Delayed Response of the Onset of the Summer Monsoon over the Bay of Bengal to Land–Sea Thermal Contrast

Weihao Sun aState Key Laboratory of Numerical Modelling for Atmospheric Sciences and Geophysical Fluid Dynamics (LASG), Institute of Atmospheric Physics, Chinese Academy of Sciences, Beijing, China
bUniversity of Chinese Academy of Sciences, Beijing, China

Search for other papers by Weihao Sun in
Current site
Google Scholar
PubMed
Close
,
Guoxiong Wu aState Key Laboratory of Numerical Modelling for Atmospheric Sciences and Geophysical Fluid Dynamics (LASG), Institute of Atmospheric Physics, Chinese Academy of Sciences, Beijing, China
bUniversity of Chinese Academy of Sciences, Beijing, China

Search for other papers by Guoxiong Wu in
Current site
Google Scholar
PubMed
Close
,
Yimin Liu aState Key Laboratory of Numerical Modelling for Atmospheric Sciences and Geophysical Fluid Dynamics (LASG), Institute of Atmospheric Physics, Chinese Academy of Sciences, Beijing, China
bUniversity of Chinese Academy of Sciences, Beijing, China
cChinese Academy of Sciences Centre for Excellence in Tibetan Plateau Earth Sciences, Beijing, China

Search for other papers by Yimin Liu in
Current site
Google Scholar
PubMed
Close
https://orcid.org/0000-0002-5050-1642
,
Jiangyu Mao aState Key Laboratory of Numerical Modelling for Atmospheric Sciences and Geophysical Fluid Dynamics (LASG), Institute of Atmospheric Physics, Chinese Academy of Sciences, Beijing, China

Search for other papers by Jiangyu Mao in
Current site
Google Scholar
PubMed
Close
,
Moran Zhuang aState Key Laboratory of Numerical Modelling for Atmospheric Sciences and Geophysical Fluid Dynamics (LASG), Institute of Atmospheric Physics, Chinese Academy of Sciences, Beijing, China

Search for other papers by Moran Zhuang in
Current site
Google Scholar
PubMed
Close
, and
Xiaolin Liu aState Key Laboratory of Numerical Modelling for Atmospheric Sciences and Geophysical Fluid Dynamics (LASG), Institute of Atmospheric Physics, Chinese Academy of Sciences, Beijing, China
dSchool of Atmospheric Sciences and Guangdong Province Key Laboratory for Climate Change and Natural Disaster Studies, Sun Yat–sen University, Zhuhai, China

Search for other papers by Xiaolin Liu in
Current site
Google Scholar
PubMed
Close
Open access

We are aware of a technical issue preventing figures and tables from showing in some newly published articles in the full-text HTML view.
While we are resolving the problem, please use the online PDF version of these articles to view figures and tables.

Abstract

The mechanisms involved in the onset of the Bay of Bengal summer monsoon (BOBSM) were studied using reanalysis data and numerical model experiments. Results revealed that the weak meridional land–sea thermal contrast (LSTC) over the northern BOB in early spring enhances the lower-tropospheric easterly belt along 10°–15°N, which is unfavorable for the BOBSM onset. The BOBSM onset is driven by the cumulative impact of this LSTC along with the LSTC in the meridional direction across the equator and in the zonal direction across the tropics, together with air–sea interactions. While the LSTC intensifies over the northern BOB, a near-surface northward cross-equatorial flow develops south of India, inducing springtime zonal flow and surface sensible heating over the southern BOB and a pair of cyclones straddling the equator over the central Indian Ocean at 700 hPa. The zonal LSTC in the tropics generates near-surface cyclones over land and anticyclones over the sea. This induces a zonal SST warm pool around 10°N, which produces vertical westerly wind shear to the north and weakens the wintertime easterly aloft and the anticyclone to its north. As the cyclone over southern India develops eastward, the cyclone below 700 hPa develops northward over the eastern BOB in response to the enhancing tropical westerly and surface sensible heating. The wintertime anticyclonic belt and easterly belt split, and the southerly carries water vapor northward over the eastern BOB, heralding the onset of the BOBSM and presenting a delayed response to the springtime LSTC changes.

© 2023 American Meteorological Society. For information regarding reuse of this content and general copyright information, consult the AMS Copyright Policy (www.ametsoc.org/PUBSReuseLicenses).

Corresponding authors: Yimin Liu, lym@lasg.iap.ac.cn; Guoxiong Wu, gxwu@lasg.iap.ac.cn

Abstract

The mechanisms involved in the onset of the Bay of Bengal summer monsoon (BOBSM) were studied using reanalysis data and numerical model experiments. Results revealed that the weak meridional land–sea thermal contrast (LSTC) over the northern BOB in early spring enhances the lower-tropospheric easterly belt along 10°–15°N, which is unfavorable for the BOBSM onset. The BOBSM onset is driven by the cumulative impact of this LSTC along with the LSTC in the meridional direction across the equator and in the zonal direction across the tropics, together with air–sea interactions. While the LSTC intensifies over the northern BOB, a near-surface northward cross-equatorial flow develops south of India, inducing springtime zonal flow and surface sensible heating over the southern BOB and a pair of cyclones straddling the equator over the central Indian Ocean at 700 hPa. The zonal LSTC in the tropics generates near-surface cyclones over land and anticyclones over the sea. This induces a zonal SST warm pool around 10°N, which produces vertical westerly wind shear to the north and weakens the wintertime easterly aloft and the anticyclone to its north. As the cyclone over southern India develops eastward, the cyclone below 700 hPa develops northward over the eastern BOB in response to the enhancing tropical westerly and surface sensible heating. The wintertime anticyclonic belt and easterly belt split, and the southerly carries water vapor northward over the eastern BOB, heralding the onset of the BOBSM and presenting a delayed response to the springtime LSTC changes.

© 2023 American Meteorological Society. For information regarding reuse of this content and general copyright information, consult the AMS Copyright Policy (www.ametsoc.org/PUBSReuseLicenses).

Corresponding authors: Yimin Liu, lym@lasg.iap.ac.cn; Guoxiong Wu, gxwu@lasg.iap.ac.cn

1. Introduction

The onset of the Asian summer monsoon (ASM) is characterized by a seasonal reversal of the prevailing winds accompanied by a significant increase in rainfall. However, this seasonal transition is not synchronous among the different subsystems of the ASM. It has been recognized that the earliest onset of the ASM occurs over the eastern Bay of Bengal (BOB) at the beginning of May, then over the South China Sea (SCS), and last over India (Lau and Yang 1997; Wu and Zhang 1998; Li and Zhang 2009). As a result of the abrupt increase in atmospheric energy and the water cycle, the onset of the ASM is always associated with extreme weather events and therefore greatly affects both society and economic growth in Asian countries (Lau and Yang 1997; Ding and Liu 2001; Bombardi et al. 2020; Li et al. 2022).

As the first area where the summer monsoon occurs, local convective activity over the BOB can influence the subsequent onset of the monsoon over the SCS and India (Liu et al. 2002; Wu et al. 2005; Liu et al. 2015). Anomalies in the onset dates of the BOB summer monsoon (BOBSM) are also related to precipitation anomalies in Asia—for example, in southwest China, southern India, and the Indochina Peninsula (Yan et al. 2003; Chen et al. 2006; Xing et al. 2016a; Hu et al. 2022). Investigating the mechanism of the onset of the BOBSM will help us to better understand the Asian monsoon system and to predict climate anomalies over the surrounding areas.

The circulations in both the lower and upper troposphere undergo abrupt changes over the BOB during the onset of the monsoon. Mao et al. (2004) showed that the BOBSM starts with the occurrence of northward tilting of the subtropical ridge surface and the reversal of the meridional temperature gradient (MTG) in the upper troposphere. The moment when the ridge surface becomes perpendicular (i.e., ∂T/∂y = 0) is defined as the onset date of the monsoon (Mao and Wu 2007). In the lower troposphere, a distinct vortex known as the monsoon onset vortex is usually generated over the southern BOB. When the onset vortex moves northward, the wintertime tropical anticyclone and the easterly wind belt to its south over the northern BOB split, southerly flow develops over the eastern BOB, and the equatorial westerly merges into the subtropical westerly, transporting a large amount of water vapor to the subtropical region (Lau et al. 1998; Liu et al. 2002).

Wu et al. (2011, 2012) found that Tibetan Plateau forcing and air–sea interaction in the Indian Ocean play key roles in the formation of the BOB monsoon onset vortex. About two pentads prior to the onset of the monsoon, a meridionally warmest sea surface temperature (SST) axis can be observed in the central BOB (Jiang and Li 2011; Yu et al. 2012). Strong sensible heating from the ocean surface associated with this warm SST provides available potential energy for the formation of the vortex. Based on a theoretical model, Xing et al. (2016b) proposed that the secondary meridionally warmest SST axis may be regarded as external forcing for the evolution of the atmospheric circulation.

The onset of the BOBSM has a significant interannual variation. Previous studies have found that the interannual variability of the onset of the BOBSM is closely related to El Niño–Southern Oscillation (ENSO) (Mao and Wu 2007; Feng et al. 2013), with an early (late) onset of the BOBSM being preceded by a La Niña (El Niño) event. This interannual relationship is modulated by the Pacific decadal oscillation (PDO). The onset of the BOBSM shows a higher correlation with ENSO during the warm phases of the PDO than the cold phases (Wu and Mao 2019).

Despite recent progress in the study of the interannual variability of the onset of the BOBSM and its causes (such as ENSO and the PDO), the mechanism behind the transition in the large-scale circulation is still not fully understood—for example, the processes and relationships in the lower- and upper-level circulations and the external forcing of the multiscale monsoon-type transition. More importantly, the reversal of land–sea thermal contrast (LSTC) along the coast of South Asia appears in late February, whereas the onset of the ASM occurs in early May, implying a time lag of more than 2 months in the seasonal changes between the atmospheric circulation and the external thermal forcing. The evolution of the atmospheric circulation in response to the changes in LSTC during the onset of the ASM is also not well understood. One of the objectives of this paper is to revisit the feedback process during the onset of the monsoon. Considering that the monsoon is a complex system linking the ocean, land, and atmosphere, and that previous research has been confined to the feedback processes over the sea, here we explore the role of the land–sea distribution in the onset of the BOBSM.

The remainder of this paper is organized as follows. Section 2 describes the data, methods, and model used in this study. Section 3 analyzes the evolution and mechanism of the atmospheric circulation during the onset of the ASM. The delayed response of the development of deep convection associated with the BOBSM onset to the LSTC over the northern BOB is revealed. Section 4 explores the role of the cross-equatorial flow (CEF) as a tropical precursor in the onset of the BOBSM. Section 5 analyzes the response of the circulation to changes in land–sea distribution in the model to verify the results obtained from data analysis. A summary and discussion are presented in section 6.

2. Data, methods, and experimental design

a. Data

The following datasets were used in this study:

  1. The Modern-Era Retrospective Analysis for Research and Applications, version 2 (MERRA-2; Gelaro et al. 2017) dataset for the time period 1980–2019, including the 3D wind field, geopotential height, air temperature, surface heating flux, and temperature tendencies, with a horizontal resolution of 0.5° × 0.625°.

  2. Daily precipitation data from the Tropical Rainfall Measurement Mission Project (TRMM) dataset (3B42 Version 7) from 1998 to 2019, with a spatial resolution of 0.25° × 0.25° (Huffman et al. 2007).

  3. The daily Optimum Interpolation SST (OISST) version 2 analysis dataset from the National Oceanic and Atmospheric Administration Satellite and Information Service, with a resolution of 1° × 1° over a 38-yr period from 1982 to 2019 (Huang et al. 2021).

b. Methods

Following Mao and Wu (2007), the onset date of the BOBSM was defined as the day when the following criteria were satisfied: 1) the 200–500-hPa average MTG over the eastern BOB (5°–15°N, 90°–100°E) changes from negative to positive and 2) the MTG remains positive for >10 days to exclude weather disturbances. Liu et al. (2013) compared the MTG with other criteria for the onset of the monsoon, including the area-averaged 850-hPa zonal wind and the outgoing longwave radiation. They showed that the MTG has a significant correlation with these other criteria and can therefore be used to reliably define the onset date of the BOBSM.

The onset dates of the BOBSM show a marked interannual variability, so a composite analysis method was used to analyze the climate-mean processes of the monsoon onset. For each year, the onset date was defined as the zero day (D0), with the dates before (after) D0 labeled as negative (positive) days. We chose days −30 to +30 relative to the onset date of the BOBSM in each year and then calculated the arithmetical mean of each variable during the onset period (from D−30 to D+30) over each corresponding day for the time period 1980–2019.

To determine the diabatic heating in the atmosphere, Q1 was calculated following the scheme deduced by Yanai et al. (1973):
Q1=Cp[Tt+VhT+(PP0)R/Cpωθp]
where Cp, T, p, θ, ω, and R are the specific heat of dry air at constant pressure, the air temperature, air pressure, potential temperature, vertical velocity in p coordinates, and the gas constant for dry air, respectively.

c. Community Earth System Model simulations

The model used in this study was the Community Earth System Model, version 2 (CESM2), of the National Center for Atmospheric Research (Danabasoglu et al. 2020). CESM2 is a fully coupled global climate model that provides state-of-the-art simulations of Earth’s climate. It consists of the Com\munity Atmosphere Model, version 6 (CAM6; Danabasoglu et al. 2020); the Community Land Model, version 5 (Lawrence et al. 2019); the Community Ice Code, version 5 (CICE5; Hunke et al. 2015); the Community Ice Sheet Model, version 2.1 (Lipscomb et al. 2019); and the Modular Ocean Model, version 6 (MOM6; Adcroft et al. 2019). The model grid scheme employed in this study was f09_tx0.66. The CAM6 experiments used the nominal 1° (1.25° longitude and 0.9° latitude) horizontal resolution with 32 vertical levels and a model top of 2.26 hPa. MOM6 has 65 vertical levels and a horizontal resolution of approximately 0.66°. CICE5 shares the same horizonal grid as MOM6.

The focus in this study was on the modulation of the onset of the BOBSM by the land–sea distribution, which was assessed in four numerical experiments: CTRL, No_Indian, No_Inch, and NO_Inch_Indian. The real land–sea distribution was used in the model in the CTRL experiment. In the other three experiments, the land grid points comprising the Indian subcontinent, the Indochina Peninsula, and the Indian subcontinent together with the Indochina Peninsula were removed, respectively, and replaced by ocean grid points. Table 1 gives the details of the target regions in the experiments. The sensitivity experiments differed from the CTRL experiment in the removal of the land distribution in certain regions of South Asia. Because the focus was on the impact of the LSTC over South Asia, the islands along the equator, including the island of Java, were retained in all the sensitivity experiments. All the experiments were integrated for 150 years, and the annual and global mean surface temperature reached a quasi-stable state long before the 100th year in each experiment (figures not shown). Therefore, the last 50 years of daily outputs were used for analysis.

Table 1.

Design of numerical experiments.

Table 1.

3. Mechanism of the BOBSM onset

a. Background conditions before BOBSM onset

Figure 1 shows the onset dates of the BOBSM for each year from 1980 to 2019 based on the MTG index defined in section 2, which indicates a strong interannual variability of onset dates of the BOBSM. The climatological onset date of the BOBSM is 1 May and the interannual onset dates of the BOBSM differ significantly. The late (early) years for the onset of the BOBSM were selected when the standard deviation was more than 1.0 (less than −1.0). On average, the onset date of the BOBSM is 15 April in the early years and 14 May in the late years; the range between the early and late onset dates is about one month. This is consistent with previous results, although there are slight differences in the index used (e.g., Xing et al. 2016a; Li et al. 2018). The difference in circulation and precipitation between the early and late monsoon onset cases will be discussed in section 4 to understand the mechanism of ASM onset over the BOB.

Fig. 1.
Fig. 1.

Onset dates of the BOBSM from 1980 to 2019. Dashed lines indicate one standard deviation from the average onset date; red (blue) means that the onset date is later (earlier) than one standard deviation.

Citation: Journal of Climate 36, 12; 10.1175/JCLI-D-22-0478.1

Because the climatological onset date of the BOBSM is in early May, the mean meteorological fields in April can be used to represent the background conditions before the onset of the summer monsoon. The one-month mean meteorological fields before the onset date (D0) are similar. Figure 2a shows the meridional vertical cross section of the temperature averaged between 80° and 100°E over the BOB sector. A northward MTG is already present in the lower atmosphere over the BOB prior to the onset of the BOBSM, indicating that the LSTC develops much earlier than the seasonal change in the atmospheric circulation. By contrast, the temperature in the upper atmosphere decreases poleward, retaining the wintertime status.

Fig. 2.
Fig. 2.

Climatology in April for (a) the pressure–latitude section of temperature (K) averaged between 80° and 100°E and (b) the 2-m surface wind (vectors; m s−1), precipitation (white contours; 5, 10, and 15 mm day−1) and the 2-m temperature (shading; °C). (c) Area-averaged vertical profile of diabatic heating (K day−1) over the “west” and “east” regions shown in (b). Red and blue denote sensible heating and latent heating, respectively, and the open triangles and closed circles denote the west and east regions, respectively. (d) Pressure–longitude cross section of the zonal circulation averaged over 10°–15°N [vectors; zonal wind in m s−1 and vertical motion (multiplied by a factor of −60) in Pa s−1], the horizontal divergence (shading; 10−6 s−1), and the zonal deviation of the equivalent potential temperature (contours; interval: 0.5 K; negative values shown as broken lines).

Citation: Journal of Climate 36, 12; 10.1175/JCLI-D-22-0478.1

Figure 2b shows the distributions of precipitation and the surface 2-m air temperature and wind. After the spring equinox, the increased solar radiation efficiently heats the tropical land surface. The air temperature over the land is clearly warmer than that over the surrounding ocean. A sea breeze therefore develops along the coast from the ocean to the land in the tropics, and remarkable southerly winds already invade Bangladesh, inducing pronounced precipitation there. A cyclonic circulation dominates the Indian subcontinent and the Indochina Peninsula, whereas an anticyclonic circulation controls the Arabian Sea and the BOB.

Figure 2c shows the vertical profiles of diabatic heating over the Indian and the Indochina peninsulas averaged between 10° and 15°N to further explore the characteristics of diabatic heating over the tropical continents. Below 800 hPa, strong sensible heating can be observed over the tropical land and the intensity of heating over the Indian Peninsula is larger than that over the Indochina Peninsula. According to the theory of atmospheric thermal adaptation (Hoskins 1991; Wu and Liu 2002), strong surface heating generates a cyclonic circulation near the surface. As shown in Fig. 2b, a low-level southwesterly wind forms along the eastern coast of India and northerly wind forms along the western coast of the Indochina Peninsula. An anticyclone is located over the northern BOB with divergence therein, which hinders the development of a southwesterly flow from the tropics.

However, over the Indochina Peninsula, there is distinct latent heating over the Indochina Peninsula with a vertical structure peaking at 700 hPa (Fig. 2c). Latent heating is weak over the Indian Peninsula. This difference in diabatic heating between the two continents indicates the mechanical effect of the Tibetan Plateau on the tropical circulation. The midlatitude westerlies are deflected by the high terrain and produce a cyclonic circulation to the south of the Tibetan Plateau, which transports dry and cold air southward to the Indian Peninsula and moist air northward to the Indochina Peninsula (Wu et al. 2012, 2015). Prior to the onset of the summer monsoon, the Indian continent is therefore relatively dry with strong surface sensible heating, whereas the Indochina Peninsula is relatively wet with shallow convection.

Figure 2d shows a pressure–longitude cross section of the zonal circulation and the zonal deviation of the equivalent potential temperature averaged over 10°–15°N. Subject to the thermal wind balance, an easterly wind prevails in the lower troposphere, whereas a westerly flow dominates in the upper troposphere in accordance with the northward MTG in the lower troposphere and the southward temperature gradient in the upper troposphere (Fig. 2a). A zone of vertical westerly wind shear is formed around 10°–15°N above 700 hPa, which indicates that, before the onset of the monsoon, a Hadley-type meridional circulation with the ascending branch near the equator and descending branch over the subtropics should appear over the BOB domain because of the conservation of angular momentum (Plumb and Hou 1992; Wu et al. 2015, 2016). Descending flow thus predominates in the free atmosphere under the background of vertical westerly wind shear over the BOB as shown in Fig. 2d, prohibiting the summer monsoon onset. Meanwhile, the strong diabatic heating over the tropical continent leads to a positive zonal deviation of the potential temperature and warmer temperatures over land below 700 hPa. Two shallow baroclinic systems are therefore located over the Indian and Indochina peninsulas with convergence in the lower layers and divergence in the free atmosphere (shading). The Rossby wave response to diabatic heating over land induces a downdraft to the west. A typical zonal circulation pattern in the tropics with strong ascending air over land and descending air over the ocean to the west is generated in the lower troposphere before the onset of the summer monsoon. Above 700 hPa, both the ascending and descending motion become weak, indicating the limited influence of surface diabatic heating above this level.

b. Evolution of MTG and meridional circulation

The prominent distribution of northward MTG in the lower atmosphere and the reversed MTG in the upper atmosphere over the BOB prior to the onset of the BOBSM as shown in Fig. 2a has a significant impact on the general circulation. As pointed out by Plumb and Hou (1992), the atmospheric circulation can be grouped into two regimes: the thermal equilibrium regime in the mid- to high latitudes, and the angular momentum conservation (AMC) regime in the tropics. Wu et al. (2015, p. 1016) further divided the AMC regime into two categories depending on the location of external thermal forcing: the Hadley type of AMC (H-AMC) with heating near the equator, and the monsoon type of AMC (M-AMC) with heating over the subtropics. Figures 3a–c present the mean meridional circulation averaged over the eastern BOB (90°–100°E) in the climatological mean January, April, and July, respectively. In January (Fig. 3a), the wintertime equatorial heating generates an H-AMC with northerly (southerly) meridional flow prevailing over the lower (upper) troposphere. Due to the inertial impact of Earth’s rotation, easterly (westerly) accelerating wind is produced in the lower (upper) troposphere. Subject to the thermal wind relation,
uz=λTy,
where λ is a positive coefficient, the vertical westerly wind shear corresponds to the generally negative MTG in the northern tropics in winter as shown in Fig. 3a. Because vertical velocity w is proportional to the vertical derivative of horizontal advection of absolute vorticity,
w[V(f+ζ)]z,
where V and ζ are the horizontal wind and relative vorticity, respectively, and because the northerly (southerly) flow implies positive (negative) planetary vorticity (f) transport, the prevailing northerly (southerly) meridional flow in the lower (upper) troposphere within the H-AMC corresponds to a vertical decrease in absolute vorticity advection, and descending air thus prevails in the tropical area.
Fig. 3.
Fig. 3.

Climate-mean pressure–latitude cross section along the eastern BOB (90°–100°E) of meridional circulation (streamlines), sensible heating (contours; 2, 4, 6, and 8 K day−1), and MTG (shading; 10−5 K m−1) in (a) January, (b) April, and (c) July, and (d) 6 days before BOBSM onset, (e) the onset day, and (f) 6 days after onset.

Citation: Journal of Climate 36, 12; 10.1175/JCLI-D-22-0478.1

In July (Fig. 3c), the summertime subtropical heating generates an M-AMC with southerly/northerly meridional flow prevailing in the lower/upper troposphere. In a similar way to the above discussion, this M-AMC corresponds to ascending air and vertical easterly wind shear with positive MTG in the tropics. April is a month of seasonal transition from winter to summer. In this month, deep convection is still active over the equatorial BOB, while shallow convection and strong surface sensible heating have developed over Bangladesh, north of the BOB (refer to Figs. 3b and 4). Consequently, negative MTG appears in the upper troposphere while positive MTG develops in the lower troposphere in the tropical region (Figs. 2a and 3b), and the meridional circulation over the eastern BOB demonstrates a double cell structure: while the H-AMC type still dominates over the free atmosphere, a shallow M-AMC type circulation has developed over the northeastern BOB below 600 hPa. Across the latitudinal zone between 15° and 25°N, vertical easterly wind shear appears below 600 hPa and vertical westerly wind shear occurs aloft. Consequently, over the northeastern BOB, vertically ascending air is confined to the lower troposphere, while descending air is intensified in the free atmosphere owing to the combined effects of the H-AMC and M-AMC, preventing the development of deep convection and prohibiting the summer monsoon onset. This implies that the monsoon may not commence till the easterly wind develops in the upper troposphere.

Fig. 4.
Fig. 4.

Composites of daily mean SST (shading; °C), precipitation (blue contours; 6, 18, and 30 mm day−1) and 2-m surface wind (streamlines), before and after BOBSM onset: (a) D−12, (b) D−6, (c) D−2, (d) D0, (e) D0 (averaged from onset early years), and (f) D0 (averaged from onset late years).

Citation: Journal of Climate 36, 12; 10.1175/JCLI-D-22-0478.1

From Eq. (2), the zonal wind speed at elevation H in the upper troposphere (uH) and its change can be calculated as
uH=u0λ0HTydz,
uHt=u0tλ0Ht(Ty)dzu0tμ0HQydz,
in which u0 is the zonal wind speed at the surface, Q is diabatic heating, and λ and μ are positive coefficients. In April, due to the strong surface heating over India, surface westerly flow is already developing over the northeastern BOB (Fig. 2b). Equation (4) indicates that a thicker layer of stronger positive MTG is in favor of easterly vertical shear. Whereas (5) indicates that only when the vertical integral of the intensifying MTG or the positive meridional gradient of diabatic heating has offset the intensification of the surface westerly wind, can the easterly wind develop in the upper troposphere. In early spring, although a positive MTG already exists below 700 hPa over the domain of 15°–25°N (Fig. 3b) due to the north–south LSTC, it is too weak and cannot offset the wintertime southward MTG aloft. Consequently, the wintertime westerly flow still prevails over the upper troposphere (Fig. 2d). In other words, monsoon onset requires stronger northward enhancement of diabatic heating or intensifying MTG accumulated in a thicker low layer.

No matter the ASM onset occurs earlier or later, the atmospheric circulation and temperature fields both experience abrupt changes from winter type to summer type. We can therefore expect that each year by the time of the ASM onset, the MTG and the type of AMC over the ASM area should undergo rapid seasonal changes. Indeed, Figs. 3d–f demonstrates the rapid changes in the thermal field and circulation during BOBSM onset. Six days before onset, the double cell structure of meridional circulation and the MTG over the eastern BOB (Fig. 3d) are similar to their counterparts in April shown in Fig. 3b. North of 15°N, due to the shallow convection and strong surface sensible heating, a shallow M-AMC type develops below 600 hPa. The H-AMC type circulation still dominates over the free atmosphere. However, some distinct differences can be observed. As the contour of maximum temperature migrates northward in the midtroposphere, the wintertime H-AMC shrinks and the descending flow in the upper troposphere over the northeastern BOB (15°–25°N) becomes weakened. The evolution continues and, by the time of onset (Fig. 3e), in the lower troposphere, the positive MTG is further intensified and precipitation over Bangladesh is enhanced (Figs. 4b–d); whereas, in the upper troposphere, the contour of maximum temperature has migrated from near the equator to north of 10°N and the negative MTG to its north has become weakened. According to Eq. (5), easterly flow develops (Fig. 6d) in association with the occurrence of southward flow in the upper troposphere (Fig. 3e). Consequently, the double cell structure is replaced by the M-AMC type circulation over the eastern BOB, and the H-AMC type circulation diminishes. After the BOBSM onset, the contour of maximum temperature has further advanced northward (Fig. 3f), and the meridional circulation tends further toward the summertime M-AMC (Fig. 3c).

c. Evolution of atmospheric circulation during the BOBSM onset

Figure 4 shows the evolution of the daily mean SST, the surface atmospheric circulation, and precipitation relative to the date of the BOBSM onset. Before the monsoon onset, a warm pool (SST > 30°C) forms around 10°N just below the surface atmospheric anticyclonic circulation in the central BOB (Figs. 4a–c). This phenomenon is associated with the premonsoon descent of air over the sea surface in the northern BOB, with the resultant clear sky and weak sea surface winds favoring an increase in the SST (Wu et al. 2011, 2012; Xing et al. 2016a).

On D−12, a clear CEF exists between 70° and 90°E over the equatorial Indian Ocean (the BOB CEF), which is stronger than the flow in other regions (Fig. 4a). In the tropics, the circulation pattern is similar to the April-mean pattern (Fig. 2b)—that is, an anticyclone–cyclone–anticyclone located from the Arabian Sea across the Indian subcontinent to the BOB. Over the BOB, a surface zonally oriented cyclonic circulation is located between the anticyclone over the central BOB and the equatorial westerly flow (Figs. 4a–c). Therefore, before the onset of the BOBSM, the southwesterly CEF has already developed over the southern BOB but is blocked by the anticyclone to the north. From D−12 to D−6 (Figs. 4a,b), significant changes occur in the tropical circulation: the BOB CEF intensifies while another northward CEF (the Somali CEF) develops between 50° and 60°E, which reinforces the northwesterly flow over the northeastern Arabian Sea. The evolution of the CEFs at 2 m is similar to the CEFs at 925 and 850 hPa. With the strengthening of the CEFs, the equatorial westerly flow intensifies and precipitation over the southern BOB increases rapidly (Figs. 4b,c). During this period, the warm pool in the central BOB migrates slowly northward and the zonal cyclonic circulation to its south also migrates northward toward the warm SST region, resulting in an eastward retreat and shrinking of the anticyclone over the northern BOB.

The rapid transition in the tropical circulation starts around three days before the onset day. From D−2 to D0 (Figs. 4c,d), with the remarkable weakening of the northern BOB anticyclone, the tropical southwesterly flow expands rapidly northward and sweeps over the zonally oriented cyclonic belt across the central BOB. After the onset of the BOBSM, the southwesterly flow dominates the whole BOB, and the center of precipitation moves from the equator to north of 10°N.

Circulation and precipitation on D0 in the monsoon onset early and late years are also shown in Figs. 4e and 4f. The most notable difference in circulation is the intensity of the Somali CEF, which is still weak in the early onset composite but strong in the late onset composite. As a result, on the onset day, the rainfall in early years is less than normal over the eastern BOB, possibly because the Somali CEF has not established. In late years, the stronger Somali CEF intensifies the BOB CEF and the associated southwesterly flow over the BOB which transports more water vapor to the eastern BOB, maintaining heavy rainfall during the monsoon onset. Thus, the two CEFs may play different roles in the cases of early and late BOBSM onset.

The circulation in the free atmosphere also undergoes an abrupt transition. Figure 5 shows the evolution of the 700 hPa streamline and diabatic heating in the lower troposphere averaged between 850 and 500 hPa. A persistent westerly flow meanders along the southern boundary of the Iranian–Tibetan Plateau in the subtropics during this period. Before the onset of the monsoon (Figs. 5a,b), a belt of interrupted anticyclones is located between the subtropical westerly belt to the north and the continuous tropical easterly belt along 10°–15°N to the south, with anticyclonic centers located over the continental regions.

Fig. 5.
Fig. 5.

Composites of daily mean diabatic heating Q1 averaged between 850 and 500 hPa (shading; K day−1) and the 700-hPa circulation (streamlines) during the onset of the BOBSM on (a) D−6, (b) D−2, (c) D0, and (d) D+2.

Citation: Journal of Climate 36, 12; 10.1175/JCLI-D-22-0478.1

It is worth noting that the formation of this easterly belt is an atmospheric response to the surface meridional LSTC (Fig. 2a). This easterly belt not only induces the vertical westerly wind shear and descending air aloft as shown in Figs. 2d, 3b, and 3d, but also separates the equatorial westerly flow from the subtropical westerly flow, prevents the northward transportation of water vapor from the BOB to the subtropical continent, and is unfavorable for the onset of the BOBSM. This reminds us that circumspection is required when we explore how the LSTC leads to the onset of the monsoon.

Precipitation feeds back strongly to the atmospheric circulation. As displayed above, before the BOBSM onset, weak precipitation already exists in the lower troposphere over the land area north of the BOB. The latent heat release there contributes to the generation of a shallow M-AMC type circulation in the lower layer as shown in Figs. 3b and 3d. Near the equator, on D−6 and earlier, the latent heating associated with rainfall is restricted to the equatorial Indian Ocean (Fig. 5a), contributing to the maintenance of the wintertime H-AMC type circulation in the tropics (Fig. 3d). Furthermore, in response to this quasi-symmetrical diabatic heating associated with deep convection, together with the strong heating over the island of Java, a symmetrical pair of cyclones develops over the equatorial BOB (Figs. 5a,b), exhibiting a Gill-type response to forcing (Gill 1980). With the strengthening of convection, strong diabatic heating of >6 K day−1 develops over the southern BOB on D−2 (Fig. 5b) and then moves northward, accompanied by a precipitation center. From D−2 to D+2 (Figs. 5bd), the diabatic heating over the Southern Hemisphere weakens, whereas the diabatic heating over the Northern Hemisphere intensifies and shifts northward. As the northern diabatic heating center shifts to north of 10°N (Fig. 5c), the Northern Hemisphere cyclone moves northward and merges into the Indo-Burma trough, which has persisted over the northern BOB. Consequently, the wintertime continuous anticyclonic ridgeline, together with the associated easterly flow, splits over the eastern BOB, southerly flow develops in situ, and the equatorial westerly flow transports large amounts of water vapor to maintain heavy rainfall over the eastern BOB and the western Indochina Peninsula (Figs. 5c,d), leading to the onset of the BOBSM.

Figures 4 and 5 both demonstrate that, during the onset process, southwesterly flow in the lower troposphere develops gradually over the eastern BOB and precipitation is increasing over the land area to the north of the BOB. These changes are important for the monsoon onset. According to Eq. (5), the continuous intensification of diabatic heating to the north of the BOB can result in easterly wind acceleration in the upper troposphere over the BOB area, creating local easterly wind vertical shear and establishing the M-AMC type meridional circulation. Ascending motion is triggered over the eastern BOB. It is apparent that the change in tropical circulation plays a significant role in the onset of the BOBSM. When the tropical weather system intrudes into the northern BOB and the southerly flow in the lower troposphere develops over the eastern BOB, the tropical circulations in both the boundary layer and the free atmosphere undergo rapid transitions. To further reveal this process, we selected 10°–15°N as the boundary between the tropical easterlies and subtropical westerlies and display pressure–longitude cross sections of the local zonal and meridional circulations in Fig. 6. There is a prominent feature caused by the zonal LSTC: the strongest cyclonic circulation over land and the strongest anticyclonic circulation over the ocean are located near the surface before the onset of the summer monsoon. Subject to the thermal wind constraint (∂υ/∂z ∝ ∂T/∂x), vertical northerly wind shear appears over the eastern coastline of the continent and the western ocean, whereas vertical southerly shear occurs over the eastern ocean and western coastline of the continent. Consequently, a cyclonic circulation in the lower layers and an anticyclonic circulation aloft can be observed over the Indian and Indochina Peninsulas, whereas an anticyclonic circulation in the lower layers and a cyclonic circulation aloft are observed over the Arabian Sea and BOB.

Fig. 6.
Fig. 6.

Composites of the 10°–15°N average pressure–longitude cross section of the zonal circulation [vectors; zonal wind in m s−1 and vertical motion (multiplied by a factor of −60) in Pa s−1] and meridional wind (shading; m s−1) for (a) D−6, (b) D−4, (c) D−2, (d) D0, (e) D0 (averaged from onset early years), and (f) D0 (averaged from late onset late years). Black shading denotes the topography.

Citation: Journal of Climate 36, 12; 10.1175/JCLI-D-22-0478.1

As the season advances, the surface heating over the Indian Peninsula increases and the cross-equatorial southerly wind in the central Indian Ocean is intensified (Fig. 2). From D−6 (Fig. 6a), the southwesterly flow in the lower layers over the western BOB begins to expand eastward, with shrinking of the low-level northerly wind over the eastern BOB (Figs. 6b,c). On D−2 and D0, this low-level northerly wind diminishes and is replaced by a strong, deep southerly flow (Fig. 6d), whereas a distinct northerly flow develops in the upper troposphere accompanied by the intensification of easterly winds there. This is because, around D0 and afterward, as a result of the onset of the monsoon, deep convective heating develops rapidly over the Indochina Peninsula with its center located at about 500 hPa (figure not shown). According to Liu et al. (2001, 2004), the circulation and diabatic heating along the anticyclonic ridgeline can be approximated by the following Sverdrup balance:
βυ(f+ζ)θz1Qz,θz0,
where Q is diabatic heating, ζ is relative vorticity, and the other terms are conventional meteorological variables. Diabatic heating increases (decreases) with height in the lower (upper) troposphere, favoring the development of a southerly (northerly) wind. This implies that, after the center of the deep monsoon convection shifts to the subtropics, strong diabatic heating in the mid- to upper troposphere can feed back to the atmosphere, resulting in the large-scale transition of the circulation. A strong, deep southerly flow thus dominates the troposphere over the eastern BOB, linking the equatorial westerly flow to the subtropical westerly flow.

These transition features in vertical circulation can also be observed in the cases of early and late BOBSM onset composite (Figs. 6e,f). In the early onset composite, because the monsoon precipitation over the eastern BOB on D0 is weaker (Fig. 4e) than normal (Fig. 4d), the ascending motion and upper-level northerly flow are weaker (Fig. 6e). Whereas in the late onset composite, because the monsoon precipitation over the eastern BOB on D0 is stronger (Fig. 4f) than normal (Fig. 4d), the ascending motion and upper-level northerly flow are stronger as well (Fig. 6f).

The impact of the tropical land–sea distribution over South Asia on the onset of the BOBSM can therefore be divided into two stages. In the first stage, changes in circulation first occur along the equator. The BOB CEF and the Somali CEF establish successionally and strengthen the southwesterly flow over the southwestern BOB, accompanied by increased precipitation in this region. In the second stage, the near-equatorial low-level cyclonic circulation over the southern Indian subcontinent and southwestern BOB intensifies and shifts northward and eastward, leading to weakening of the anticyclone over the northern BOB. Accordingly, deep convection expands from the tropical Indian Ocean to the eastern BOB, contributing to the intensification of the southerly flow there. The equatorial westerly flow is therefore linked to the subtropical westerly flow and abundant water vapor is transported northward, contributing to the onset of the ASM.

4. Influence of air–sea feedback on the boundary layer circulation

a. Background condition before the BOBSM onset

The above results show the importance of the low-level circulation in triggering the cyclone in the lower layer and its subsequent northward movement together with deep convection. We therefore investigated the mechanism responsible for the evolution of the low-level wind. Figure 7 shows the evolution of the sea surface sensible heat flux from D−6 to D0, during which period the low-level circulation over the northern BOB experiences a sharp transition. The sensible heating is very weak (<10 W m−2) over the northern BOB, beneath the surface anticyclone on D−6 (Fig. 7a). By contrast, strong sensible heating (>16 W m−2) is observed across the central Indian Ocean beneath the strong equatorial westerly flow. Because the sea–air temperature difference in this region does not differ much from that in the tropical Indian Ocean (not shown), the intense sea surface sensible heat flux in this region is mainly a result of strengthening of the local surface westerly flow. A zone of positive relative vorticity lies to its north. The convergence of air on the eastern front of this equatorial westerly flow, together with the prominent surface heating, generates a cyclonic circulation with a positive relative vorticity in the near-surface layer (Figs. 7b,c).

Fig. 7.
Fig. 7.

Daily evolution of the 2-m surface wind (vectors; m s−1), the sea surface sensible heating flux (shading; W m−2), and the relative vorticity (contours; interval: 4 × 10−6 s−1): (a) D−6, (b) D−4, (c) D−2, and (d) D0. Heavy contours denote zero relative vorticity.

Citation: Journal of Climate 36, 12; 10.1175/JCLI-D-22-0478.1

The equatorial westerly flow and associated surface sensible heating intensify because the Somali CEF also develops during this period (Figs. 7b–d). The strengthened southwesterly flow of the eastern cyclonic circulation then increases the sensible heat flux of the sea surface and the convergence to the north, forming an air–sea feedback loop presenting as the concurrent intensification of the southerly flow, the surface sensible heat flux, and the relative vorticity. As a consequence, the area of high sensible heating gradually advances northeastward. By D0 (Fig. 7d), a remarkable positive vorticity develops, and a strong southwesterly flow prevails over the eastern BOB, and the monsoon-type circulation begins.

b. Vorticity budget

The process responsible for the transition in the low-level circulation is investigated in detail using vorticity budget analysis. Separating the vorticity tendency into the vorticity advection associated with the rotational and divergent winds and the vorticity generation associated with divergence (e.g., Haltiner 1971), the vorticity budget equation can be written as
ζt=fVxVψ(ζ+f)Vx(ζ+f),
where ζ is the relative vorticity, f is the Coriolis parameter, Vψ is the rotational wind vector, and Vx is the divergent wind vector. From left to right, the terms on the right-hand side of Eq. (7) correspond to the generation of vorticity due to convergence, the horizontal vorticity advection associated with the rotational wind, and the horizontal vorticity advection associated with the divergent wind, respectively. Imbalances in these vorticity budget terms can be interpreted as driving the evolution of the relative vorticity.

Figure 8 shows the distributions of the vorticity budget terms before and at the time of monsoon onset. The horizontal vorticity advection is divided into components associated with the rotational and divergent winds, with the zero line shown by heavy contours. From D−2 to D0, the positive relative vorticity tendency appears over the eastern BOB (Figs. 8a,b), consistent with the expansion of the southwesterly flow from the tropics (Figs. 7c,d). The contributions of the horizontal advection terms associated with both the rotational and divergent winds are negative in this area of positive relative vorticity tendency (Figs. 8e–h). The generation of vorticity associated with convergence mainly contributes to the positive vorticity tendency over the eastern BOB (Figs. 8c,d). Because the regions of surface wind convergence in Fig. 8 are in good agreement with the regions of sea surface sensible heating in Figs. 7c and 7d, we infer that the sea surface sensible heat flux is the main driver of the transition in the low-level circulation, and the horizontal advection associated with the rotational and divergent winds acts as a balancing term for the positive relative vorticity tendency.

Fig. 8.
Fig. 8.

Distribution of the terms in the vorticity budget (10−5 s−1), Eq. (7), at the 2-m surface on (left) D−2 and (right) D0: (a),(b) relative vorticity tendency ∂ζ/∂t, (c),(d) −f∇ ⋅ Vx; (e),(f) −Vψ ⋅ ∇ζβvψ; (g),(h) −Vx ⋅ ∇ζβvx.

Citation: Journal of Climate 36, 12; 10.1175/JCLI-D-22-0478.1

c. Role of CEF in triggering monsoon onset

Our analysis so far has shown that the convergence associated with the strong sea surface heat flux is the main driver of the evolution of the relative vorticity over the eastern BOB during the period in which the tropical circulation changes rapidly. This evolution of the sea surface sensible heat flux over the BOB depends on the surface equatorial westerly wind. Because the intensity of this westerly wind is closely associated with the CEF over the central Indian Ocean, the CEF may be considered as a tropical precursor of BOBSM onset.

To verify this hypothesis, we first define the region (5°S–5°N, 50°–60°E) as the area corresponding to the Somali CEF, and the region (5°S–5°N, 80°–90°E) as the area corresponding to the BOB CEF. We then calculate the area-averaged meridional wind speed at 925 hPa to represent the intensity of the CEF. Figure 9a shows the annual cycle of the area-averaged meridional wind. Figures 9b and 9c show the arithmetical mean of the annual cycle for the early onset and late onset years of the monsoon as defined in section 3. The absolute magnitude of the Somali CEF is >4 m s−1 in the winter months and approximately 10 m s−1 in the summer months, which is much stronger than the BOB CEF, particularly in summer. The northward CEF in the BOB region starts earlier than that in the Somali region by about five pentads, whereas the southward CEF in the BOB region begins later than that in the Somali region by about five pentads, indicating that the duration of the northward CEF in the BOB region is longer than that in the Somali region by about 10 pentads. In the BOB region, the southward CEF is strongest in January, then decreases gradually and changes to northward in early spring, after which its intensity increases continuously.

Fig. 9.
Fig. 9.

Annual cycle of the CEF intensity (m s−1) for (a) the climatology, (b) the early years, and (c) the late years. Red and purple curves denote the area-average over (5°S5°N, 50°60°E) and (5°S5°N, 80°90°E), respectively. The green curve denotes the area-mean precipitation (mm day−1) averaged over (10°20°N, 90°100°E). The black dashed vertical line indicates the monsoon onset pentad.

Citation: Journal of Climate 36, 12; 10.1175/JCLI-D-22-0478.1

When the intensity reaches approximately 2 m s−1, precipitation over the eastern BOB area (Fig. 9, green curves) increases rapidly and BOBSM onset occurs. This indicates that the time at which the BOB CEF reaches the 2 m s−1 threshold matches well with the onset date of the BOBSM. This is true for the climatology (pentad 25), the early onset (pentad 21), and the late onset (pentad 27) years (Figs. 9b,c). In the early BOBSM onset years, only the BOB CEF exists and there is no Somali CEF. The increasing speed of the BOB CEF from 0 to 2 m s−1 in spring is faster in early onset years (Fig. 9b) than in normal years (Fig. 9a) and slower in late onset years (Fig. 9c). All these results indicate that the intensity of the BOB CEF may be an important tropical precursor for the onset of the BOBSM, whereas the Somali CEF may reinforce the BOBSM. Strong CEFs turn eastward as a result of the Coriolis force and strengthen the equatorial westerly flow, providing suitable background conditions for air–sea feedback as discussed above.

5. Role of the South Asian continent in the onset of the monsoon

Since the CEFs play important roles in the summer monsoon onset process, in this section we analyze the factors affecting the CEF and the onset of the BOBSM. Previous research has shown that the interhemispheric thermal contrast and the differences between surface characteristics, including the land–sea distribution and terrain, can affect the formation and intensity of the CEF (Krishnamurti et al. 1976; Rodwell and Hoskins 1995; He et al. 2007; Kitoh 2017; Xu et al. 2010, and references cited therein). During the transition to the monsoon-type circulation, the intertropical convergence zone in the tropical Indian Ocean disappears because of the development of the strong surface CEF from the winter hemisphere to the summer hemisphere, which transports abundant water vapor to higher latitudes, resulting in the occurrence of a tropical rainy season. This section explores the effect of the South Asian continent on the CEFs and the onset of the BOBSM using CESM2.

a. Model validation

Before we compare the results of the CTRL and sensitivity experiments, we first examine the model’s performance in simulating the ASM onset. Figures 10a and 10b are Hovmöller diagrams of the precipitation averaged over 10°–20°N using TRMM data and the result of the CTRL experiment. The rainy season for tropical Asia in CESM2 (Fig. 10b) first occurs over the eastern BOB and the western Indochina Peninsula at the end of April to early May, then extends eastward to the SCS in mid-May, in good agreement with the observations (Fig. 10a). In the monsoon regions, the centers of precipitation are always located on the windward side of high terrain, consistent with the results of Wang and Chang (2012).

Fig. 10.
Fig. 10.

Hovmöller diagrams of the daily mean precipitation averaged over 10°–20°N (shading; mm day−1) in (a) TRMM, and in the numerical experiments of (b) the CTRL run, (c) the No_Indian run, (d) the No_Inch run, and (e) the No_Inch_Indian run.

Citation: Journal of Climate 36, 12; 10.1175/JCLI-D-22-0478.1

The intense precipitation over the eastern BOB after the onset of the monsoon in the observations (Fig. 10a) is simulated well in the CTRL experiment (Fig. 10b). The deep convection over the eastern BOB induces a strong downdraft to the west, blocking the westward spread of the rain belt from the BOB. The onset of the monsoon in India therefore occurs in late May, as reported by Liu et al. (2015). This phenomenon is also captured by the model. Although the model shows deviations in simulating the detailed pattern of precipitation, mainly due to its coarse resolution, in general it can capture the main features of the onset of the ASM and its evolution when compared with the TRMM data. The model could therefore be used to conduct sensitivity experiments to investigate the effect of the land–sea distribution on the onset of the BOBSM.

b. Effect of LSTC on the onset of the BOBSM

To investigate the role of the LSTC in the onset of the BOBSM, the land grid points comprising the tropical continent of the Indian and Indochina peninsulas around the BOB were replaced by the ocean grid individually and then jointly to execute sensitivity experiments (as described in section 2). Figures 10c–e are Hovmöller diagrams of the precipitation in the sensitivity experiments. In the No_Indian experiment, in which the Indian Peninsula is replaced by ocean, the rainy season initially occurs over the SCS and then propagates westward continuously, and reaches the eastern BOB by late-May, which is delayed by about 20 days (Fig. 10c). Meanwhile, the blocking phenomenon of the westward propagation of the rain belt along 80°E observed in TRMM and CTRL disappears when the Indian subcontinent is removed.

In the No_Inch experiment, the earliest onset of the monsoon also occurs over the SCS in mid-May (Fig. 10d), and then propagates westward. The BOBSM onset is delayed to late-May but the blocking of the westward propagation of the rain belt along 80°E is simulated well when the Indian subcontinent is retained. In the No_Inch_Indian experiment, the rainy season first occurs over the SCS but is slightly later than in the other experiments. The BOBSM onset is further delayed until June. The above experimental results indicate that both the ASM onset timing and the onset sequence are sensitive to the distribution of the land in tropical Asia. After the tropical continent around the BOB is removed, the BOBSM onset is significantly delayed.

To understand how the removal of the Indian continent can delay the BOBSM onset, in the following analysis we focus on the circulation difference between the CTRL experiment and the No_Indian experiment. In April, the warm Indian Peninsula in the CTRL experiment generates a low pressure in the lower troposphere over the central Indian Peninsula, resulting in a pronounced northward pressure gradient force in the tropical north Indian Ocean and along the equator (Fig. 11a). The strongest equatorial northward pressure gradient force is located over the central Indian Ocean in accordance with the strongest in situ interhemispheric surface thermal contrast. Consequently, the earliest CEF occurs over the central Indian Ocean (Figs. 11c–f). Figure 11b shows the annual cycle of the BOB CEF and precipitation over the eastern BOB region (10°–20°N, 90°–100°E) in the No_Indian experiment compared with the CTRL experiment. Taking 2 m s−1 as the threshold of the onset of the BOB monsoon, as in the data analysis, the onset date in the CTRL experiment is around the end of April and the beginning of May (pentads 25–26). When the Indian Peninsula is removed in the No_Indian experiment, the intensity of the BOB CEF becomes significantly weaker than in the CTRL experiment in spring between pentads 10 and 30 because the northward meridional pressure gradient is significantly weakened between 70° and 90°E (Fig. 11a) and the onset of the BOB wet season is also delayed.

Fig. 11.
Fig. 11.

(a) Difference in the April-mean meridional geopotential gradient at 925 hPa (shading; 10−6 m s−2) between the CTRL and No_Indian experiments. (b) Annual cycle of the BOB CEF (m s−1) at 925 hPa (red curve) and the area-averaged precipitation (mm day−1) in the eastern BOB (10°–20°N, 90°–100°E; green curve) as shown in panel (a); the solid and dashed curves indicate the CTRL and No_Indian experiments, respectively. (c)–(f) Pentad-averaged differences in the wind at 925 hPa (vectors; m s−1) between the CTRL and No_Indian experiments. (g),(h) Pressure–latitude sections of the meridional circulation (streamlines) and MTG (shading; 10−5 K m−1) over the eastern BOB (90°–100°E) at the times of onset in the CTRL and No_Indian experiments, respectively.

Citation: Journal of Climate 36, 12; 10.1175/JCLI-D-22-0478.1

To understand how the Indian continent affects the transition in the low-level circulation, the difference in the pentad-mean 925-hPa wind between the CTRL experiment and the No_Indian sensitivity experiment is shown in Figs. 11c–f. The low-level cyclonic circulation induced by heating of the Indian subcontinent is already prominent in early April (Fig. 11c) and strengthens rapidly as the season evolves. By pentad 23 (Fig. 11d), this cyclonic circulation has connected to the CEF over the central Indian Ocean and the southwesterly flow over the western BOB has developed, leading to weakening of the anticyclone over the eastern BOB. This southwesterly flow then forms a coalition between the cyclone over India and the BOB CEF and intensifies rapidly over the western and central BOB (Figs. 11e,f). By early May, the southwesterly flow has swept over the whole of the BOB, reaching the western coast of the Indochina Peninsula. Using the MTG index described in section 2, we select the onset date in the CTRL experiment and present a pressure–latitude cross section of the circulation and MTG over the eastern BOB (90°–100°E). On the onset day, despite the model bias in simulating the meridional circulation, the dominance of the southerly flow in the lower troposphere over the eastern BOB is reasonably represented. In the temperature distribution, the maximum positive MTG in the lower troposphere already appears to the north of 20°N, and the contour of maximum temperature in the free atmosphere is already located to the north of 10°N. The main feature of the thermal field is similar to its counterparts on D−6 and D0 during the monsoon onset in the reanalysis as shown in Figs. 3d and 3e, indicating that the upper-tropospheric circulation is under transition to the monsoon-type AMC. In the No_Indian experiment at the same time, the wintertime Hadley cell is surviving, and the associated convergence zone is still near the equator. Northerly wind still prevails in the lower levels. These differences in the experiments suggest that the CEF in the lower troposphere is an important tropical precursor for the establishment of monsoon-type circulation. The results thus confirm our earlier conclusions obtained from data analysis that the meridional thermal contrast across the equator and the zonal LSTC across tropical Asia are important for the onset of the BOBSM.

6. Summary and discussion

After the spring equinox, the surface temperature along the southern coast of Asia is warmer than that over the north Indian Ocean (Fig. 2a) as a result of the faster increase in incoming solar radiation over land than over the ocean. This meridional LSTC over the northern BOB is too weak in early spring and can only enhance the easterly flow in the lower troposphere over the central north Indian Ocean (Figs. 2d and 4a), inducing vertical westerly wind shear and descending air aloft and preventing the transport of water vapor from the tropics to the subtropics, and therefore does not favor the commencement of the ASM. The onset of the BOBSM requires extra forcing to create an appropriate environment for the rapid circulation changes. Previous studies have shown the importance of surface sensible heating over the Indian Peninsula and of the warm sea surface of the central BOB to the onset of the BOBSM. This study used composite analysis based on MERRA-2 data and numerical sensitivity experiments with an Earth system model (CESM2) to explore how the zonal LSTC over South Asia and the meridional LSTC across the equator in spring control the onset of the ASM.

The climatological onset of the BOBSM is in early May. The tropical Asian continent is distinctly warmer than the surrounding ocean before the onset of the monsoon. Strong sensible heating over the Indian and Indochina peninsulas leads to a cyclonic circulation in the lower layers and an anticyclonic circulation aloft, accompanied by ascending air over the land. An anticyclonic circulation in the lower layers and a cyclonic circulation aloft are accompanied by downward motion over the northern BOB. This circulation pattern is insufficient for the onset of the monsoon over the eastern BOB. The anticyclone over the northern BOB gradually decays as the CEF develops over the central Indian Ocean and the cyclonic circulation surrounding India is intensified. The southwesterly flow to the west of the anticyclone is enhanced and expands eastward, whereas the northerly flow to the east shrinks gradually on both horizontal and vertical scales and is eventually replaced by a southerly flow by the time of the onset of the monsoon (Fig. 6). Deep southerly winds develop strongly from the surface to the upper troposphere and the onset of the summer monsoon occurs over the eastern BOB. This study has shed new light on the underlying mechanism responsible for the onset of the ASM, as summarized schematically in Fig. 12.

  1. Thermal contrast across the equator. Across the equatorial Indian Ocean, the strongest meridional LSTC is located over the central Indian Ocean, where the Indian subcontinent is located just to the north of the equator. After the spring equinox, increased solar radiation efficiently heats the Indian Peninsula, which enhances the interhemispheric geopotential gradient. The earliest northward CEF (the BOB CEF) near the surface develops in this region (Fig. 7a). The southerly flow turns eastward as a result of the Coriolis force, becoming the remarkable springtime near-equatorial surface zonal flow (Figs. 7 and 12a). A prominent belt of surface sensible heating is formed beneath the surface zonal flow and a zone of surface positive vorticity is located to the north of this zonal flow.

  2. Asymmetrical atmospheric heating along the equator. The equatorial convection, together with diabatic heating over the island of Java in spring, triggers a pair of cyclonic circulations straddling the equator in the central Indian Ocean at 700 hPa, which intensifies the near-equatorial zonal flow in the lower layers (Fig. 5). However, the easterly wind on the north side of the northern cyclone is located at 10°–15°N (Fig. 12a), which separates the equatorial westerly flow from the subtropical westerly flow on the southern side of the Tibetan Plateau and prevents the onset of the monsoon.

  3. Warm SST pool of the BOB. Before the onset of the monsoon, a warm SST pool is formed near 10°N over the northern BOB (Fig. 4) as a result of forcing by the Tibetan Plateau and air–sea interactions over northern Indian Ocean (Wu et al. 2011, 2012). Subject to the thermal wind constraint, vertical westerly wind shear is located to the north of the warm pool (Fig. 12a), leading to weakening of the easterly wind aloft along the 10°–15°N latitude belt. At the same time, as the surface sensible heating and the westerly flow in the southern BOB are enhanced and migrate northward, the surface relative vorticity increases in front of the westerly jet (Fig. 7). The existing anticyclonic circulation over the northern and eastern BOB is therefore weakened.

  4. Reinforcement of the LSTC and onset of the BOBSM. The strong surface sensible heating over the Indian continent in spring and the associated forcing by the Tibetan Plateau generate a prominent cyclonic circulation in the lower troposphere. This continental cyclone is rapidly enhanced as summer approaches and relates to the BOB CEF over the central Indian Ocean. At the same time, the development of the Somali CEF reinforces the impact of the LSTC in both the zonal and meridional directions (Figs. 4c,d, 7c,d, and 12b). Consequently, the southwesterly flow over the western BOB develops and expands eastward, while the northerly flow over the eastern BOB shrinks and is eventually replaced by a southerly wind. During this process, the center of deep convection over the tropical Indian Ocean moves to the eastern BOB. As a response to this diabatic heating, a southerly wind dominates almost the whole troposphere. The wintertime tropical anticyclonic belt, together with the continuous easterly belt to its south, split over the eastern BOB and the equatorial westerly flow merges into the higher-latitude westerly belt, transporting large amounts of moisture from the Indian Ocean to subtropical Asia, resulting in heavy precipitation in situ (Fig. 11b). The onset of the ASM therefore begins over the eastern BOB.

  5. BOB CEF. The CEF over the central Indian Ocean can be regarded as a tropical precursor for the onset of the BOBSM. In early (late) monsoon onset years, the intensity of the BOB CEF is anomalously stronger (weaker) in spring and the transition of the circulation from winter to summer is earlier (later) than the climatological mean.

  6. Differences between early and late monsoon onset years. Significant differences in the early and late years of monsoon onset are reflected in the presence or absence of Somali CEF. In the early years, the Somali CEF has not developed and the rainfall over the eastern BOB is weaker than climatology, which is opposite in the late years. Thus, the two CEFs play different roles in the BOBSM onset.

Fig. 12.
Fig. 12.

Schematic diagram showing the roles of the land–sea thermal contrast in the meridional direction at the equator and in the zonal direction across the tropics in the onset of the BOBSM. (a) In early April, CEF first occurs over the central Indian Ocean (BOB CEF), which generates a band of westerly flow and a surface sensible heat flux in the southern BOB with cyclonic vorticity to the north. The zonal contrast generates a warm SST pool in the central BOB, which weakens the anticyclonic circulation and easterly flow to the north in the lower troposphere. (b) Before the onset of the BOBSM, the development of the Somali CEF enhances the impacts of the land–sea thermal contrast in both the meridional and zonal directions, leading to the onset of the BOBSM; see text for details.

Citation: Journal of Climate 36, 12; 10.1175/JCLI-D-22-0478.1

This study investigated the impact of the land–sea distribution over South Asia on the onset of the ASM by analyzing the physical processes, mainly in the lower troposphere. The onset of the monsoon requires coupling of the atmospheric circulation between the upper and lower troposphere. In the sensitivity experiments, it was also noticed that when the Indochina Peninsula is removed, the BOBSM onset is also delayed significantly. The mechanism of how the Indochina Peninsula can influence the BOBSM onset, and the land–sea distribution in South Asia can influence the upper tropospheric circulation during the onset of the ASM, requires further study.

Acknowledgments.

This study was jointly supported by the Guangdong Major Project of Basic and Applied Basic Research (2020B0301030004) and the Strategic Priority Research Program of Chinese Academy of Sciences (XDB40030204, XDB40030205).

Data availability statement.

The MERRA-2 reanalysis data are available from https://disc.gsfc.nasa.gov/datasets?project=MERRA-2. The TRMM precipitation data are available from https://disc2.gesdisc.eosdis.nasa.gov/data/TRMM_L3/TRMM_3B42/. The OISST data are available from https://psl.noaa.gov/data/gridded/data.noaa.oisst.v2.highres.html.

REFERENCES

  • Adcroft, A. J., and Coauthors, 2019: The GFDL global ocean and sea ice model OM4.0: Model description and simulation features. J. Adv. Model. Earth Syst., 11, 31673211, https://doi.org/10.1029/2019MS001726.

    • Search Google Scholar
    • Export Citation
  • Bombardi, R. J., V. Moron, and J. S. Goodnight, 2020: Detection, variability, and predictability of monsoon onset and withdrawal dates: A review. Int. J. Climatol., 40, 641667, https://doi.org/10.1002/joc.6264.

    • Search Google Scholar
    • Export Citation
  • Chen, Y., Y.-H. Ding, Z.-N. Xiao, and H.-Y. Yan, 2006: The impact of water vapor transport on the summer monsoon onset and abnormal rainfall over Yunnan province in May (in Chinese). Chin. J. Atmos. Sci., 30, 2537, https://doi.org/10.3878/j.issn.1006-9895.2006.01.03.

    • Search Google Scholar
    • Export Citation
  • Danabasoglu, G., and Coauthors, 2020: The Community Earth System Model version 2 (CESM2). J. Adv. Model. Earth Syst., 12, e2019MS001916, https://doi.org/10.1029/2019MS001916.

    • Search Google Scholar
    • Export Citation
  • Ding, Y., and Y. Liu, 2001: Onset and the evolution of the summer monsoon over the South China Sea during SCSMEX field experiment in 1998. J. Meteor. Soc. Japan, 79, 255276, https://doi.org/10.2151/jmsj.79.255.

    • Search Google Scholar
    • Export Citation
  • Feng, J., D. Hu, and L. Yu, 2013: Role of western Pacific oceanic variability in the onset of the Bay of Bengal summer monsoon. Adv. Atmos. Sci., 30, 219234, https://doi.org/10.1007/s00376-012-2040-9.

    • Search Google Scholar
    • Export Citation
  • Gelaro, R., and Coauthors, 2017: The Modern-Era Retrospective Analysis for Research and Applications, version 2 (MERRA-2). J. Climate, 30, 54195454, https://doi.org/10.1175/JCLI-D-16-0758.1.

    • Search Google Scholar
    • Export Citation
  • Gill, A. E., 1980: Some simple solutions for heat‐induced tropical circulation. Quart. J. Roy. Meteor. Soc., 106, 447462, https://doi.org/10.1002/qj.49710644905.

    • Search Google Scholar
    • Export Citation
  • Haltiner, G. J., 1971: Numerical Weather Prediction. John Wiley and Sons, 317 pp.

  • He, J., J. Ju, Z. Wen, J. , and Q. Jin, 2007: A review of recent advances in research on Asian monsoon in China. Adv. Atmos. Sci., 24, 972992, https://doi.org/10.1007/s00376-007-0972-2.

    • Search Google Scholar
    • Export Citation
  • Hoskins, B. I., 1991: Towards a PV-θ view of the general circulation. Tellus, 43A, 2735, http://doi.org/10.3402/tellusa.v43i4.11936.

    • Search Google Scholar
    • Export Citation
  • Hu, P., W. Chen, S. Chen, Y. Liu, L. Wang, and R. Huang, 2022: The leading mode and factors for coherent variations among the subsystems of tropical Asian summer monsoon onset. J. Climate, 35, 15971612, https://doi.org/10.1175/JCLI-D-21-0101.1.

    • Search Google Scholar
    • Export Citation
  • Huang, B., C. Liu, V. Banzon, E. Freeman, G. Graham, B. Hankins, T. Smith, and H.-M. Zhang, 2021: Improvements of the Daily Optimum Interpolation Sea Surface Temperature (DOISST) version 2.1. J. Climate, 34, 29232939, https://doi.org/10.1175/JCLI-D-20-0166.1.

    • Search Google Scholar
    • Export Citation
  • Huffman, G. J., and Coauthors, 2007: The TRMM Multi-Satellite Precipitation Analysis (TMPA): Quasi-global, multiyear, combined-sensor precipitation estimates at fine scales. J. Hydrometeor., 8, 3855, https://doi.org/10.1175/JHM560.1.

    • Search Google Scholar
    • Export Citation
  • Hunke, E. C., W. H. Lipscomb, A. K. Turner, N. Jeffery, and S. Elliott, 2015: CICE: The Los Alamos Sea ice model. Documentation and software user’s manual version 5.1. Tech. Rep. LA‐CC‐06‐012, T‐3 Fluid Dynamics Group, Los Alamos National Laboratory, 116 pp., https://www.ccpo.odu.edu/∼klinck/Reprints/PDF/cicedoc2015.pdf.

  • Jiang, X., and J. Li, 2011: Influence of the annual cycle of sea surface temperature on the monsoon onset. J. Geophys. Res., 116, D10105, https://doi.org/10.1029/2010JD015236.

    • Search Google Scholar
    • Export Citation
  • Kitoh, A., 2017: The Asian monsoon and its future change in climate models: A review. J. Meteor. Soc. Japan, 95, 733, https://doi.org/10.2151/jmsj.2017-002.

    • Search Google Scholar
    • Export Citation
  • Krishnamurti, T. N., J. Molinari, and H. L. Pan, 1976: Numerical simulation of the Somali jet. J. Atmos. Sci., 33, 23502362, https://doi.org/10.1175/1520-0469(1976)033<2350:NSOTSJ>2.0.CO;2.

    • Search Google Scholar
    • Export Citation
  • Lau, K. M., and S. Yang, 1997: Climatology and interannual variability of the Southeast Asian summer monsoon. Adv. Atmos. Sci., 14, 141162, https://doi.org/10.1007/s00376-997-0016-y.

    • Search Google Scholar
    • Export Citation
  • Lau, K.-M., H.-T. Wu, and S. Yang, 1998: Hydrologic processes associated with the first transition of the Asian summer monsoon: A pilot satellite study. Bull. Amer. Meteor. Soc., 79, 18711882, https://doi.org/10.1175/1520-0477(1998)079<1871:HPAWTF>2.0.CO;2.

    • Search Google Scholar
    • Export Citation
  • Lawrence, D. M., and Coauthors, 2019: The Community Land Model version 5: Description of new features, benchmarking, and impact of forcing uncertainty. J. Adv. Model. Earth Syst., 11, 42454287, https://doi.org/10.1029/2018MS001583.

    • Search Google Scholar
    • Export Citation
  • Li, C.-Z., S. Yang, W. Mo, J. Zhang, and W. Wei, 2022: Seasonal predictions for May rainfall over southern China by the NCEP CFSv2. J. Trop. Meteor., 28, 2944, https://doi.org/10.46267/j.1006-8775.2022.003.

    • Search Google Scholar
    • Export Citation
  • Li, J., and L. Zhang, 2009: Wind onset and withdrawal of Asian summer monsoon and their simulated performance in AMIP models. Climate Dyn., 32, 935968, https://doi.org/10.1007/s00382-008-0465-8.

    • Search Google Scholar
    • Export Citation
  • Li, K., Y. Liu, Z. Li, Y. Yang, L. Feng, S. Khokiattiwong, W. Yu, and S. Liu, 2018: Impacts of ENSO on the Bay of Bengal summer monsoon onset via modulating the intra-seasonal oscillation. Geophys. Res. Lett., 45, 52205228, https://doi.org/10.1029/2018GL078109.

    • Search Google Scholar
    • Export Citation
  • Lipscomb, W. H., and Coauthors, 2019: Description and evaluation of the Community Ice Sheet Model (CISM) v. 2.1. Geosci. Model Dev., 12, 387424, https://doi.org/10.5194/gmd-12-387-2019.

    • Search Google Scholar
    • Export Citation
  • Liu, B., G. Wu, J. Mao, and J. He, 2013: Genesis of the South Asian high and its impact on the Asian summer monsoon onset. J. Climate, 26, 29762991, https://doi.org/10.1175/JCLI-D-12-00286.1.

    • Search Google Scholar
    • Export Citation
  • Liu, B., Y. Liu, G. Wu, J. Yan, J. He, and S. Ren, 2015: Asian summer monsoon onset barrier and its formation mechanism. Climate Dyn., 45, 711726, https://doi.org/10.1007/s00382-014-2296-0.

    • Search Google Scholar
    • Export Citation
  • Liu, Y., J. C. L. Chan, J. Mao, and G. Wu, 2002: The role of Bay of Bengal convection in the onset of the 1998 South China Sea summer monsoon. Mon. Wea. Rev., 130, 27312744, https://doi.org/10.1175/1520-0493(2002)130<2731:TROBOB>2.0.CO;2.

    • Search Google Scholar
    • Export Citation
  • Liu, Y., G. Wu, and R. Ren, 2004: Relationship between the subtropical anticyclone and diabatic heating. J. Climate, 17, 682698, https://doi.org/10.1175/1520-0442(2004)017<0682:RBTSAA>2.0.CO;2.

    • Search Google Scholar
    • Export Citation
  • Liu, Y. M., G. X. Wu, H. Liu, and P. Liu, 2001: Condensation heating of the Asian summer monsoon and the subtropical anticyclone in the Eastern Hemisphere. Climate Dyn., 17, 327338, https://doi.org/10.1007/s003820000117.

    • Search Google Scholar
    • Export Citation
  • Mao, J., and G. Wu, 2007: Interannual variability in the onset of the summer monsoon over the eastern Bay of Bengal. Theor. Appl. Climatol., 89, 155170, https://doi.org/10.1007/s00704-006-0265-1.

    • Search Google Scholar
    • Export Citation
  • Mao, J., J. C. L. Chan, and G. Wu, 2004: Relationship between the onset of the South China Sea summer monsoon and the structure of the Asian subtropical anticyclone. J. Meteor. Soc. Japan, 82, 845859, https://doi.org/10.2151/jmsj.2004.845.

    • Search Google Scholar
    • Export Citation
  • Plumb, R. A., and A. Y. Hou, 1992: The response of a zonally symmetric atmosphere to subtropical thermal forcing: Threshold behavior. J. Atmos. Sci., 49, 17901799, https://doi.org/10.1175/1520-0469(1992)049<1790:TROAZS>2.0.CO;2.

    • Search Google Scholar
    • Export Citation
  • Rodwell, M. J., and B. J. Hoskins, 1995: A model of the Asian summer monsoon. Part II: Cross-equatorial flow and PV behavior. J. Atmos. Sci., 52, 13411356, https://doi.org/10.1175/1520-0469(1995)052<1341:AMOTAS>2.0.CO;2.

    • Search Google Scholar
    • Export Citation
  • Wang, Z., and C.-P. Chang, 2012: A numerical study of the interaction between the large-scale monsoon circulation and orographic precipitation over South and Southeast Asia. J. Climate, 25, 24402455, https://doi.org/10.1175/JCLI-D-11-00136.1.

    • Search Google Scholar
    • Export Citation
  • Wu, C., S. Yang, A. Wang, and S. Fong, 2005: Effect of condensational heating over the Bay of Bengal on the onset of South China Sea monsoon in 1998. Meteor. Atmos. Phys., 90, 3747, https://doi.org/10.1007/s00703-005-0115-1.

    • Search Google Scholar
    • Export Citation
  • Wu, G., and Y. Zhang, 1998: Tibetan Plateau forcing and the timing of the monsoon onset over South Asia and the South China Sea. Mon. Wea. Rev., 126, 913927, https://doi.org/10.1175/1520-0493(1998)126<0913:TPFATT>2.0.CO;2.

    • Search Google Scholar
    • Export Citation
  • Wu, G., and Y. Liu, 2002: Thermal adaptation, overshooting, dispersion, and subtropical anticyclone Part I: Thermal adaptation and overshooting (in Chinese). Chin. J. Atmos. Sci., 24, 433446.

    • Search Google Scholar
    • Export Citation
  • Wu, G., Y. Guan, T. Wang, Y. Liu, J. Yan, and J. Mao, 2011: Vortex genesis over the Bay of Bengal in spring and its role in the onset of the Asian summer monsoon. Sci. China Earth Sci., 54 (1), 19, https://doi.org/10.1007/s11430-010-4125-6.

    • Search Google Scholar
    • Export Citation
  • Wu, G., Y. Guan, Y. Liu, J. Yan, and J. Mao, 2012: Air–sea interaction and formation of the Asian summer monsoon onset vortex over the Bay of Bengal. Climate Dyn., 38, 261279, https://doi.org/10.1007/s00382-010-0978-9.

    • Search Google Scholar
    • Export Citation
  • Wu, G., and Coauthors, 2015: Tibetan Plateau climate dynamics: Recent research progress and outlook. Natl. Sci. Rev., 2, 100116, https://doi.org/10.1093/nsr/nwu045.

    • Search Google Scholar
    • Export Citation
  • Wu, G., B. He, Y. Liu, Q. Bao, R. Ren, and B. Liu, 2016: Recent progresses on dynamics of the Tibetan Plateau and Asian summer monsoon (in Chinese). Chin. J. Atmos. Sci., 40, 2232, https://doi.org/10.3878/j.issn.1006-9895.1504.15129.

    • Search Google Scholar
    • Export Citation
  • Wu, X., and J. Mao, 2019: Decadal changes in interannual dependence of the Bay of Bengal summer monsoon onset on ENSO modulated by the Pacific decadal Oscillation. Adv. Atmos. Sci., 36, 14041416, https://doi.org/10.1007/s00376-019-9043-8.

    • Search Google Scholar
    • Export Citation
  • Xing, N., J. Li, X. Jiang, and L. Wang, 2016a: Local oceanic precursors for the summer monsoon onset over the Bay of Bengal and the underlying processes. J. Climate, 29, 84558470, https://doi.org/10.1175/JCLI-D-15-0825.1.

    • Search Google Scholar
    • Export Citation
  • Xing, N., J. Li, and L. Wang, 2016b: Effect of the early and late onset of summer monsoon over the Bay of Bengal on Asian precipitation in May. Climate Dyn., 47, 19611970, https://doi.org/10.1007/s00382-015-2944-z.

    • Search Google Scholar
    • Export Citation
  • Xu, Z., Y. Qian, and C. Fu, 2010: The role of land-sea distribution and orography in the Asian monsoon. Part II: Orography. Adv. Atmos. Sci., 27, 528542, https://doi.org/10.1007/s00376-009-9045-z.

    • Search Google Scholar
    • Export Citation
  • Yan, H., Z. Xiao, and L. Wang, 2003: Activities of Bay of Bengal monsoon and beginning date of rain season in Yunnan (in Chinese). Plateau Meteor., 22, 624630.

    • Search Google Scholar
    • Export Citation
  • Yanai, M., S. Esbensen, and J.-H. Chu, 1973: Determination of bulk properties of tropical cloud clusters from large-scale heat and moisture budgets. J. Atmos. Sci., 30, 611627, https://doi.org/10.1175/1520-0469(1973)030<0611:DOBPOT>2.0.CO;2.

    • Search Google Scholar
    • Export Citation
  • Yu, W.-D., J.-W. Shi, L. Liu, K.-P. Li, Y.-L. Liu, and H.-W. Wang, 2012: The onset of the monsoon over the Bay of Bengal: The observed common features for 2008–2011. Atmos. Ocean. Sci. Lett., 5, 314318, https://doi.org/10.1080/16742834.2012.11447009.

    • Search Google Scholar
    • Export Citation
Save
  • Adcroft, A. J., and Coauthors, 2019: The GFDL global ocean and sea ice model OM4.0: Model description and simulation features. J. Adv. Model. Earth Syst., 11, 31673211, https://doi.org/10.1029/2019MS001726.

    • Search Google Scholar
    • Export Citation
  • Bombardi, R. J., V. Moron, and J. S. Goodnight, 2020: Detection, variability, and predictability of monsoon onset and withdrawal dates: A review. Int. J. Climatol., 40, 641667, https://doi.org/10.1002/joc.6264.

    • Search Google Scholar
    • Export Citation
  • Chen, Y., Y.-H. Ding, Z.-N. Xiao, and H.-Y. Yan, 2006: The impact of water vapor transport on the summer monsoon onset and abnormal rainfall over Yunnan province in May (in Chinese). Chin. J. Atmos. Sci., 30, 2537, https://doi.org/10.3878/j.issn.1006-9895.2006.01.03.

    • Search Google Scholar
    • Export Citation
  • Danabasoglu, G., and Coauthors, 2020: The Community Earth System Model version 2 (CESM2). J. Adv. Model. Earth Syst., 12, e2019MS001916, https://doi.org/10.1029/2019MS001916.

    • Search Google Scholar
    • Export Citation
  • Ding, Y., and Y. Liu, 2001: Onset and the evolution of the summer monsoon over the South China Sea during SCSMEX field experiment in 1998. J. Meteor. Soc. Japan, 79, 255276, https://doi.org/10.2151/jmsj.79.255.

    • Search Google Scholar
    • Export Citation
  • Feng, J., D. Hu, and L. Yu, 2013: Role of western Pacific oceanic variability in the onset of the Bay of Bengal summer monsoon. Adv. Atmos. Sci., 30, 219234, https://doi.org/10.1007/s00376-012-2040-9.

    • Search Google Scholar
    • Export Citation
  • Gelaro, R., and Coauthors, 2017: The Modern-Era Retrospective Analysis for Research and Applications, version 2 (MERRA-2). J. Climate, 30, 54195454, https://doi.org/10.1175/JCLI-D-16-0758.1.

    • Search Google Scholar
    • Export Citation
  • Gill, A. E., 1980: Some simple solutions for heat‐induced tropical circulation. Quart. J. Roy. Meteor. Soc., 106, 447462, https://doi.org/10.1002/qj.49710644905.

    • Search Google Scholar
    • Export Citation
  • Haltiner, G. J., 1971: Numerical Weather Prediction. John Wiley and Sons, 317 pp.

  • He, J., J. Ju, Z. Wen, J. , and Q. Jin, 2007: A review of recent advances in research on Asian monsoon in China. Adv. Atmos. Sci., 24, 972992, https://doi.org/10.1007/s00376-007-0972-2.

    • Search Google Scholar
    • Export Citation
  • Hoskins, B. I., 1991: Towards a PV-θ view of the general circulation. Tellus, 43A, 2735, http://doi.org/10.3402/tellusa.v43i4.11936.

    • Search Google Scholar
    • Export Citation
  • Hu, P., W. Chen, S. Chen, Y. Liu, L. Wang, and R. Huang, 2022: The leading mode and factors for coherent variations among the subsystems of tropical Asian summer monsoon onset. J. Climate, 35, 15971612, https://doi.org/10.1175/JCLI-D-21-0101.1.

    • Search Google Scholar
    • Export Citation
  • Huang, B., C. Liu, V. Banzon, E. Freeman, G. Graham, B. Hankins, T. Smith, and H.-M. Zhang, 2021: Improvements of the Daily Optimum Interpolation Sea Surface Temperature (DOISST) version 2.1. J. Climate, 34, 29232939, https://doi.org/10.1175/JCLI-D-20-0166.1.

    • Search Google Scholar
    • Export Citation
  • Huffman, G. J., and Coauthors, 2007: The TRMM Multi-Satellite Precipitation Analysis (TMPA): Quasi-global, multiyear, combined-sensor precipitation estimates at fine scales. J. Hydrometeor., 8, 3855, https://doi.org/10.1175/JHM560.1.

    • Search Google Scholar
    • Export Citation
  • Hunke, E. C., W. H. Lipscomb, A. K. Turner, N. Jeffery, and S. Elliott, 2015: CICE: The Los Alamos Sea ice model. Documentation and software user’s manual version 5.1. Tech. Rep. LA‐CC‐06‐012, T‐3 Fluid Dynamics Group, Los Alamos National Laboratory, 116 pp., https://www.ccpo.odu.edu/∼klinck/Reprints/PDF/cicedoc2015.pdf.

  • Jiang, X., and J. Li, 2011: Influence of the annual cycle of sea surface temperature on the monsoon onset. J. Geophys. Res., 116, D10105, https://doi.org/10.1029/2010JD015236.

    • Search Google Scholar
    • Export Citation
  • Kitoh, A., 2017: The Asian monsoon and its future change in climate models: A review. J. Meteor. Soc. Japan, 95, 733, https://doi.org/10.2151/jmsj.2017-002.

    • Search Google Scholar
    • Export Citation
  • Krishnamurti, T. N., J. Molinari, and H. L. Pan, 1976: Numerical simulation of the Somali jet. J. Atmos. Sci., 33, 23502362, https://doi.org/10.1175/1520-0469(1976)033<2350:NSOTSJ>2.0.CO;2.

    • Search Google Scholar
    • Export Citation
  • Lau, K. M., and S. Yang, 1997: Climatology and interannual variability of the Southeast Asian summer monsoon. Adv. Atmos. Sci., 14, 141162, https://doi.org/10.1007/s00376-997-0016-y.

    • Search Google Scholar
    • Export Citation
  • Lau, K.-M., H.-T. Wu, and S. Yang, 1998: Hydrologic processes associated with the first transition of the Asian summer monsoon: A pilot satellite study. Bull. Amer. Meteor. Soc., 79, 18711882, https://doi.org/10.1175/1520-0477(1998)079<1871:HPAWTF>2.0.CO;2.

    • Search Google Scholar
    • Export Citation
  • Lawrence, D. M., and Coauthors, 2019: The Community Land Model version 5: Description of new features, benchmarking, and impact of forcing uncertainty. J. Adv. Model. Earth Syst., 11, 42454287, https://doi.org/10.1029/2018MS001583.

    • Search Google Scholar
    • Export Citation
  • Li, C.-Z., S. Yang, W. Mo, J. Zhang, and W. Wei, 2022: Seasonal predictions for May rainfall over southern China by the NCEP CFSv2. J. Trop. Meteor., 28, 2944, https://doi.org/10.46267/j.1006-8775.2022.003.

    • Search Google Scholar
    • Export Citation
  • Li, J., and L. Zhang, 2009: Wind onset and withdrawal of Asian summer monsoon and their simulated performance in AMIP models. Climate Dyn., 32, 935968, https://doi.org/10.1007/s00382-008-0465-8.

    • Search Google Scholar
    • Export Citation
  • Li, K., Y. Liu, Z. Li, Y. Yang, L. Feng, S. Khokiattiwong, W. Yu, and S. Liu, 2018: Impacts of ENSO on the Bay of Bengal summer monsoon onset via modulating the intra-seasonal oscillation. Geophys. Res. Lett., 45, 52205228, https://doi.org/10.1029/2018GL078109.

    • Search Google Scholar
    • Export Citation
  • Lipscomb, W. H., and Coauthors, 2019: Description and evaluation of the Community Ice Sheet Model (CISM) v. 2.1. Geosci. Model Dev., 12, 387424, https://doi.org/10.5194/gmd-12-387-2019.

    • Search Google Scholar
    • Export Citation
  • Liu, B., G. Wu, J. Mao, and J. He, 2013: Genesis of the South Asian high and its impact on the Asian summer monsoon onset. J. Climate, 26, 29762991, https://doi.org/10.1175/JCLI-D-12-00286.1.

    • Search Google Scholar
    • Export Citation
  • Liu, B., Y. Liu, G. Wu, J. Yan, J. He, and S. Ren, 2015: Asian summer monsoon onset barrier and its formation mechanism. Climate Dyn., 45, 711726, https://doi.org/10.1007/s00382-014-2296-0.

    • Search Google Scholar
    • Export Citation
  • Liu, Y., J. C. L. Chan, J. Mao, and G. Wu, 2002: The role of Bay of Bengal convection in the onset of the 1998 South China Sea summer monsoon. Mon. Wea. Rev., 130, 27312744, https://doi.org/10.1175/1520-0493(2002)130<2731:TROBOB>2.0.CO;2.

    • Search Google Scholar
    • Export Citation
  • Liu, Y., G. Wu, and R. Ren, 2004: Relationship between the subtropical anticyclone and diabatic heating. J. Climate, 17, 682698, https://doi.org/10.1175/1520-0442(2004)017<0682:RBTSAA>2.0.CO;2.

    • Search Google Scholar
    • Export Citation
  • Liu, Y. M., G. X. Wu, H. Liu, and P. Liu, 2001: Condensation heating of the Asian summer monsoon and the subtropical anticyclone in the Eastern Hemisphere. Climate Dyn., 17, 327338, https://doi.org/10.1007/s003820000117.

    • Search Google Scholar
    • Export Citation
  • Mao, J., and G. Wu, 2007: Interannual variability in the onset of the summer monsoon over the eastern Bay of Bengal. Theor. Appl. Climatol., 89, 155170, https://doi.org/10.1007/s00704-006-0265-1.

    • Search Google Scholar
    • Export Citation
  • Mao, J., J. C. L. Chan, and G. Wu, 2004: Relationship between the onset of the South China Sea summer monsoon and the structure of the Asian subtropical anticyclone. J. Meteor. Soc. Japan, 82, 845859, https://doi.org/10.2151/jmsj.2004.845.

    • Search Google Scholar
    • Export Citation
  • Plumb, R. A., and A. Y. Hou, 1992: The response of a zonally symmetric atmosphere to subtropical thermal forcing: Threshold behavior. J. Atmos. Sci., 49, 17901799, https://doi.org/10.1175/1520-0469(1992)049<1790:TROAZS>2.0.CO;2.

    • Search Google Scholar
    • Export Citation
  • Rodwell, M. J., and B. J. Hoskins, 1995: A model of the Asian summer monsoon. Part II: Cross-equatorial flow and PV behavior. J. Atmos. Sci., 52, 13411356, https://doi.org/10.1175/1520-0469(1995)052<1341:AMOTAS>2.0.CO;2.

    • Search Google Scholar
    • Export Citation
  • Wang, Z., and C.-P. Chang, 2012: A numerical study of the interaction between the large-scale monsoon circulation and orographic precipitation over South and Southeast Asia. J. Climate, 25, 24402455, https://doi.org/10.1175/JCLI-D-11-00136.1.

    • Search Google Scholar
    • Export Citation
  • Wu, C., S. Yang, A. Wang, and S. Fong, 2005: Effect of condensational heating over the Bay of Bengal on the onset of South China Sea monsoon in 1998. Meteor. Atmos. Phys., 90, 3747, https://doi.org/10.1007/s00703-005-0115-1.

    • Search Google Scholar
    • Export Citation
  • Wu, G., and Y. Zhang, 1998: Tibetan Plateau forcing and the timing of the monsoon onset over South Asia and the South China Sea. Mon. Wea. Rev., 126, 913927, https://doi.org/10.1175/1520-0493(1998)126<0913:TPFATT>2.0.CO;2.

    • Search Google Scholar
    • Export Citation
  • Wu, G., and Y. Liu, 2002: Thermal adaptation, overshooting, dispersion, and subtropical anticyclone Part I: Thermal adaptation and overshooting (in Chinese). Chin. J. Atmos. Sci., 24, 433446.

    • Search Google Scholar
    • Export Citation
  • Wu, G., Y. Guan, T. Wang, Y. Liu, J. Yan, and J. Mao, 2011: Vortex genesis over the Bay of Bengal in spring and its role in the onset of the Asian summer monsoon. Sci. China Earth Sci., 54 (1), 19, https://doi.org/10.1007/s11430-010-4125-6.

    • Search Google Scholar
    • Export Citation
  • Wu, G., Y. Guan, Y. Liu, J. Yan, and J. Mao, 2012: Air–sea interaction and formation of the Asian summer monsoon onset vortex over the Bay of Bengal. Climate Dyn., 38, 261279, https://doi.org/10.1007/s00382-010-0978-9.

    • Search Google Scholar
    • Export Citation
  • Wu, G., and Coauthors, 2015: Tibetan Plateau climate dynamics: Recent research progress and outlook. Natl. Sci. Rev., 2, 100116, https://doi.org/10.1093/nsr/nwu045.

    • Search Google Scholar
    • Export Citation
  • Wu, G., B. He, Y. Liu, Q. Bao, R. Ren, and B. Liu, 2016: Recent progresses on dynamics of the Tibetan Plateau and Asian summer monsoon (in Chinese). Chin. J. Atmos. Sci., 40, 2232, https://doi.org/10.3878/j.issn.1006-9895.1504.15129.

    • Search Google Scholar
    • Export Citation
  • Wu, X., and J. Mao, 2019: Decadal changes in interannual dependence of the Bay of Bengal summer monsoon onset on ENSO modulated by the Pacific decadal Oscillation. Adv. Atmos. Sci., 36, 14041416, https://doi.org/10.1007/s00376-019-9043-8.

    • Search Google Scholar
    • Export Citation
  • Xing, N., J. Li, X. Jiang, and L. Wang, 2016a: Local oceanic precursors for the summer monsoon onset over the Bay of Bengal and the underlying processes. J. Climate, 29, 84558470, https://doi.org/10.1175/JCLI-D-15-0825.1.

    • Search Google Scholar
    • Export Citation
  • Xing, N., J. Li, and L. Wang, 2016b: Effect of the early and late onset of summer monsoon over the Bay of Bengal on Asian precipitation in May. Climate Dyn., 47, 19611970, https://doi.org/10.1007/s00382-015-2944-z.

    • Search Google Scholar
    • Export Citation
  • Xu, Z., Y. Qian, and C. Fu, 2010: The role of land-sea distribution and orography in the Asian monsoon. Part II: Orography. Adv. Atmos. Sci., 27, 528542, https://doi.org/10.1007/s00376-009-9045-z.

    • Search Google Scholar
    • Export Citation
  • Yan, H., Z. Xiao, and L. Wang, 2003: Activities of Bay of Bengal monsoon and beginning date of rain season in Yunnan (in Chinese). Plateau Meteor., 22, 624630.

    • Search Google Scholar
    • Export Citation
  • Yanai, M., S. Esbensen, and J.-H. Chu, 1973: Determination of bulk properties of tropical cloud clusters from large-scale heat and moisture budgets. J. Atmos. Sci., 30, 611627, https://doi.org/10.1175/1520-0469(1973)030<0611:DOBPOT>2.0.CO;2.

    • Search Google Scholar
    • Export Citation
  • Yu, W.-D., J.-W. Shi, L. Liu, K.-P. Li, Y.-L. Liu, and H.-W. Wang, 2012: The onset of the monsoon over the Bay of Bengal: The observed common features for 2008–2011. Atmos. Ocean. Sci. Lett., 5, 314318, https://doi.org/10.1080/16742834.2012.11447009.

    • Search Google Scholar
    • Export Citation
  • Fig. 1.

    Onset dates of the BOBSM from 1980 to 2019. Dashed lines indicate one standard deviation from the average onset date; red (blue) means that the onset date is later (earlier) than one standard deviation.

  • Fig. 2.

    Climatology in April for (a) the pressure–latitude section of temperature (K) averaged between 80° and 100°E and (b) the 2-m surface wind (vectors; m s−1), precipitation (white contours; 5, 10, and 15 mm day−1) and the 2-m temperature (shading; °C). (c) Area-averaged vertical profile of diabatic heating (K day−1) over the “west” and “east” regions shown in (b). Red and blue denote sensible heating and latent heating, respectively, and the open triangles and closed circles denote the west and east regions, respectively. (d) Pressure–longitude cross section of the zonal circulation averaged over 10°–15°N [vectors; zonal wind in m s−1 and vertical motion (multiplied by a factor of −60) in Pa s−1], the horizontal divergence (shading; 10−6 s−1), and the zonal deviation of the equivalent potential temperature (contours; interval: 0.5 K; negative values shown as broken lines).

  • Fig. 3.

    Climate-mean pressure–latitude cross section along the eastern BOB (90°–100°E) of meridional circulation (streamlines), sensible heating (contours; 2, 4, 6, and 8 K day−1), and MTG (shading; 10−5 K m−1) in (a) January, (b) April, and (c) July, and (d) 6 days before BOBSM onset, (e) the onset day, and (f) 6 days after onset.

  • Fig. 4.

    Composites of daily mean SST (shading; °C), precipitation (blue contours; 6, 18, and 30 mm day−1) and 2-m surface wind (streamlines), before and after BOBSM onset: (a) D−12, (b) D−6, (c) D−2, (d) D0, (e) D0 (averaged from onset early years), and (f) D0 (averaged from onset late years).

  • Fig. 5.

    Composites of daily mean diabatic heating Q1 averaged between 850 and 500 hPa (shading; K day−1) and the 700-hPa circulation (streamlines) during the onset of the BOBSM on (a) D−6, (b) D−2, (c) D0, and (d) D+2.

  • Fig. 6.

    Composites of the 10°–15°N average pressure–longitude cross section of the zonal circulation [vectors; zonal wind in m s−1 and vertical motion (multiplied by a factor of −60) in Pa s−1] and meridional wind (shading; m s−1) for (a) D−6, (b) D−4, (c) D−2, (d) D0, (e) D0 (averaged from onset early years), and (f) D0 (averaged from late onset late years). Black shading denotes the topography.

  • Fig. 7.

    Daily evolution of the 2-m surface wind (vectors; m s−1), the sea surface sensible heating flux (shading; W m−2), and the relative vorticity (contours; interval: 4 × 10−6 s−1): (a) D−6, (b) D−4, (c) D−2, and (d) D0. Heavy contours denote zero relative vorticity.

  • Fig. 8.

    Distribution of the terms in the vorticity budget (10−5 s−1), Eq. (7), at the 2-m surface on (left) D−2 and (right) D0: (a),(b) relative vorticity tendency ∂ζ/∂t, (c),(d) −f∇ ⋅ Vx; (e),(f) −Vψ ⋅ ∇ζβvψ; (g),(h) −Vx ⋅ ∇ζβvx.

  • Fig. 9.

    Annual cycle of the CEF intensity (m s−1) for (a) the climatology, (b) the early years, and (c) the late years. Red and purple curves denote the area-average over (5°S5°N, 50°60°E) and (5°S5°N, 80°90°E), respectively. The green curve denotes the area-mean precipitation (mm day−1) averaged over (10°20°N, 90°100°E). The black dashed vertical line indicates the monsoon onset pentad.

  • Fig. 10.

    Hovmöller diagrams of the daily mean precipitation averaged over 10°–20°N (shading; mm day−1) in (a) TRMM, and in the numerical experiments of (b) the CTRL run, (c) the No_Indian run, (d) the No_Inch run, and (e) the No_Inch_Indian run.

  • Fig. 11.

    (a) Difference in the April-mean meridional geopotential gradient at 925 hPa (shading; 10−6 m s−2) between the CTRL and No_Indian experiments. (b) Annual cycle of the BOB CEF (m s−1) at 925 hPa (red curve) and the area-averaged precipitation (mm day−1) in the eastern BOB (10°–20°N, 90°–100°E; green curve) as shown in panel (a); the solid and dashed curves indicate the CTRL and No_Indian experiments, respectively. (c)–(f) Pentad-averaged differences in the wind at 925 hPa (vectors; m s−1) between the CTRL and No_Indian experiments. (g),(h) Pressure–latitude sections of the meridional circulation (streamlines) and MTG (shading; 10−5 K m−1) over the eastern BOB (90°–100°E) at the times of onset in the CTRL and No_Indian experiments, respectively.

  • Fig. 12.

    Schematic diagram showing the roles of the land–sea thermal contrast in the meridional direction at the equator and in the zonal direction across the tropics in the onset of the BOBSM. (a) In early April, CEF first occurs over the central Indian Ocean (BOB CEF), which generates a band of westerly flow and a surface sensible heat flux in the southern BOB with cyclonic vorticity to the north. The zonal contrast generates a warm SST pool in the central BOB, which weakens the anticyclonic circulation and easterly flow to the north in the lower troposphere. (b) Before the onset of the BOBSM, the development of the Somali CEF enhances the impacts of the land–sea thermal contrast in both the meridional and zonal directions, leading to the onset of the BOBSM; see text for details.