising motion (contours) is widespread over the North China plains throughout the day and night. It reaches a peak of up to -0.1 Pa s^-1 at 0200 LST because of a strong convergence in the plains, as shown in Fig. 10g, compared to that of -0.06 Pa s^-1 under BLO(+) (Fig. 5g). Strong rising motion over the plains is sustained until 0800 LST (Fig. 10h) and corresponds to the enhanced diurnal variations of low-level winds under BLO(+)$\vert$Monsoon(+), with the largest amplitude at 0800 LST (Fig. 11b). Meanwhile, the rising motion under BLO(+)$\vert$Monsoon(+) is displaced to the center of the North China plains, compared to that under MPS(+)$\vert$Monsoon(+) (c.f., Figs. 10g, h and 10c, d). These results are consistent with previous studies in that a strong monsoon may increase both the daily mean and diurnal variation of low-level convergence over eastern China (Chen et al., 2013).

Figure10. As in Fig. 5 but under (a-d) MPS(+)$\vert$Monsoon(+) and (e-h) BLO(+)$\vert$Monsoon(+).

Figure11. Diurnal variations of the 925-hPa wind deviations (units: m s^-1) with the daily mean removed under (a) MPS(+) over the North China plains and (b) BLO(+) over the Yangtze River valley. Grey circles enclose the points at the same hour.

The increasing amplitude of the BLO with monsoon intensity may be explained physically as follows. It is well-recognized that inertial oscillation is closely related to the frictional drag of turbulent mixing in the boundary layer (Blackadar, 1957). A large background wind speed is helpful to exerting a frictional effect (Shapiro et al., 2016), and a strong monsoon with relatively warm and windy conditions may also favor the occurrence of BLO(+) (Chen et al., 2013). As a result, the amplitude of inertial oscillation is proportional to the geostrophic wind speed or large-scale mean flow (Xue et al., 2018). In this study, we also found that BLO(+) days tend to occur in strong monsoon flow (Table 1), with the wind amplitude much larger than that under other conditions (Fig. 11b). This indicates that the regional forcings, particularly the BLO, may work with the strong monsoon flow to enlarge diurnal variations of low-level winds, strengthening the upward motion over the North China plains in the AM hours.

2

4.2. Combined influence of monsoon and regional forcings on DVP

-->

4.2. Combined influence of monsoon and regional forcings on DVP

Figures 12a and d show the spatial pattern of rainfall and its diurnal variation under Monsoon(+). The daily rainfall amount increases remarkably over the plains to ~13 mm d^-1 (Fig. 12a), compared to the summer mean of ~7 mm d^-1 (Fig. 6a). The increase in monsoon rainfall mainly occurs during 2000-0800 LST (Fig. 12d). Monsoon activity thus produces a large amount of summer precipitation in the North China plains, particularly at night, which is similar to the situation in Central China (Chen et al., 2013). It is related to the enhanced moisture transport/convergence by strong monsoon at night (Figs. 13a-d).

Figure12. As in Fig. 6 but under (a, d) Monsoon(+), (b, e) MPS(+)$\vert$Monsoon(+) and (c, f) BLO(+)$\vert$Monsoon(+).

Figure13. As in Fig. 7 but under (a-d) Monsoon(+), (e-h) MPS(+)$\vert$Monsoon(+) and (i-l) BLO(+)$\vert$Monsoon(+). Note that the color bar is changed.

Figure 12b shows that, under MPS(+)$\vert$Monsoon(+), the daily rainfall amount in the southern part of the North China plains increases to a maximum of ~20 mm d-1, while the spatial pattern of rainfall is analogous to that under MPS(+) (Fig. 6b). Figure 12e shows that the rain rate at the slope is enhanced to ~0.50 mm h^-1 at 1700-2300 LST, which is around double that under MPS(+) (Fig. 6e). Afternoon-evening rainfall is relatively suppressed in the open plains, 400 km away from the mountains, due to an enhanced divergence of the MPS (Fig. 12e), compared to that in Fig. 12d. Figure 12e also shows that rainfall systems in the plains at 300 km away from the mountains develop after midnight and propagate eastward. They produce morning rainfall of up to 0.70 mm\;h^-1 for the plains 400-600 km away from mountains. These rainfall features are closely related to the nighttime enhancement of wind convergence in the plains (Fig. 10c). The diurnal phase of rainfall is highly similar to that under MPS(+), despite the increasing rain intensity (c.f., Figs. 12e and 6e). Therefore, under MPS(+)$\vert$Monsoon(+), the active monsoon flow mainly intensifies the precipitation and its diurnal amplitude, while the strong MPS still largely regulates the diurnal phase pattern.
Figure 12c shows that, under BLO(+)$\vert$Monsoon(+), daily precipitation increases markedly in the plains up to 15 mm d-1, compared to that of ~10 mm d-1 under BLO(+) (Fig. 6c). The large rainfall is also displaced more northward, to the center of the North China plains, than that under Monsoon(+) (c.f., Figs. 12c and a). Figure 12f further shows that a large portion of rainfall in the plains occurs during 2000-0800 LST, with the maximum rain rate increasing from ~0.45 mm h^-1 (Fig. 6f) to ~0.8 mm h^-1. The increased nocturnal rainfall is strongly linked to the enhanced BLO amplitude and low-level convergence under a strong monsoon (Figs. 10f-h and 11b). These results suggest that the regional BLO may couple with large-scale monsoon flow to intensify both the daily mean and diurnal amplitude of rainfall, particularly through increasing the nocturnal rain rate in the plains. The nocturnal rainfall under BLO(+)$\vert$Monsoon(+) also has a longer duration and is more widespread than that under MPS(+)$\vert$Monsoon(+) (c.f., Figs. 12f and e). Thus, the strong BLO under a strong monsoon becomes more effective in regulating the regional precipitation than the MPS.

2

4.3. Diurnal variations of water vapor regulated jointly by monsoon and regional forcings

-->

4.3. Diurnal variations of water vapor regulated jointly by monsoon and regional forcings

To further clarify the processes producing the DVP features revealed in section 4.2, we examine the diurnal variations of low-level water vaper under strong monsoon conditions. Figures 13a-d show that under Monsoon(+) the water vapor flux convergence is prevalent in most of the plains throughout the day and night. It becomes strongest at 0200 LST with a magnitude of up to -1.5× 10^-3 kg m² s^-1, which is double of that at 2000 LST. Figures 14a and b also show that the low-level specific humidity in the plains increases from ~10.0 g kg^-1 under Monsoon(-) to ~11.5 g kg^-1 under Monsoon(+). The humidity decreases slightly at 0200-0800 LST in the plains (Fig. 14a) due to the enhanced moisture sink (i.e., precipitation minus evaporation) associated with the water vapor flux convergence (Chen et al., 2013). Such a high humidity and an enhanced convergence correspond to the widespread rainfall in the plains during 2000-0800 LST (Fig. 12d). In contrast, over the mountains, the water vapor flux is divergent at 0200-0800 LST (Figs. 13c and d) and the specific humidity decreases from 11.1 g kg^-1 at 2000 LST to 10.4 g kg^-1 at 0800 LST (Fig. 14a), resulting in the suppressed rainfall in the morning hours.

Figure14. As in Fig. 8 but under (a) Monsoon(+) and (b) Monsoon(-) with strong regional forcings.

Figures 13e-h show that, under MPS(+)$\vert$Monsoon(+), the water vapor flux is convergent throughout the day in the southern part of the plains (to the south of 34^°N) as a result of strong daily monsoon flow. In the northern part of the plains, the water vapor flux is mainly divergent at 1400-2000 LST (Figs. 13e and f) due to the downward branch of the MPS (Figs. 9a and b). This feature is different from that under Monsoon(+) (c.f., Figs. 13e, f and 13a, b). It indicates that the regional-scale MPS tends to offset the effect of large-scale monsoon flow in terms of moisture convergence during the daytime. At 0200 LST, water vapor flux convergence strengthens significantly in the plains, with the magnitude up to -1.5× 10^-3 kg m² s^-1 (Fig. 13g). This convergence is more evident than that under MPS(+) in terms of intensity and area (c.f., Figs. 13g and 7c). The enhanced convergence can last until 0800 LST (Fig. 13h). Figure 14a shows that specific humidity in the plains is sustained as high as ~11.1 g kg^-1, which is 10% higher than that under MPS(+) (Fig. 8a) and 15% higher than that under MPS(+)$\vert$Monsoon(-) (Fig. 14b). Such enhanced moisture content and convergence due to the strong monsoon combining with the MPS explains the nocturnal rainfall in the plains (Figs. 12e and 13g) being more intense than that under MPS(+) (Figs. 6e and 7c).
Figures 13i and j show that under BLO(+)$\vert$Monsoon(+) the water vapor flux convergence is evident in most areas of the plains at 1400-2000 LST. The spatial pattern is similar to that under Monsoon(+) (Figs. 13a and b) but different from that under MPS(+)$\vert$Monsoon(+) (Figs. 13e and f). In particular, the flux convergence is up to -1.0× 10^-3 kg m² s^-1 in the plains to the north of 34^°N at 2000 LST (Fig. 13j), which is much stronger than that in Figs. 13b and f. Such enhanced convergence is due to the strong monsoon flow and deviation of easterly wind at 2000 LST that are mainly convergent in the plains rather than in the mountains (Fig. 13j). The monsoon and BLO seem to work together to enhance the water vapor flux convergence in the plains early at 2000 LST. This feature is clearly different from the situation in which the MPS and monsoon offset each other (Fig. 13f). The enhanced flux convergence at 2000 LST explains well the precipitation growth early at 2000 LST being widespread in the plains in Figs. 12f and 13j, in stark contrast to that in Figs. 12e and 13f. The flux convergence becomes strongest around 0200 LST in the plains to the north of 34^°N (Fig. 13k), with an intensity much stronger than those under all other conditions (Figs. 13c, g and 7c, g). As a result of the enhanced convergence and humidity at 2000-0800 LST (Fig. 14a), the intense precipitation is widespread in the plains and has a relatively long duration, as shown in Fig. 12f and Figs. 13j-l. Therefore, the combination of the monsoon and BLO (i.e., BLO(+)$\vert$Monsoon(+)) may be most effective in regulating the diurnal cycle of rainfall over the North China plains.
We may further estimate the relative importance of the monsoon and regional forcings through comparing the daily mean and diurnal amplitude of water vapor flux convergence in the plains. The daily mean of flux convergence by monsoon flow is approximately -0.66× 10^-3 kg m² s^-1 (Figs. 13a-d), while the diurnal amplitudes by the BLO and MPS are approximately -0.54× 10^-3 kg m² s^-1 and -0.42× 10^-3 kg m² s^-1, respectively (Fig. 7). This suggests that the influence of the regional forcings can be comparable to that of the monsoon flow. For both the daily mean and the diurnal deviation at 0200 LST, flux convergence increases markedly under a strong monsoon and the regional forcings (Figs. 13e-l). The daily-mean enhancement of the flux convergence by monsoon flow is attributable to both large-scale convergence and increased humidity (Fig. 14). The diurnal enhancement by regional forcings is mainly due to the horizontal convergence of diurnally varying low-level winds, whereas the diurnal variation in humidity is quite small (Fig. 14). Furthermore, the flux convergence by the BLO is larger than that by the MPS, particularly at 2000-0200 LST to the north of 34^°N (c.f., Figs. 13f, g and 13j, k), suggesting a relatively large importance of the BLO.

5. Summary and discussion

Using a 19-yr archive of satellite rainfall and reanalysis data, we have classified two regional forcings (the MPS and BLO) into different intensities and examined their associated diurnal variations of atmospheric conditions and precipitation over the North China plains. We further investigated their influence under strong monsoon conditions. The major findings can be summarized as follows:
(1) Both the MPS and BLO strongly regulate the diurnal variation of rainfall over the North China plains, with increased nocturnal rainfall, but they exhibit some obvious differences. Under a strong MPS, the nocturnal rainfall in the plains exhibits a topographic contrast to the afternoon rainfall over the mountains, despite the less change in daily rainfall amount compared to the climate mean. Rainfall systems mainly develop at the foothills after midnight and propagate eastward, in an association with the upward branch of the nighttime MPS. They produce early-morning rainfall over the adjacent plains within ~550 km from the mountains. As a comparison, under a strong BLO, rainfall systems over the plains strengthen from early evening, producing stronger and more extensive morning rainfall over the areas up to ~800 km away from the mountains. The nocturnal rainfall is produced by the enhancement of water vapor flux convergence by the regional forcings at night, though the BLO may induce moisture convergence with a wider area and some hours earlier than that of the MPS. It is mainly caused by the convergence of the diurnally varying low-level winds, whereas the diurnal variation in humidity is quite small. In the evening, the BLO and MPS have a contrasting effect because of the convergence of the BLO-related easterly from the ocean to the plains and the divergence of the MPS-related upslope winds in the plains. From late night to morning, both the reversed nighttime MPS and the BLO-related southerly anomaly with relatively high humidity facilitate the moisture convergence over the plains.
(2) On strong monsoon days, the daily rainfall amount increases remarkably in the North China plains. It results from the water vapor flux convergence and high humidity that is enhanced by the southerly monsoon flow. Both the daily mean and diurnal amplitude of rainfall in the plains are strengthened under the combined influence of the monsoon and MPS/BLO, though there is less change in the spatial pattern and propagation features of rainfall. In particular, daily rainfall increases from ~10 mm d^-1 under the BLO to ~15 mm d^-1 under a strong monsoon with the BLO. The increase in nocturnal rainfall collocates well with the enhancement of water vapor flux convergence in the plains. The relative importance of monsoon flow and regional forcings in regulating rainfall can be measured in terms of the daily mean and diurnal amplitude of flux convergence. The diurnal amplitudes of flux convergence by the BLO and MPS have a magnitude comparable to the daily mean by the monsoon flow. Under a strong monsoon and the BLO (MPS), flux convergence is evident (still suppressed) at 2000 LST in the open plains, which explains the growth of nocturnal rainfall early in the evening (delayed to after midnight). Thus, the BLO seems more effective than the MPS in regulating the regional climate under monsoon conditions.
Many previous studies have focused on the summer mean of the DVP and related regional circulations. The DVP process actually regulated by both MPS and BLO, though the effect of the MPS has been emphasized (e.g., He and Zhang, 2010; Bao et al., 2011; Bao and Zhang, 2013). Our study confirms such a strong connection of the regional forcings with North China precipitation. We further examine two regional forcings (the BLO and MPS) in a more specific manner and estimate their relative importance, which has been less well studied. The BLO is shown to have a larger overall impact on the precipitation over the North China plains than the MPS, which is consistent with some recent studies (e.g., Chen et al., 2017; Xue et al., 2018; Zhang et al., 2019b).
In this study, we have carried out a composite analysis on the influence of the monsoon and regional forcings. As monsoon activity exhibits a large variability from intraseasonal to interdecadal time scales (e.g., Ding, 1992; Ding and Chan, 2005), we expect that it may regulate the intensity of regional forcings through large-scale processes. Previous studies have noted that the diurnal variations of rainfall over eastern China have pronounced seasonal change, interannual variation and an interdecadal trend, in a close association with monsoon activity (Yuan et al., 2013a; Chen et al., 2014a, Chen et al., 2018). Further studies on the long-term variation of regional forcings under monsoon conditions may provide additional insights into regional climate change. Meanwhile, we note that the regional forcings and their influence under monsoon conditions strongly control the preferred location and timing of rainfall. A better representation of them, as well as related convection processes, in models may help to improve the simulation of the rainfall diurnal cycle with evident regionality over China (Yuan et al., 2013b; Du and Chen, 2018; Zhang et al., 2019a, b). Therefore, the detailed roles of the monsoon and regional forcings in the occurrence of extreme weather deserve further observation analysis and numerical study.

本站小编 Free考研考试/2022-01-02

Corresponding author: Guixing CHEN,chenguixing@mail.sysu.edu.cn

HTML

2.1. Satellite-derived rainfall and atmospheric reanalysis data

2.2. Definition of the intensities of the MPS, BLO and monsoon affecting North China

3.1. Differences in atmospheric circulation under a strong MPS and BLO

3.2. DVP under MPS and BLO conditions

3.3. Diurnal variations of the water vapor transport and moisture content

4.1. Characteristics of regional forcings on strong monsoon days

4.2. Combined influence of monsoon and regional forcings on DVP

4.3. Diurnal variations of water vapor regulated jointly by monsoon and regional forcings

相关话题/Diurnal Variations Precipitation

领限时大额优惠券,享本站正版考研考试资料!

Differences between Convective and Stratiform Precipitation Budgets in a Torrential Rainfall Event

Seasonal Variations of Observed Raindrop Size Distribution in East China

Warm-Season Diurnal Variations of Total, Stratiform, Convective, and Extreme Hourly Precipitation ov

Role of the Nocturnal Low-level Jet in the Formation of the Morning Precipitation Peak over the Dabi

Variation in Principal Modes of Midsummer Precipitation over Northeast China and Its Associated Atmo

A 31-year Global Diurnal Sea Surface Temperature Dataset Created by an Ocean Mixed-Layer Model

Temporal and Spatial Variations in the Climate Controls of Vegetation Dynamics on the Tibetan Platea

Determination of Intraseasonal Variation of Precipitation Microphysics in the Southern Indian Ocean

Subseasonal Change in the Seesaw Pattern of Precipitation between the Yangtze River Basin and the Tr

Changes in the Proportion of Precipitation Occurring as Rain in Northern Canada during Spring-Summer