1. Introduction
The Asian monsoon is one of the most energetic components of the global atmospheric circulation systems. The large-scale component of this multiscale circulation system is driven by seasonally varying sensible and latent heat (e.g., Murakami and Ding 1982; Luo and Yanai 1983, 1984; He et al. 1987), with the Tibetan Plateau playing an important role as an elevated heat source (Chen et al. 1985; Li and Yanai 1996; Ye and Gao 1979; Wu and Zhang 1998). Previous studies have shown that, on average, the summer monsoon onset first occurs in the eastern Bay of Bengal (BoB) and the Indochina Peninsula in the second to third pentad of May, which is followed by the monsoon onset over the South China Sea in late May and the Indian summer monsoon onset in early June (e.g., Webster et al. 1998; Wu and Zhang 1998; Qian and Yang 2000; Wang and LinHo 2002). Based on global model simulations with and without mountains, Hahn and Manabe (1975) showed that the large-scale topography enhances the latent heating over Tibet and helps to maintain the South Asian low, which contributes to the rapid development of the monsoon circulation. Li and Yanai (1996) found that the onset of the Asian summer monsoon is concurrent with the reversal of the meridional temperature gradient in the upper troposphere south of the Tibetan Plateau. This reversal results from the large temperature increase over Eurasia due to the sensible heating over the plateau region in spring, with no appreciable temperature change in the Indian Ocean. In a case study of the Asian monsoon onset in 1989, Wu and Zhang (1998) suggested that the earliest monsoon onset in the BoB is linked to the thermal and mechanical forcing of the Tibetan Plateau. Based on experiments using a simplified regional climate model, Sato and Kimura (2007) suggested that the Tibetan Plateau induces midtropospheric subsidence over northern India through the thermal and mechanical effects and that the gradual weakening and retreat of the descent before July is consistent with the northward migration of the monsoon rainfall.
At the regional scale, much of the Asian monsoon domain, especially South and Southeast Asia, is characterized by complex local terrain. Although it has been well known that mountains can induce heavy orographic precipitation, the importance of the regional topography to the large-scale monsoon circulation was not recognized until recently. Using the high-resolution Tropical Rainfall Measuring Mission (TRMM) precipitation and Quick Scatterometer (QuikSCAT) surface wind data, Chang et al. (2005) showed that heavy monsoon rainfall does not follow the seasonal evolution of the large-scale monsoon circulation over the Asia–Australia land bridge. Instead, it is anchored on the windward side of the high terrain, and the annual cycle of the monsoon rainfall over Southeast Asia is thus the result of interaction between the seasonal reversing large-scale monsoon circulation and the regional terrain. This was further confirmed by Xie et al. (2006). In their regional climate model simulation, orographic rainbands, which were not adequately resolved due to the coarse model resolution, were mimicked by a prescribed heat source and were shown to have far-reaching impacts on the large-scale monsoon circulation.
Flow over mountains and orographic precipitation have been investigated in many previous studies (e.g., Smith 1980, 1989; Ogura and Yoshizaki 1988). With a large moist Froude number or small nondimensionalized mountain height, flow tends to ascend over a mountain, and the upslope lifting may trigger or enhance precipitation.1 A small Froude number (or a large nondimensionalized mountain height) is associated with large mountains, strong static stability or weak inflow speed, and upstream blocking may take place with moderate lifting. Precipitation, however, can be enhanced by the so-called feeder–seeder mechanism, by which raindrops from a passing precipitating system above collect additional moisture as they fall through the low-level orographic cap clouds and thus increase precipitation (e.g., Bergeron 1960). Precipitation is not only enhanced in the immediate vicinity of mountains. Riehl (1979) indicated the presence of upstream deep convective towers well off the Western Ghats during the monsoon season. Grossman and Durran (1984) suggested that deep convection offshore was due to gentle lifting of potentially unstable air associated with upstream blocking, while deep convection is inhibited near the mountain crests by strong vertical wind shear and large-scale subsidence aloft. Chu and Lin (2000) examined the interaction of a conditionally unstable flow with a two-dimensional mesoscale mountain. They identified three flow regimes based on the orographic convective systems: upstream propagating convective systems for small moist Froude number, quasi-stationary convective systems for moderate moist Froude number, and both quasi-stationary- and downstream propagating convective systems for large moist Froude number. Chen and Lin (2005) further studied the effects of the convective available potential energy (CAPE) on a moist flow over a three-dimensional mesoscale mountain and found that relatively strong mean flow produces a quasi-stationary mesoscale convective system (MCS) and maximum rainfall on the windward slope, instead of on the mountain peak or over the lee side as suggested by two-dimensional simulations. These studies have improved our understanding of the mechanisms of orographic precipitation. However, most of them focused on mesoscale precipitating systems, and the interaction between orographic precipitation and the large-scale monsoon circulation remains unclear.
The objective of this study is to examine the interaction between the large-scale monsoon circulation and orographic precipitation over South and Southeast Asia and its role in the Asian summer monsoon onset. The summer of 2000 was chosen for regional climate model simulations as a typical monsoon season in terms of the seasonal progression of the monsoon. The model description and experimental design are presented in section 2. Section 3 provides a brief overview of the observed seasonal march of the monsoon in summer 2000. The numerical model simulations and the dynamical mechanisms are discussed in sections 4 and 5, followed by a summary in section 6.
2. Model description
The numerical model used in this study is version 3 of the regional climate model (RCM) developed by the National Center for Atmospheric Research (RegCM3) (e.g., Giorgi et al. 1999). The dynamical core of the RegCM3 is essentially equivalent to the hydrostatic version of the fifth-generation Pennsylvania State University–National Center for Atmospheric Research Mesoscale Model (MM5) (Grell et al. 1994). It is a primitive equation model with a terrain-following σ vertical coordinate. Surface processes are represented via the Biosphere–Atmosphere Transfer Scheme (BATS) (Dickinson et al. 1993). An explicit planetary boundary layer is formulated following the nonlocal vertical diffusion scheme of Holtslag et al. (1990). A detailed atmospheric radiative calculation package (Briegleb 1992) is adopted for climate studies. Resolvable scale precipitation is represented via the scheme of Pal et al. (2000), which includes a prognostic equation for cloud water and allows for fractional grid box cloudiness, accretion, and reevaporation of falling precipitation. Convective precipitation is represented using the Emanuel scheme (Emanuel 1994). RegCM3 has been successfully applied to different climate regimes (e.g., Kato et al. 1999; Small et al. 1999). In particular, Sun et al. (1999a,b) tested the model over East Africa, where the seasonally varying intertropical convergence zone (ITCZ) superimposes on the complex topography, and it was shown that the model reproduced the spatial patterns and seasonal variations of monsoon rainfall reasonably well.
The RegCM3 is driven by sea surface temperature forcing interpolated from the weekly mean SST data from the NOAA Optimum Interpolation (OI) SST V2, and the lateral boundary forcing and initial conditions are derived from the six-hourly 40-yr European Centre for Medium-Range Weather Forecasts Re-Analysis (ERA-40). The model domain is 15°S–25°N, 40°–130°E, which covers the South and Southeast Asian monsoon regions. The cross-equatorial flow over the eastern Indian Ocean and the Somali jet over the western Indian Ocean are simulated by the model. The model has 17 vertical levels and the horizontal resolution of the model is 30 km, which can resolve the regional features of the terrain reasonably well (see Fig. 3a).
Several experiments are designed to explore the impacts of wind–terrain interaction on monsoon onset, precipitation distribution, and large-scale circulation. In all experiments the model was integrated from 1 April to 30 June, a time period covering monsoon onsets from Southeast Asia to India. The simulation before 20 April was excluded as the model’s spin up or initial adjustment period. In the control run, the global digital elevation model GTOPO30 dataset from the U.S. Geological Survey (USGS) is used to specify the model topography. In the sensitivity tests, the high terrain is modified or removed while other land surface conditions remain the same. Elevation changes are usually accompanied by changes in land surface conditions, such as vegetation, albedo, and soil moisture. In this study we are interested in the mechanical forcing of high terrain and the impacts of orographic precipitation. For simplicity, the static land surface conditions are kept unchanged in all experiments, except that the removal of topography changes the land surface temperature and the associated surface heat fluxes.
There are various definitions for monsoon onset (e.g., Fasullo and Webster 2003; Goswami et al. 2006; Wang et al. 2009). In this study, we are interested in the evolution of precipitation, so the monsoon onset is defined as the commencement of persistent precipitation, with daily mean precipitation more than 0.2 mm h−1 in TRMM or 0.4 mm h−1 in RegCM model simulations, for at least two pentads. A higher threshold is used for RegCM simulations because the model overpredicts precipitation (see section 4).
3. Distribution of monsoon precipitation and wind–terrain interaction
In this section we briefly describe the observed seasonal progression of the summer monsoon over South and Southeast Asia in year 2000. Figure 1 shows the pentad mean precipitation derived from TRMM 3B42 (Simpson et al. 1996) and the surface wind from QuikSCAT level-3 data. Both data have a spatial resolution of 0.25° × 0.25°.
Pentad mean TRMM precipitation (shading, mm h−1) and QuikSCAT surface wind (m s−1) for the year 2000.
Citation: Journal of Climate 25, 7; 10.1175/JCLI-D-11-00136.1
The earliest monsoon onset, marked by a rapid increase in precipitation amount and the establishment of the low-level southwesterly flow, occurred over the Indochina Peninsula in mid-May. The commencement of convection, however, occurred in April before monsoon onset. Wen and Zhang (2007) suggested that premonsoon precipitation and monsoon onset are associated with different background circulations and mechanisms. The premonsoon convection may be triggered by the tropical quasi-biweekly oscillation propagating from the south and cold surges invading from the north. The monsoon onset, as shown in Fig. 1, was preceded by a pair of cyclonic circulations straddling over the equator during 25 April–9 May. The circulation in the Southern Hemisphere was a well-defined vortex. Its Northern Hemisphere counterpart was not a closed cyclone but had strong cyclonic curvature south of the Indian subcontinent. The structure of the cyclonic circulation is similar to the “onset vortex” identified by Ding and Liu (2001). The cyclonic circulation over the northern Indian Ocean was associated with a southwesterly flow over the BoB and a westerly flow over the eastern equatorial Indian Ocean. The former transferred moisture to the Indochina Peninsula, while the later fueled convection over Sumatra and Borneo. Also shown in the figure is Typhoon Damrey with significant precipitation over the Philippine Sea during 5–9 May. Moisture convergence and precipitation over Southeast Asia, however, were reduced in the same pentad. During 10–14 May a cross-equatorial flow began to develop over the eastern Indian Ocean (80°–90°E), and the westerly flow advanced northward and prevailed from the BoB to the South China Sea after Typhoon Damrey drifted northwestward. Precipitation persisted over the Indochina Peninsula and expanded both westward to the eastern BoB and eastward to the South China Sea, which led to the monsoon onset in those regions.
The Indian monsoon onset occurred in the last pentad of May, characterized by the prevailing westerly flow over the Arabian Sea and an abrupt increase of precipitation over the west coast of India. This is consistent with the onset date, 29 May, determined by the hydrological onset and withdraw index (HOWI) (Fasullo and Webster 2003). Precipitation off the west coast of the Indian subcontinent peaked during 4–8 June and decreased in the following pentad as precipitation over the eastern BoB, the Indochina Peninsula, and the South China Sea increased in mid-June (the pentad 14–18 June not shown). The convection center thus appeared to shift eastward. This onset sequence is consistent with that depicted by Wang and Linho (2002) and Ding and Chan (2005). An interesting feature shown in Fig. 1 is that heavy precipitation tends to occur on the windward side of the high terrain, as pointed out by Chang et al. (2005, 2006).
The seasonal march of the convection center is clearly illustrated in the Hovmöller diagram of precipitation in Fig. 2. Monsoon onset first occurred over the Indochina Peninsula in the second pentad of May and extended eastward to the South China Sea and westward to the BoB in mid-May. The onset over India in late May was preceded by a transient precipitation event in mid-May. The heavy precipitation in early June was followed by a break period from mid to late June. Figure 2 also shows westward and eastward propagating signals. The former is probably associated with convectively coupled Rossby waves, and the latter may be related to the MJO (Wheeler and Kiladis 1999) and prevailing low-level westerlies, which may induce downstream propagation of orographic convection (Chu and Lin 2000; Chen and Lin 2005). A Hovmöller diagram of 12-yr climatological mean daily precipitation during 1998–2009 (not shown) is similar to Fig. 2a in terms of the seasonal progression of the monsoon precipitation, but the propagating features are weaker in the climatological mean.
(a) Hovmöller diagram of daily mean TRMM precipitation (mm h−1) averaged over 10°–20°N and (b) maximum terrain height between 10° and 20°N.
Citation: Journal of Climate 25, 7; 10.1175/JCLI-D-11-00136.1
The maximum terrain height between 10° and 20°N is shown in Fig. 2b. Consistent with Fig. 1, heavy precipitation tends to be anchored on the windward side of high terrain.
4. Monsoon onset and wind–terrain interaction in the RegCM simulations
In this section we will examine the effects of the mesoscale terrain on the monsoon circulation in the RegCM simulations, focusing on the monsoon onset. In the control run, the model’s lower boundary condition is specified by the topography data from GTOPO30. The Western Ghats over the Indian subcontinent, the Tenasserim Range over the western Indochina Peninsula, and the Annamite Range over eastern Indochina are reasonably represented (Fig. 3a). Three sensitivity experiments were designed by systematically removing the high terrain in the Southeast–South Asian monsoon region. In experiment No-BRDG high terrain over the Maritime Continent land bridge is removed (Fig. 3b), in experiment No-SEA high terrain over the Southeast Asia is removed (Fig. 3c), and in experiment No-Topo all of the high terrain in the Southeast–South Asian monsoon region is removed (Fig. 3c). When high terrain is removed within a subdomain, the land surface height is set to 10 m uniformly, and a cosine function profile is used to relax it to the height outside of the subdomain over a 240-km (eight grid points) transition zone along the lateral boundaries.
Topography (m) for (a) the control (CTRL) run, (b) the run with the high terrains over the central Maritime Continent removed (No-BRDG), (c) the run with high terrains over Southeast Asia removed (No-SEA), and (d) the run with all high terrains removed over South and Southeast Asia (No-Topo).
Citation: Journal of Climate 25, 7; 10.1175/JCLI-D-11-00136.1
Figure 4a shows the vertical structure of the zonal wind along 15°N averaged over 20 April–30 June in the control simulation. In the monsoon region, the mean flow is characterized by upper-level easterlies and low-level westerlies. The westerly flow east of 80°E is much deeper, extending up to 350 hPa. This is related to the early monsoon onset over Southeast Asia. The deep westerly structure over India and the Arabian Sea is established only after the Indian summer monsoon onset, which is consistent with the ERA-40 reanalysis. The mesoscale model simulation also shows gravity wave response along mountain ranges, which is not resolved in the 2.5° × 2.5° ERA-40 data. The distribution of the vertical velocity and precipitation are shown in Figs. 4b and 4c, respectively. Strong updrafts and heavy precipitation occur upslope of the Western Ghats, the Tenasserim Range, and the Annamite Range. Over the eastern BoB, precipitation and weak ascent extend a few hundred kilometers upstream off the coastal mountains. Weak but extensive ascent is also present over the South China Sea.
Vertical cross sections along 15°N: (a) zonal wind (m s−1), (b) omega (10−4 Pa s−1), and (c) precipitation (thick curve, mm h−1) averaged during 20 Apr–30 Jun. The thin curve in (c) represents terrain height along 15°N (m: with the ordinate on the right).
Citation: Journal of Climate 25, 7; 10.1175/JCLI-D-11-00136.1
The Hovmöller diagram of precipitation averaged over 10°–20°N is shown in Fig. 5a. Similar to what is shown by the TRMM data (Fig. 2), the monsoon onset occurs first over the Indochina Peninsula in early May after a period of premonsoon precipitation, and then extends westward to the BoB and eastward to the South China Sea in mid-May. The onset over India occurs in late May. In all of these regions the monsoon onset is characterized by a rapid increase of precipitation on the windward side of the high terrain, consistent with Chang et al. (2005). However, the model-simulated rainfall is a significant overestimate compared to the TRMM data, with the simulated seasonal mean precipitation up to 20% larger. This may be partly due to the lateral boundary forcing derived from ERA-40 reanalysis. Wang and Yang (2008), in a study using the Weather Research and Forecasting (WRF) model, found that the simulation of the seasonal progression of precipitation forced by the ERA-40 reanalysis is better than that driven by the National Centers for Environmental Prediction (NCEP)–Department of Defense (DOD) reanalysis, but the water vapor convergence over the Southeast Asian monsoon region is 47% higher. This bias in moisture convergence may lead to the overprediction of precipitation.
Hovmöller diagram of daily mean precipitation (mm h−1) averaged over 10°–20°N in (a) the CTRL run, (b) the No-BRDG run, (c) the No-SEA run, and (d) the No-Topo run.
Citation: Journal of Climate 25, 7; 10.1175/JCLI-D-11-00136.1
In experiment No-BRDG (Fig. 5b), the high terrain over the Maritime Continent is removed. Compared to the control run, precipitation is enhanced over the Indochina Peninsula and the windward side of the Western Ghats. The seasonal march of monsoon rainfall remains broadly the same as in the control run.
When the high terrain over Southeast Asia is removed, precipitation is significantly reduced not only in the windward side of the Tenasserim Range but also over the central and eastern Indochina (Fig. 5c). As shown in Fig. 2 and Fig. 5a, convective systems over the Indochina Peninsula tend to originate windward of the Tenasserim Range and then propagate inland (downstream) in May, which is similar to the third (III) flow regime identified by Chu and Lin (2000) with a high moist Froude number. Removal of high terrain weakens the eastward propagating convective systems and thus reduces precipitation over the entire peninsula. The Indochina Peninsula remains relatively dry until early June except for some transient convective features in late April and early May, and the monsoon onset is delayed to early June after the Indian monsoon onset. By contrast, the precipitation over India remains largely unchanged, and the monsoon onset occurs in late May as in the control run. In summary, the westward march of the rainband prior to the Indian monsoon onset is absent, while the eastward march of the rainband following the Indian monsoon onset does not change much except that precipitation is no longer confined along the windward side of the western Indochina Peninsula.
Figure 5d shows the seasonal variations of precipitation in the No-Topo simulation (with mountains all removed over South–Southeast Asia). Compared to No-SEA (Fig. 5c), the major difference is the significantly reduced precipitation along the west coast of India. A closer look, however, reveals that precipitation over central India (80°–90°E), although weaker, has a temporal evolution similar to the No-SEA experiment or the control run in that the monsoon onset still occurs in late May, preceded by premonsoon precipitation. This suggests that the Indian monsoon onset is controlled by the large-scale land–sea thermal contrast while the monsoon onset over the Indochina Peninsula is driven by the local wind–terrain interaction. This may explain Xu and Chan’s (2001) findings that the monsoon onset over the BoB occurs before the reversal of the meridional temperature gradient between the Indian Ocean and Tibetan Plateau while the Indian monsoon onset is accompanied by the reversal of the temperature gradient.
Figure 6 shows the Hovmöller diagram of the 850-hPa zonal wind. In both the control experiment and No-BRDG experiment, the low-level westerlies is first established over the BoB and the Indochina Peninsula in April prior to the monsoon onset in these regions. In the No-SEA and No-Topo experiments, the topography over Southeast Asia is removed, and the establishment of the low-level westerlies is delayed until late May. Besides, compared to the control run, the westerly flow upstream of the Western Ghats is slightly weakened in the No-Topo experiment, consistent with reduced precipitation windward of the Western Ghats (Fig. 5).
As in Fig. 5, but for the 850-hPa zonal wind (m s−1) averaged between 5° and 15°N.
Citation: Journal of Climate 25, 7; 10.1175/JCLI-D-11-00136.1
To investigate why the monsoon onset over the Indochina Peninsula is delayed in the No-SEA and No-Topo experiments, we examine the Hovmöller diagram of the upper-tropospheric (500–200 hPa) temperature difference between 5° and 20°N (Fig. 7), which represents the regional-scale meridional temperature gradient. The large-scale meridional temperature gradient between the Tibetan Plateau and the Indian Ocean can be regarded as being the same in all experiments, as the same lateral boundary conditions and SST forcing are used to drive the model, and the large-scale meridional temperature gradient reversal is mainly due to the temperature increase over the Tibetan Plateau (Li and Yanai 1996).
As in Fig. 5, but for the upper-tropospheric (200–500 hPa) temperature (K) at 20°N minus that at 5°N.
Citation: Journal of Climate 25, 7; 10.1175/JCLI-D-11-00136.1
In the control run (Fig. 7a), the meridional temperature gradient reverses over the Indochina Peninsula and the eastern BoB by 20 April and over the Arabian Sea in mid May. It is interesting to note the negative temperature difference (i.e., temperature at 20°N lower than at 5°N) in the mountainous regions, which is likely due to the adiabatic cooling effect associated with vigorous orographic updrafts. Over India, this negative temperature difference persists until the end of May, and then the latent heating associated with orographic precipitation becomes dominant and the temperature gradient is reversed. Over the Indochina Peninsula, the adiabatic cooling is dominant throughout the simulation along the Tenasserim Range and the Annamite Range, with strong warming upstream. While a vigorous updraft occurs close to steep terrain as shown in Fig. 4, orographic precipitation begins to occur well upstream of the mountain (Smith 1982) and contributes to the upper-tropospheric warming and the temperature gradient reversal.
The evolution of the temperature gradient in experiment No-BRDG is similar to that in the control run except that the strong positive temperature difference is absent around 115°E. This is due to the removal of the mountains over the island of Borneo. The orographic lifting associated with the mountains induces adiabatic cooling and a positive temperature difference in the control run.
In the No-SEA experiment, the adiabatic cooling associated with orographic cooling is no longer present with the removal of the high terrain over the Indochina Peninsula, and the temperature gradient is not reversed until June over the Indochina Peninsula and the eastern BoB. A closer examination of the control run reveals that reversal of the temperature gradient over the Indochina Peninsula is mainly due to the upper-tropospheric warming along 20°N (not shown). The evolution of temperature, wind, and precipitation suggests that high terrain enhances the premonsoon precipitation over the Indochina Peninsula, which warms the upper troposphere and facilitates the earliest summer monsoon onset over the Indochina Peninsula and the eastern BoB. Despite the significant change over Southeast Asia due to terrain removal, the temperature gradient over India is still reversed in mid May, similar to the control run and the No-BRDG experiment. This suggests that the upper-tropospheric warming over India is largely independent of orographic precipitation to the east.
In the No-Topo experiment, high terrain is also removed over the Indian subcontinent. The adiabatic cooling over the mountainous region disappears, but the temperature evolution is broadly the same as in the other experiments in that the temperature gradient reversal over the eastern Arabian Sea and the Indian subcontinent still occurs in mid May. This is consistent with the precipitation evolution and suggests that orographic precipitation does not play a dominant role in the upper-tropospheric warming and monsoon onset over India.
To further examine the interaction between orographic precipitation and the monsoon circulation, precipitation and 850-hPa zonal wind fields averaged from 20 April to 30 June are shown in Fig. 8 and Fig. 9. Compared to the control run, precipitation is reduced in experiment No-BRDG over the Malay Peninsula, Sumatra, and Borneo and is enhanced windward of the Western Ghats and the Tenasserim Range. As shown in Fig. 9b, the westerly flow shifts slightly northward over the equatorial eastern Indian Ocean and is strengthened over the eastern BoB and the eastern Arabian Sea. Given the small magnitude of the differences, further experiments are necessary to evaluate whether these changes are robust.
Precipitation (mm h−1) averaged from 20 Apr to 30 Jun in (a) the CTRL run, (b) the No-BRDG run, (c) the No-SEA run, and (d) the No-Topo run.
Citation: Journal of Climate 25, 7; 10.1175/JCLI-D-11-00136.1
The 850-mb zonal wind (m s−1) averaged from 20 Apr to 30 Jun in (a) the CTRL run, (b) the No-BRDG run, (c) No-SEA run, and (d) the No-Topo run.
Citation: Journal of Climate 25, 7; 10.1175/JCLI-D-11-00136.1
In experiment No-SEA (Fig. 8c and Fig. 9c), the seasonal mean precipitation is reduced over the Indochina Peninsula as has been shown in Fig. 5, while precipitation is enhanced over the BoB. It is interesting to note that the 850-hPa westerly flow is significantly weakened upstream of the Indochina Peninsula all the way to the eastern Arabian Sea. The weakened westerly flow over the Arabian Sea is likely responsible for the reduced precipitation along the west coast of India.
In experiment No-Topo, precipitation is significantly reduced over India, especially along the windward side of the Western Ghats (Fig. 8d). Precipitation over the eastern and northern BoB is also reduced. Maximum precipitation is broadly collocated with strong low-level westerlies over the southeastern Arabian Sea, southern India, and the southwestern BoB. The low-level westerly monsoon flow is overall weaker than in the other experiments (Fig. 9d).
In summary, this group of experiments suggests the following interaction between high terrain, precipitation, and low-level westerlies: i) removal of high terrain reduces precipitation both upstream and downstream; ii) the decrease in precipitation is accompanied by the weakening of the low-level westerly flow over a large longitude range (up to 30°) upstream of the high terrain; iii) wind–terrain–precipitation interaction plays a key role in the earliest Asian summer monsoon onset over the Indochina Peninsula but it is not essential for the monsoon onset over India. These experiments also imply a positive feedback between orographic precipitation and the upstream low-level wind: a stronger inflow leads to stronger orographic precipitation, which in turn strengthens the low-level westerly via diabatic heating. This positive feedback will be further explored in the next section.
5. Feedback of convection to the large-scale wind
To elucidate the interaction between the large-scale monsoon circulation and orographic precipitation, we carried out sensitivity tests with the terrain height over Southeast Asia (indicated by a box in Fig. 3a) varying from 0 to 1.75 times of that in the control run. We focus on Southeast Asia because the experiments discussed in section 4 show that mountains in this region have a strong impact on local precipitation and upstream low-level wind. Besides, this region is more than 45° longitude away from the western boundary of the model domain, and the impact of the upstream boundary is relatively weak.
Figure 10 shows the variations of precipitation (left panels), 850-hPa zonal wind (middle panels), and the vertical cross section of zonal wind averaged over 5°–15°N (right panels). Both precipitation and zonal wind are averaged from 20 April to 30 June. As the terrain height increases from 0% (flat land surface) to 75%, the zonal wind below 600 hPa weakens over the BoB. As the terrain height is further increased from 75% to 175%, the low-level westerlies increase monotonically with terrain height. The precipitation field, although noisy, varies with terrain height in a similar way as the low-level westerlies.
(left) Precipitation (mm h−1), (middle) 850-mb zonal wind (m s−1), and (right) vertical profile of zonal wind averaged between 5° and 15°N in the sensitivity tests, where the terrain height over the Indochina Peninsula (see Fig. 3a) varies from (top to bottom) zero to 1.75 times of that in the control run. All variables are averaged from 20 Apr to 30 Jun.
Citation: Journal of Climate 25, 7; 10.1175/JCLI-D-11-00136.1
To better illustrate the variations of precipitation and low-level westerlies in the sensitivity experiments, Fig. 11 shows the 850-hPa zonal wind averaged over the BoB (5°–15°N, 80°–95°E), precipitation averaged over the Indochina Peninsula (10°–20°N, 92°–108°E), and precipitation averaged along the Tenasserim Range (12°–19.5°N, 96°–99°E) as a function of the terrain height (0%–175%). Consistent with Fig. 10, the zonal wind first decreases from 10.6 to 9.3 m s−1 when the terrain height is increased from 0% to 75%, and then increases from 9.3 to 11.6 m s−1 when the terrain height is increased from 75% to 175%. The precipitation along the Tenasserim Range reaches its minimum when terrain height is set as 50% and then begins to increase. The total precipitation over the Indochina Peninsula shows a similar trend.
Zonal wind averaged over 5°–15°N, 80°–95°E (open squares with the ordinate on the right: m s−1), precipitation averaged over the entire Indochina (10°–20°N, 92°–108°E) (closed triangles with the ordinate on the far left: mm day−1), and precipitation averaged along the Tenasserim Range (12°–19.5°N, 96°–99°E) (closed circles with the ordinate on the near left: mm day−1) in the sensitivity tests. The abscissa indicates the terrain heights in the sensitivity tests, varying from 0 to 1.75 of that in the control run.
Citation: Journal of Climate 25, 7; 10.1175/JCLI-D-11-00136.1
These experiments suggest two opposite impacts of high terrain on precipitation and low-level westerlies. On one hand, high terrain tends to decelerate the low-level inflow because of the surface drag effect (surface friction, gravity wave drag, and upstream flow blocking); on the other hand, orographic lifting can enhance precipitation, and the resultant orographic precipitation, acting as a diabatic heat source for the atmosphere, can strengthen the upstream inflow in the lower troposphere. Previous studies (e.g., Miglietta and Buzzi 2001) have noted that orographic precipitation can induce additional upslope lifting and reduce the height of diving streamlines in the vicinity of the mountain. Our experiments suggest that the regional mountains over the Indochina Peninsula, despite their moderate height and narrow width, can affect the large-scale monsoon circulation through persistent, heavy orographic precipitation.
Figures 10 and 11 suggest that the surface drag effect is dominant with small mountain heights and that the orographic heating effect is dominant with large mountain heights. The competition of the two opposing effects may be understood by considering the amount of orographic lifting required to induce precipitation. Only when the upslope lifting is strong enough to drive air parcels above the lifting condensation level (LCL) can condensation and cloud formation occur. If parcels are further driven above the level of free convection (LFC), deep convection may be triggered and continuous precipitation may be sustained given sufficient moisture supply. As shown in Table 1, when the terrain heights in the sensitivity tests are set less than their actual values, the nondimensionalized mountain height hND is less than 1. For small nondimensionalized mountain heights, linear theory (Smith 1980) suggests that the upslope lifting height can be approximated by the mountain height, which begins to exceed the LCL in experiment 0.75SEA. This is when precipitation begins to increase with the terrain height, followed by the increase of low-level westerlies (Fig. 11). Previous studies also suggested that airflow will transit to the “flow over” regime to the “flow around” regime when hND exceeds a critical value (1.2~1.3) (Smith and Gronas 1993; Baines and Smith 1993). Besides, orographic precipitation cannot keep increasing with upslope lifting due to limited moisture content in the atmosphere. This may explain why the precipitation curve begins to level off when the terrain height is set to 150% or larger in Fig. 11.
Mountain height (H) over the western Indochina Peninsula, nondimensionalized mountain height (hND), LCL, and LFC in the sensitivity tests. Nondimensionalized mountain height is calculated based on the zonal wind and moist buoyancy frequency averaged below 850 hPa over (12°–15°N, 93.5°–97°E) (an upstream region over the open ocean with relatively uniform low-level zonal wind) from 20 Apr to 30 Jun. LFL and LCL are derived based on the mean sounding averaged over the same region and during the same period.
6. Conclusions and discussion
We used regional climate model simulations to study the effects of terrain on the monsoon circulation, in particular, the monsoon onset over South and Southeast Asia. The control simulation, which was forced by the ERA-40 reanalysis data and the USGS topographic data (GTOPO30), reproduced the seasonal progression of precipitation in the year 2000 reasonably well. The model monsoon onset first occurs over the Indochina Peninsula and then over the eastern BoB, the South China Sea, and India, which is consistent with observations.
In the sensitivity tests, topography was removed systematically over the Maritime Continent, the Indochina Peninsula and South Asia. When topography on the Indochina Peninsula is removed, the monsoon onset over the Indochina Peninsula is significantly delayed. The onset first occurs at the western coast of India and then advances eastward to Indochina, and precipitation is significantly reduced over the entire Indochina Peninsula. The reversal of the upper-tropospheric temperature gradient (temperature difference between 5° and 20°N) is also delayed to June. Our diagnosis suggests that premonsoon orographic precipitation over the Indochina Peninsula may play an important role in reversing the regional meridional temperature gradient, establishing low-level westerlies, and triggering the earliest summer monsoon onset in this region.
When mountains are removed from the South–Southeast Asian monsoon region, precipitation is further reduced and becomes more homogeneous, instead of being anchored along the windward side of the mountains. However, the Indian monsoon onset, defined in terms of precipitation increase, does not change (i.e., still occurs in late May). This suggests that the Indian monsoon onset is likely controlled by the large-scale meridional thermal contrast and that the local wind–terrain interaction and orographic precipitation do not play any major role in the seasonal progression of Indian summer monsoon.
The model experiments also show that changes in precipitation are accompanied by changes in the low-level westerlies upstream of high terrain, which can be expected as a result of a positive feedback between the orographic precipitation and atmospheric circulation: a stronger westerly inflow enhances upslope lifting and orographic precipitation, and diabatic heating associated with orographic precipitation in turn strengthens the upstream low-level westerlies and moisture supply. On the other hand, mountains can also decelerate airflow due to the surface drag effect. In our sensitivity test, as the terrain height increases from 0% to 175% of that in the control run, precipitation over the Indochina Peninsula and the low-level westerlies over the BoB first decrease and then increase. With low mountain heights, the upslope lifting is not sufficient to drive air parcels above the LCL and induce cloud formation, so the surface drag effect is dominant. The increased surface drag associated with an increasing terrain height tends to reduce the low-level westerlies and the associated moisture supply and thus reduces the orographic precipitation. With large mountain heights, the orographic lifting can drive air parcels above the LCL or even above the LFC and thus enhance precipitation or even trigger deep convection. The resultant orographic precipitation can in turn strengthen the low-level inflow through the diabatic forcing effect. In the sensitivity tests, the terrain height is increased up to 175%, and a positive feedback is found between the upstream westerlies in the lower troposphere and the orographic precipitation. It is likely, however, that the surface drag or blocking effect may become dominant again if the terrain height increases further.
Although heavy orographic precipitation is confined along the windward side of high terrain, its impacts can be far reaching. The sensitivity tests show that the change of the upstream low-level westerlies due to precipitation variations over the Indochina Peninsula extends up to 3000 km westward to India and the eastern Arabian Sea, which induces variations of orographic precipitation along the Western Ghats. These simulations suggest that a realistic representation of the regional terrain is critical to modeling the distribution and seasonal variations of monsoon precipitation.
This study is based on simulations of one single season. Ensemble simulations or extended regional climate simulations are desirable if computational resources are available. The horizontal resolution of the simulations is also too coarse to resolve convection explicitly, which introduces some uncertainty in the quantitative results of this study, but we feel that the mechanisms of wind–terrain–precipitation interaction are likely robust. These mechanisms may have some important implication for the coevolution of orographic precipitation patterns and topography on the geological time scale. Future studies may also include the effects of static stability, moisture distribution, and vertical wind shear, which are not considered in this study.
Acknowledgments
This work was supported by the U.S. Office of Naval Research, under Awards N000141110446, N0001408WR20125, N0001409AF00002, and N0001410AF00002, and by the National Research Council of Taiwan under Grant NSC 100-2111-M-002-007.
REFERENCES
Baines, P. G., and R. B. Smith, 1993: Upstream stagnation points in stratified flow past obstacles. Dyn. Atmos. Oceans, 18, 105–113.
Bergeron, T., 1960: Problems and methods of rainfall investigation, address of the honorary chairman of the conference. Physics of Precipitation, Geophys. Monogr., Vol. 5, Amer. Geophys. Union, 5–30.
Briegleb, B., 1992: Delta-Eddington approximation for solar radiation in the NCAR community climate model. J. Geophys. Res., 97, 7603–7612.
Chang, C.-P., Z. Wang, J. McBride, and C. H. Liu, 2005: Annual cycle of Southeast Asia–Maritime Continent rainfall and the asymmetric monsoon transition. J. Climate, 18, 287–301.
Chang, C.-P., Z. Wang, and H. Hendon, 2006: The Asian winter monsoon. The Asian Monsoon, B. Wang, Ed., Praxis Publishing, 89–128.
Chen, L., E. R. Reiter, and Z. Feng, 1985: The atmospheric heat source over the Tibetan plateau: May–August 1979. Mon. Wea. Rev., 113, 1771–1790.
Chen, S.-H., and Y.-L. Lin, 2005: Effects of moist Froude number and CAPE on a conditionally unstable flow over a mesoscale mountain ridge. J. Atmos. Sci., 62, 331–350.
Chu, C.-M., and Y.-L. Lin, 2000: Effects of orography on the generation and propagation of mesoscale convective systems in a two-dimensional conditionally unstable flow. J. Atmos. Sci., 57, 3817–3837.
Dickinson, R., A. Henderson-Sellers, and P. Kennedy, 1993: Biosphere–Atmosphere Transfer Scheme (BATS) version 1e as coupled to the NCAR community climate model. National Center for Atmospheric Research Tech. Note NCAR/TN-387+STR, 72 pp.
Ding, Y., and Y. Liu, 2001: Onset and evolution of the summer monsoon season over the South China Sea during SCSMEX field experiment in 1998. J. Meteor. Soc. Japan, 79, 255–276.
Ding, Y., and J. C. L. Chan, 2005: The East Asia summer monsoon: An overview. Meteor. Atmos. Phys., 89, 117–142.
Emanuel, K. A., 1994: The physics of tropical cyclogenesis over the eastern Pacific. Tropical Cyclone Disasters, J. Lighthill et al., Eds., Peking University Press, 588 pp.
Fasullo, J., and P. J. Webster, 2003: A hydrological definition of Indian monsoon onset and withdrawal. J. Climate, 16, 3200–3211.
Giorgi, F., Y. Huang, K. Nishizawa, and C.-B. Fu, 1999: A seasonal cycle simulation over eastern Asia and its sensitivity to radiative transfer and surface processes. J. Geophys. Res., 104, 6403–6423.
Goswami, B. N., G. Wu, and T. Yasunari, 2006: The annual cycle, intraseasonal oscillations, and roadblock to seasonal predictability of the Asian summer monsoon. J. Climate, 19, 5078–5099.
Grell, G. A., J. Dudhia, and D. R. Stauffer, 1994: A description of the Fifth-Generation Penn State/NCAR Mesoscale Model (MM5). National Center for Atmospheric Research Tech. Note NCAR/TN-398+STR, 117 pp.
Grossman, R. L., and D. R. Durran, 1984: Interaction of low-level flow with the western Ghat Mountains and offshore convection in the summer monsoon. Mon. Wea. Rev., 112, 652–672.
Hahn, D. G., and S. Manabe, 1975: The role of mountains in the South Asian Monsoon Circulation. J. Atmos. Sci., 32, 1515–1541.
He, H., J. W. McGinnis, Z. Song, and M. Yanai, 1987: Onset of the Asian Summer Monsoon in 1979 and the effect of the Tibetan Plateau. Mon. Wea. Rev., 115, 1966–1995.
Holtslag, A., E. de Bruijn, and H.-L. Pan, 1990: A high-resolution airmass transformation model for short-range weather forecasting. Mon. Wea. Rev., 118, 1561–1575.
Kato, H., H. Hirakuchi, K. Nishizawa, and F. Giorgi, 1999: Performance of NCAR RegCM in the simulation of June and January climate over Eastern Asia and the high-resolution effect of the model. J. Geophys. Res., 104, 6455–6476.
Li, C., and M. Yanai, 1996: The onset and interannual variability of the Asian summer monsoon in relation to land–sea thermal contrast. J. Climate, 9, 358–375.
Luo, H., and M. Yanai, 1983: The large-scale circulation and heat sources over the Tibetan Plateau and surrounding areas during the early summer of 1979. Part I: Precipitation and kinematic analyses. Mon. Wea. Rev., 111, 922–944.
Luo, H., and M. Yanai, 1984: The large-scale circulation and heat sources over the Tibetan Plateau and surrounding areas during the early summer of 1979. Part II: Heat and moisture budgets. Mon. Wea. Rev., 112, 966–989.
Miglietta, M. M., and A. Buzzi, 2001: A numerical study of moist stratified flows over isolated topography. Tellus, 53A, 481–499.
Murakami, T., and Y. Ding, 1982: Wind and temperature changes over Eurasia during the early summer of 1979. J. Meteor. Soc. Japan, 60, 182–196.
Ogura, Y., and M. Yoshizaki, 1988: Numerical Study of orographic–convective precipitation over the Eastern Arabian Sea and the Ghat Mountains during the summer monsoon. J. Atmos. Sci., 45, 2097–2122.
Pal, J., E. Small, and E. Eltahir, 2000: Simulation of regional-scale water and energy budgets: Representation of subgrid cloud and precipitation processes within regcm. J. Geophys. Res., 105 (D24), 29 579–29 594.
Qian, W., and S. Yang, 2000: Onset of the regional monsoon over Southeast Asia. Meteor. Atmos. Phys., 75, 29–38.
Riehl, H., 1979: Climate and Weather in the Tropics. Academic Press, 611 pp.
Sato, T., and F. Kimura, 2007: How does the Tibetan Plateau affect the transition of Indian monsoon rainfall? Mon. Wea. Rev., 135, 2006–2015.
Simpson, J., C. Kummerow, W.-K. Tao, and R. F. Adler, 1996: On the Tropical Rainfall Measuring Mission (TRMM) satellite. Meteor. Atmos. Phys., 60, 19–36.
Small, E., F. Giorgi, and L. Sloan, 1999: Regional climate model simulation of precipitation in central Asia: Mean and interannual variability. J. Geophys. Res., 104, 6563–6582.
Smith, R. B., 1980: Linear theory of stratified hydrostatic flow past an isolated mountain. Tellus, 32, 348–364.
Smith, R. B., 1982: A differential advection model of orographic rain. Mon. Wea. Rev., 110, 306–309.
Smith, R. B., 1989: Mountain-induced stagnation points in hydrostatic flow. Tellus, 41A, 270–274.
Smith, R. B., and S. Gronas, 1993: Stagnation points and bifurcation in 3D mountain airflow. Tellus, 45A, 28–43.
Sun, L.-Q., F. Semazzi, F. Giorgi, and L. Ogallo, 1999a: Application of the NCAR regional climate model to eastern Africa 1. Simulation of the short rains of 1988. J. Geophys. Res., 104, 6529–6548.
Sun, L.-Q., F. Semazzi, F. Giorgi, and L. Ogallo, 1999b: Application of the NCAR regional climate model to eastern Africa 2. Simulation of interannual variability of short rains. J. Geophys. Res., 104, 6549–6562.
Wang, B., and LinHo, 2002: Rainy season of the Asian–Pacific monsoon. J. Climate, 15, 386–398.
Wang, B., and H. Yang, 2008: Hydrological issues in lateral boundary conditions for regional climate modeling: Simulation of East Asian summer monsoon in 1998. Climate Dyn., 31, 477–490, doi:10.1007/s00382-008-0385-7.
Wang, B., Q. Ding, and V. Joseph, 2009: Objective definition of the Indian summer Monsoon onset using large scale winds. J. Climate, 22, 3303–3316.
Webster, P. J., V. O. Magana, T. N. Palmer, J. Shukla, R. A. Tomas, M. Yanai, and T. Yasunari, 1998: Monsoons: Processes, predictability, and the prospects for prediction. J. Geophys. Res., 103, 14 451–14 510.
Wen, M., and R. Zhang, 2007: Role of the quasi-biweekly oscillation in the onset of convection over the Indochina Peninsula. Quart. J. Roy. Meteor. Soc., 133, 433–444.
Wheeler, M., and G. N. Kiladis, 1999: Convectively coupled equatorial waves: Analysis of clouds and temperature in the wavenumber–frequency domain. J. Atmos. Sci., 56, 374–399.
Wu, G., and Y. Zhang, 1998: Tibetan Plateau forcing and timing of the monsoon onset over South Asia and the South China Sea. Mon. Wea. Rev., 126, 913–927.
Xie, S.-P., H. Xu, N. H. Saji, Y. Wang, and W. T. Liu, 2006: Role of narrow mountains in large-scale organization of Asian monsoon convection. J. Climate, 19, 3420–3429.
Xu, J., and J. C. L. Chan, 2001: First transition of the Asian summer monsoon in 1998 and the effect of the Tibet–tropical ocean thermal contrast. J. Meteor. Soc. Japan, 79, 241–253.
Ye, D. Z., and Y. X. Gao, 1979. Meteorology of Qinghai-Xizhang (Tibet) Plateau. Science Press, 278 pp.
Moist Froude number is defined as 
