1. Introduction
The tropical Indian Ocean forms the major part of the largest warm pool on Earth, and it differs considerably from other oceans. The Indian Ocean is closed to the north by the Asian continent, preventing large meridional heat transfer to higher latitudes in the Northern Hemisphere, and the relatively stable anti–Walker circulation means that the Indian Ocean lacks steady equatorial easterlies (Schott and McCreary 2001). Moreover, the tropical Indian Ocean is a nonnegligible member of the Asian summer monsoon (ASM) system, which is the strongest monsoon system on Earth. On the one hand, the tropical Indian Ocean plays a key role in modulating the South Asian summer monsoon (Wang 2006; Yang et al. 2007; Lutsko et al. 2019) and East Asian summer monsoon (Li et al. 2008; Xie et al. 2009; Liu and Duan 2017), the two subsystems of the ASM. On the other hand, the dominant currents of the Indian Ocean—for example, the Somali Current and the Southwest/Northeast Monsoon Current—are largely controlled by the monsoon system and exhibit annual reversals (Schott et al. 2009).
The relationship of the Indian Ocean’s interannual variability with the El Niño–Southern Oscillation (ENSO) cycle and with the ASM has received much attention. For example, the Indian Ocean is known to prolong the influence of El Niño into boreal spring and summer via the capacitor effect and is an important pathway for ENSO to influence the ASM (Klein et al. 1999; Yang et al. 2007; Li et al. 2008; Du et al. 2009). Tropical Indian Ocean warming acts as a capacitor, anchoring atmospheric anomalies over the Indo–western Pacific Ocean, and its related adiabatic adjustment emanates Kelvin waves into the Pacific Ocean. In addition, northern Indian Ocean warming is crucial for maintaining the Kelvin waves and anomalous northwest Pacific anticyclone in post–El Niño summers (Xie et al. 2009, 2016). Furthermore, since the Indian Ocean dipole phenomenon was proposed (Saji et al. 1999; Webster et al. 1999; Chang and Li 2000), the role of Indian Ocean warming is no longer considered to be a passive response to El Niño, but instead an important self-sustaining mechanism that involves ocean processes (Li et al. 2003). Although existing theories cannot explain all historic Indian Ocean dipole events (Gualdi et al. 2003; Yamagata et al. 2004), most tend to consider the strength of the monsoon circulation and the ocean processes as being important factors (Saji et al. 1999; Xie et al. 2002; Li et al. 2003). From this point of view, the interaction between ocean processes in the Indian Ocean and the leading modes of variability of the atmosphere over the Indian Ocean basin deserves further study.
The Indian Ocean is closely related to the ASM (Li et al. 2008); however, as the most complex monsoon system in the world, the ASM system is not yet sufficiently understood (Wang 2006; IPCC 2014). Many existing studies on the details of the mechanisms involved are based on linear statistical analysis and process descriptions. The implicated internal processes are diverse and need to be explored. The lack of knowledge about the ASM makes it hard to diagnose and tune numerical models to obtain better simulation performance for ASM regions. In fact, some state-of-the-art GCMs show a large bias for circulation and precipitation over ASM areas (Sperber et al. 2013). Furthermore, the interannual predictability of the ASM is still low (Wang et al. 2009). A better understanding of the subprocesses within the Indian Ocean, such as air–sea interactions, is needed to guide the improvement of simulation results (Lau and Nath 1996; Alexander et al. 2002; Duan et al. 2008; He et al. 2019).
In addition to air–sea interactions, another factor in the ASM system that needs to be considered is the thermal forcing of the Tibetan Plateau (TP). Although many previous studies have focused on the mechanical effects of the TP (Bolin 1950; Hoskins and Karoly 1981; Wu 1984; Rodwell and Hoskins 2001; Boos and Kuang 2010, 2013), Wu et al. (Wu et al. 2007, 2015; G. Wu et al. 2012) suggest that the lifting heating effect of the TP is also an important factor affecting the Asian climate, especially when it acts as a “sensible heat pump” during spring and summer (Wu et al. 1997). Here, we focus on thermal effects not only because of their greater role in altering the TP’s surrounding atmospheric circulations, but also for their showing significant impact on the ASM with considerable interannual variability (Ye and Gao 1979; Yanai and Li 1994; Zhao and Chen 2001; Duan and Wu 2005; Wang et al. 2016; Wu et al. 2016), possibly even from the perspective of modulating the effect of ENSO on the ASM via teleconnection (Z. Wu et al. 2012; Jin et al. 2018).
The early use of general circulation models to study the impact of Asian topography can be traced back to Hahn and Manabe (1975). By eliminating all land elevations in an atmospheric general circulation model (AGCM), they found that the TP is associated with middle- and upper-atmosphere heating, and that both the South Asian monsoon and Hadley cell are influenced significantly in summer. Yasunari et al. (2006), Tang et al. (2013), and Chen and Bordoni (2014) also explored the role the TP plays in the ASM, by removing related topography in the AGCM. Performing global mountain uplifting experiments with coupled general circulation models (CGCMs), several studies (Abe et al. 2003, 2005; Kitoh 2004, 2007; Okajima and Xie 2007; Kitoh et al. 2010) have shown that global topography is important in forming the mean climate state of the ASM and SST over the Indian Ocean and Pacific Ocean. Focusing more on the impacts of the TP, Abe et al. (2013) found that the surrounding air–sea interactions and SST in adjacent oceans can be influenced by removing the TP in a CGCM. Fallah et al. (2016) also showed that removing the TP from another CGCM resulted in changes to the Atlantic meridional overturning circulation and North Atlantic SST when they are in a balanced state. Recently, Baldwin et al. (2019) found that SSTs in the western North Pacific and upwelling in the Arabian Sea (AS) are influenced by removing Asian topography. These results indicate that, to a certain degree, the TP can alter the state of oceans through air–sea interactions, as most ocean currents are so-called wind-driven circulations. Ye et al. (1979) pointed out that the TP heat source is closely related to the local Hadley circulation over the Indian Ocean, and even remote teleconnections in the Southern Hemisphere. However, the underlying mechanism is still not clear. Although, commonly, ocean signals influence the atmosphere in the tropics rather than vice versa (Cayan 1992), the theory that ENSO signals are stored in the Indian Ocean basin mode through an atmospheric bridge (Yang et al. 2007) in spring implies the possibility of the Indian Ocean being affected by the atmosphere. Hu and Duan (2015) also pointed out that Indian Ocean sea surface temperature (SST) and TP thermal heating are closely related to each other in summer and are both important to the interannual variability of the East ASM. In general, both the Indian Ocean and TP play important roles in shaping the climate on both regional and global scales, and the Indian Ocean can be viewed as key for understanding the ASM during spring and summer. However, only a few studies have attempted to explore the ocean’s response to the TP thermal effect. Wang et al. (2019) performed several experiments with a regional model and found that TP heating significantly enhances the southwesterly monsoon circulation over the northern Indian Ocean, and further cools the sea surface in summer. A clear comparison of these related studies is provided in Table 1.
Related studies about large-scale topography and the Indian Ocean. (CCSRUT is Center for Climate System Research at the University of Tokyo; OMLM is ocean mixed layer model; expansions of other acronyms are available online at https://www.ametsoc.org/PubsDataPolicy.)
Our recent study (Zhao et al. 2018) indicated that the TP thermal forcing and tropical Indian Ocean SST anomalies start to act as correlated processes in May. Here, we attempt to explore the short-term response (within one month) of the Indian Ocean in terms of mixed layer and upper-level currents to the TP thermal forcing. However, the ASM is regarded as a complex system that includes many nonlinear processes, so linear statistical models (such as correlation analysis, linear regression, and the Granger causality paradigm) can be misleading. Only in a pure linear system or a “synchrony” phenomenon can linear statistical diagnosis be helpful (Sugihara et al. 2012). Therefore, we opted to use numerical experiments to support our viewpoint, providing statistical results only for reference.
Compared to previous work (listed in Table 1), this study improves existing insights in four aspects. First, we shed more light on the changes in the upper ocean and the related air–sea interactions, rather than SST. This can be conveniently assessed by performing a series of numerical experiments with a full land–air–sea coupled model. Second, instead of reshaping the topography in the model when exploring the related issues mentioned above, we choose to modify the surface heat source anomaly over the TP only, as it is the dominant component of the total atmospheric heat source in spring (Ye and Gao 1979; Yanai et al. 1992). Although modifying the topography in a model can help us understand the underlying mechanisms of large-scale topography, in practical climate predictions the TP thermal forcing has considerable interannual variability (Zhao and Chen 2001; Zhao et al. 2018), while the change in TP elevation is regarded as fixed. Third, most previous results have attempted to understand the ocean response in a balanced state; in other words, their conclusions are valid only as a mixture of many nonlinear processes with different time scales from a climatological perspective. Here, we observe the short-term ocean response to TP thermal forcing using a branch-run technique in a CGCM, so that the changes in two subsystems (the TP and Indian Ocean) in this complex system are relatively independently separated. Fourth, we choose to discuss the situation in May, a very special month in the ASM. On the one hand, it is part of the traditional boreal spring season, during which surface sensible heat is considered to dominate the atmospheric heat source over the TP. On the other hand, distinct differences exist in the background circulation between March/April and May. Furthermore, the TP surface heat source and the Indian Ocean are closely related in May (Zhao et al. 2018).
The rest of the paper is organized as follows. In section 2, the data and numerical models used in this study are introduced. The model validation and experimental design are also described in this section. In section 3, we analyze the response of the Indian Ocean and possible mechanisms according to the model results. Finally, a summary is presented, and the potential limitations of the present work and avenues for future research are discussed in section 4.
2. Methods
a. Data
The data used in this study are as follows.
Six-hourly meteorological observations at 73 stations across the TP collected by the China Meteorological Administration. Variables include surface air temperature, ground surface temperature, and wind speed at 10 m above the surface. Quality control procedures have been applied to eliminate erroneous data and ensure homogeneity (Duan et al. 2011).
NCEP Global Ocean Data Assimilation System (GODAS) data. These data provide a detailed pentad proxy of sea currents and temperature states (Behringer and Xue 2004). To diagnose the mixed layer, potential temperature and current speed are used with a resolution of 0.333° × 1.0°. More details can be found at https://www.esrl.noaa.gov/psd/data/gridded/data.godas.html.
Monthly SSTs from the Hadley Center Global Sea Ice and Sea Surface Temperature dataset, version 1.1 (Rayner et al. 2003), with a resolution of 1.0° × 1.0°. More details can be found at https://climatedataguide.ucar.edu/climate-data/sst-data-hadisst-v11.
Wind speeds and radiation flux from the Japanese 55-yr Reanalysis (JRA-55) project conducted by the Japan Meteorological Agency (Ebita et al. 2011; Kobayashi et al. 2015), with a resolution of 1.25° × 1.25°. These data are employed to support the analysis of air–sea interactions and surface heat flux exchange in the heat budget equation.
NCEP–DOE Reanalysis-2 data, as Zhong et al. (2013) showed that the vertical distributions of heat sources in the NCEP reanalysis datasets are reasonably similar to observations. The vertical diffusion heating rate, with a T62 Gaussian grid resolution and 28 sigma levels, is used to evaluate the vertical diffusion profile. More details can be found at https://www.cpc.ncep.noaa.gov/products/wesley/reanalysis2/index.html.
The period for all the above datasets used in this study is 1980–2014, to cover the observation period and available GODAS datasets.
b. Model
The CGCM employed in this study is the Community Earth System Model (CESM), version 1.2.0. This model is widely used in climate variability studies and has made significant contributions to regional and global assessments of climate changes (see http://www.cesm.ucar.edu/models/cesm1.2/ for more details). With the help of a finite-volume dynamical core, the capability of simulating the climate mean in the Community Atmosphere Model, version 4.0 (CAM4), is greatly improved, especially when it is used in a CGCM (Gent et al. 2011; Neale et al. 2013; Kay et al. 2015). The ocean component is based on the Parallel Ocean Program, version 2.1 (POP2), of the Los Alamos National Laboratory, and is commonly used in CGCMs. In addition, the Community Land Model, version 4.0, is the land model for the CESM.
c. Model evaluations
As the performance of CAM4 has been well documented by Neale et al. (2013), we evaluate only the air–sea interface and the ocean performance in our experiment. The mean climate state in May during 1980–2014 for SSTs provided by the Hadley Center and the mean climate state of the surface wind from JRA-55 are adopted as observations in our evaluation. We find that surface winds are well simulated both in direction and magnitude in our control run. The southeasterly wind over the southern Indian Ocean and the southwesterly wind over the AS and the Bay of Bengal (BOB) are all well simulated (Fig. 1b). The model produces a pattern of warm pools similar to that in the Hadley SST dataset (Fig. 1a).
(a) Climatological mean of Hadley SST (shaded; °C) and surface wind (vectors; m s−1) in JRA-55 during 1980–2013 in May. (b) As in (a), but produced by the CESM control run. (c) The 0–45-m averaged potential temperature (shaded; °C) and sea currents (vectors; cm s−1) from GODAS. (d) As in (c), but produced by the CESM control run. (e) Vertical diffusion profile in the NCEP–DOE Reanaysis-2 dataset, averaged over the TP areas with elevations greater than 1500 m. Black represents the climate mean; red and blue denote its range from the strongest and weakest years, respectively. (f) Vertical diffusion profile from CESM averaged over the TP areas with elevations greater than 1500 m. Black represents the control run; red, the TP_S ensembles; and blue, the TP_W ensembles.
Citation: Journal of Climate 32, 20; 10.1175/JCLI-D-18-0880.1
The output of POP2 is also compared with its counterpart in GODAS. As we focus on the simultaneous response within one month in this study, the deep levels in the ocean do not have enough time to respond significantly, so we evaluate only the upper level. The potential temperature and sea current averaged for the top 45 m are shown in Figs. 1c and 1d. It can be seen that the model results capture the southward current in the subsurface layer in both the AS and BOB; the equatorial eastward current appears broader in the simulation, and Wyrtki jets (Wyrtki 1973) are also well simulated. Hence, POP2 can reasonably reproduce the sea currents and temperature patterns in the upper Indian Ocean.
d. Experimental design
As we focus on the short-term response of the ocean to the TP thermal forcing in May, we perform an ensemble of short-term integrations of the CGCM, rather than running long simulations and comparing the equilibrated states. Owing to the short memory of the atmosphere, the experiment’s atmospheric signal dampens within several months and does not last through a year. This means that if we set a forcing in spring for each year and integrate for 15 years, the atmospheric state will be random every winter, which means that the continuous integration of the AGCM for the remaining years can be considered as multiple repeated sampling, because the SST boundary conditions remain climatological for each year and only a random beginning condition is set by random signals from the last year. Therefore, 15 years of integration can be regarded as 15 independently identical instances. However, for a CGCM, if we set an atmospheric forcing in spring for a certain year, the response observed in the following years or even decades is the direct response to not only the original atmospheric forcing, but also the sea temperature and ocean process anomalies caused by the atmospheric forcing. The corresponding ocean response can last until the second year, and thus can even change the circulation for the following several years or decades. This kind of nonlinear response accumulates for years, which makes the following years not independently identical. So, the average of the 15 years can only be viewed as one instance of the climate state. In other words, this mean climate state includes many nonlinear processes and some time-delayed effects from the past several years, which prevents us from investigating a more separable and independent mechanism in a complex system. Therefore, in this study, we choose to use the branch-run technique to avoid this situation, by starting from each 1 May in the 15 years after the CGCM spinup and integrating for only 30 days, with each year as an ensemble member, to observe the short-term response in each year; each set of individual sensitivity experiments includes 15 ensemble members with different initial states as a random sampling process.
To maintain the surface energy balance process, the experiments in this study did not modify the surface temperature or the sensible heat directly. Instead, the vertical diffusion was changed to indirectly control the surface heating over the TP, following Wu et al. (2007). In other words, we modified the process of land heating the atmosphere by changing the parameter k in the k∂θ/∂z term, leaving ∂θ/∂z to vary freely. This is basically reasonable because the sensible heat in this model is obtained by integrating the vertical diffusion in the atmospheric column. This method is the same as that used in G. Wu et al. (2012) and He et al. (2015). The domain-averaged observed vertical diffusion profiles over the TP range approximately from 0.5 to 1.5 times the mean climate state (Fig. 1e). The interannual variability estimated by standard deviation reaches 23.9% of the mean state. These results provide a reference when we use numerical experiments for further investigations. The details of the experimental design are as follows. Using CAM4 coupled with POP2 and CLM to integrate for 115 years, we take the first 100 years as the spinup and the last 15 years as the control run and start the branch runs from 1 May in these 15 years as 15 ensemble members for each sensitivity experiment. Allowing free feedback in air–land interactions, the vertical diffusion is amplified to 1.5 times its original value during each 30-day branch run in the strong TP heating experiment (TP_S cases; vertical diffusion is shown as the red curve in Fig. 1f). Correspondingly, the vertical diffusion is reduced to 0.5 times its original values in the weak TP heating experiment (TP_W cases; vertical diffusion is shown as the blue curve in Fig. 1f). As the surface thermal forcing is not forced to be 1.5 or 0.5 times the fixed state, the related processes can change accordingly with the corresponding feedback boundary layer processes.
In summary, the model was run 15 times from 15 years, with each ensemble member differing in its initial conditions in an experiment. Each year, the simulation in different experiments was branched in May with the same initial conditions, and month-long simulations were run with strong and weak thermal diffusion. Ultimately, there were three sets of ensembles, with normal, strong, and weak TP thermal conditions, each with 15 months of May. We compare the ocean’s change in the strong and weak cases in the following sections.
3. Results
a. Response of SST and related atmospheric circulations
To investigate the ocean response to the TP thermal effect, we need to first observe the changes at the air–sea interface. SST is considered an important indicator of thermal forcing for many weather and climate systems (Cayan 1992; Hastenrath 2012). Considering that ocean processes are much slower than atmospheric processes, the upper layers in the northern Indian Ocean may take approximately 15 days to respond (see section 3c for details); therefore, we observe the SST difference during the second half of May between TP_S and TP_W (Fig. 2). In the northern Indian Ocean, SST is significantly colder in the TP_S cases compared with in the TP_W cases. This SST response is similar to the result from a regional model in summer (Wang et al. 2019).
Difference fields of SST between the TP_S and TP_W experiments. Dotted areas indicate statistical significance at the 90% confidence level.
Citation: Journal of Climate 32, 20; 10.1175/JCLI-D-18-0880.1
Observing the changes in the atmosphere will improve our understanding of the cooling response in the northern Indian Ocean. Difference fields of geopotential height and circulation show that a stronger TP surface heat source enhances the anticyclonic circulation or so-called South Asian high at 200 hPa (Fig. 3a). At 850 hPa, however, it is present as a cyclonic circulation anomaly around the TP, with northerly wind to its west and westerly wind to its south over the Indian Ocean (Fig. 3b). In the transition from 200 to 850 hPa, the circulation anomaly at 500 hPa behaves more like the one at 200 hPa (figure not shown here). All these responses are basically consistent with the findings from previous work (Wu et al. 2007; G. Wu et al. 2012) and can be well explained by potential vorticity theory and the atmospheric thermal adaptation paradigm (Hoskins et al. 1985; Liu and Wu 2000; Liu et al. 2017).
Difference fields of circulation (vectors; m s−1) and geopotential height (shaded; gpm) between the TP_S and TP_W experiments at (a) 200 and (b) 850 hPa. Dotted areas and black vectors indicate statistical significance at the 90% confidence level according to the t test. Red dashed lines denote the TP’s topography. (c) Difference of zonally averaged (75°–110°E) meridional circulation (vectors; υ in m s−1, −ω in 10−2 Pa s−1) between the TP_S and TP_W experiments. (d) Difference fields of meridionally averaged (10°S–10°N) zonal circulation between the TP_S and TP_W experiments. Red vectors denote the difference passed the 90% significance level in a two-sided t test. Difference fields of (e) precipitation (mm s−1) and (f) surface temperature (°C) between the TP_S and TP_W experiments.
Citation: Journal of Climate 32, 20; 10.1175/JCLI-D-18-0880.1
The overall meridional circulation along the TP also shows a monsoonal-type circulation response, with an ascending branch located over the top of the TP and a descending one over the tropical Indian Ocean (Fig. 3c). This is easily explained, as the air mass continuity equation shows that where there is an ascending branch there must be a descending branch. In addition, the so-called anti–Walker circulation along the equatorial Indian Ocean is impacted. The anti–Walker circulation usually manifests as an ascending branch over the Maritime Continent, a descending branch on the east coast of Africa, east winds at high levels, and westerly winds at low levels. Such a vertical circulation is closely coupled with the Walker circulation over the Pacific Ocean (Bjerknes 1969). The anti–Walker circulation is located to the south of the stationary wave excited by the TP heat source, so it is accelerated by the high-level easterlies and low-level westerlies (Fig. 3d). However, the equatorial westerly response below 700 hPa is not significant, so it does not cause significant changes directly in the equatorial thermocline (figure not shown).
The eastern part of the TP shows an increase in precipitation, together with significant precipitation reductions in the western TP, the Indian peninsula, the southern Arabian Peninsula, and Somalia (Fig. 3e). Surface temperature shows a positive anomaly centered in the western part of the TP and the Indian peninsula, and the eastern part of the Mediterranean region shows a significant increase. Besides, this result is consistent with Rodwell and Hoskins (1996), in that the Mediterranean and the Sahara area are affected by thermal forcing from Asia. The change in land surface temperature (Fig. 3f) is also closely related with precipitation, in that where precipitation is less, the land surface will be drier and the temperature higher. The response of the land and ocean surface temperature to the south of the TP is completely opposite. Enhanced evaporation cools the SST, while drier land becomes warmer, intensifying the land–sea thermal contrast between the eastern Indian peninsula and the BOB. However, due to the enhanced westerly, the cold advection from the AS cools the western Indian peninsula surface.
b. Mixed layer sea temperature diagnosis
Although it is impossible for all levels in the ocean to respond to the atmospheric changes in less than one month, the upper layer that interacts with the atmosphere in a more direct way might show significant changes. In this section, we investigate the corresponding adjustments in the mixed layer.
Obviously, the difference in the temperature tendency of the mixed layer is very close to that of SST. Because of the signal change coming from the atmosphere, the mixed layer of the ocean is influenced from the top down, starting from the surface, and the relatively uniform overall change of the mixed layer temperature is consistent with that at the surface. As the starting points in each pair of experiments in TP_W and TP_S are the same, the difference in the temperature tendency can also reflect the final temperature changes. The left-hand side of Eq. (2) denotes the rate of local temperature change. The difference in this term between TP_S and TP_W is shown in Fig. 4a. This difference is consistent with the SST difference in Fig. 2, in which the northern Indian Ocean shows a significant negative SST difference and the southern Indian Ocean presents a positive SST anomaly. To further understand the reason for this change, we calculate the difference fields of the five terms on the right-hand side of Eq. (2) successively (Figs. 4b–f). The results show that the heat flux term (Fig. 4e) dominates this kind of temperature difference. In addition, the vertical mixing effect also contributes to temperature changes in the BOB, with cooling effects over the northwestern coast and warming effects over the northeastern coast (Fig. 4d). This phenomenon can be explained by the surface wind–induced vertical motion anomaly discussed in section 3c. Note that the residual term is small enough to be neglected, which justifies our diagnosis.
Difference fields of the (a) time tendency term of the mixed layer potential temperature, (b) zonal temperature advection, (c) meridional temperature advection, (d) vertical entrainment term, (e) heat flux exchange, and (f) residual (K s−1) between TP_S and TP_W. Dotted areas indicate the difference passed the 90% significance level in a two-sided t test.
Citation: Journal of Climate 32, 20; 10.1175/JCLI-D-18-0880.1
To further understand the underlying reason for the heat flux changes in Fig. 4e, the four components of the sea surface heat flux are diagnosed (Fig. 5). It is clear that the latent heat flux (Fig. 5a) and solar radiation flux (Fig. 5d) are responsible for the mixed layer temperature change. However, the difference field of latent heat flux is more important as it imitates the pattern and center values of the total heat flux differences (Fig. 4e), which exhibit negative values in the northern Indian Ocean. Although not sufficiently significant compared with the latent heat flux changes, the solar radiation flux (Fig. 5d) is more important than the sensible heat flux (Fig. 5b) and longwave radiation flux (Fig. 5c), especially in the northern AS. According to the discussion about wind stress in section 3d, the surface wind response are significant and consistent with the mean climate state in the northern Indian Ocean (Figs. 1a,b). Therefore, accelerated surface wind leads to more evaporation, which enables the ocean to release more energy to the atmosphere, cooling the water. A greater surface wind speed also enhances the mixed layer entrainment and makes the mixed layer deeper (as shown in the following section).
Difference fields of the (a) surface latent heat flux, (b) surface sensible heat flux, (c) surface longwave radiation flux, and (d) surface shortwave radiation flux (K s−1) between TP_S and TP_W. The downward direction is chosen as the positive direction. Dotted areas indicate the difference passed the 90% significance level in a two-sided t test.
Citation: Journal of Climate 32, 20; 10.1175/JCLI-D-18-0880.1
After the BOB summer monsoon onset, the surface wind is as being southwesterly during May (Mao and Wu 2007). Meanwhile, a stronger TP surface heat source tends to enhance the cyclonic circulation in the lower level of the atmosphere (Fig. 3b), which results in the westerly wind anomaly to the south of the TP and accelerates the southwesterly wind on the sea surface in the northern Indian Ocean, consistent with the result from Wu et al. (2007).
In the observational data, we also find some evidence that supports the above findings. First, a TP surface heat source index is defined as the average for 73 stations across the TP (Duan et al. 2008). A composite analysis of each term in the heat equation is performed according to the TP surface heat source index in May. The strong (weak) years are defined as years in which the TP surface heat source index is greater (less) than 0.7 times its standard deviation. To exclude the interference of ENSO, which has strong interannual variability, the ENSO years are omitted, as they can greatly alter the SST in the Indian Ocean and interfere with the effect of the TP surface heat source. Strong TP heating years are 1986, 2004, 2012, and 2014; weak years are 1981, 1984, 1990, and 2013. The composite analysis result shows that the northern AS and BOB tend to be colder in the strong TP surface heat source years (Fig. 6a), which is basically similar to our model result (Fig. 4a). We also notice that the warm anomaly area in the AS from the composite (Fig. 6a) is not consistent with that in Fig. 4a. The residual term provides some clues to this phenomenon (Figs. 4f and 6f). It may be that some other signals as well as TP thermal forcing are captured, or the pentad GODAS dataset is unable to rule out the errors as we did with a daily output from the model. However, we think that overall the observed results support the model results, not only because of the small residual term, which comes from the daily output, but also because of the reasonable radiation flux term, which shows a similar result both in the composite analysis (Fig. 6e) and model difference field (Fig. 4e)—negative patterns in the eastern side of the AS and BOB. For horizontal advection and vertical entrainment, the differences between strong years and weak years seem to be scattered and trivial near the equator (Figs. 6b–d). Given the large residual in the observations, perhaps more data samples and higher data resolutions are needed.
As in Fig. 4, but for the composite differences between the TP sensible heat strong years and weak years calculated using GODAS and JRA-55. Dotted areas indicate the difference passed the 90% significance level in a two-sided t test.
Citation: Journal of Climate 32, 20; 10.1175/JCLI-D-18-0880.1
c. Vertical structure of changes in the northern Indian Ocean
In the above sections, we discuss the horizontal patterns of the changes in the mixed layer and notice that the northern Indian Ocean is significantly colder in the strong TP surface heat source years. Since TP thermal heating can affect the ASM, we further consider the response within the Indian Ocean. Next, we investigate the vertical structure of temperature and sea current changes.
Considering that the AS and BOB are at nearly identical latitudes, we take the average over latitudes 10°–20°N and find that the longitude–depth profile provides more information about the vertical structure of the ocean response. As discussed above, the SST tends to be colder in the northern Indian Ocean when the TP surface heat source is stronger than normal. In fact, the cold response can penetrate 30 m in the upper ocean (Fig. 7a). In addition, the cooling zone is located in the eastern part of the AS, while in the BOB it is limited to the western part of the basin. At depths of 40–50 m, a mass of warm water is squeezed by the upper cold water. This depth is exactly the depth of the mixed layer bottom where the colder subsurface water is separated from the uniformly mixed layer. When sea surface wind speed increases, the oceanic mixed layer becomes deeper as a result of wind-induced mixing; thus, the original subsurface cold water can now mix with the upper warm water and exhibit warm anomalies beneath the cooling SST. In addition, it is unsurprising that this part of the warm anomaly fails to pass the 90% confidence level, because the variance of the potential temperature at the bottom of the mixed layer is very large as it changes with the mixed layer depth and varies between two different layers.
Difference fields of the (a) potential temperature (K), (b) vertical motion (cm s−1), (c) zonal ocean current velocity (cm s−1), and (d) meridional ocean current velocity (cm s−1) averaged across 10°–20°N between the TP_S and TP_W experiments. Dotted areas indicate the difference passed the 90% significance level in a two-sided t test.
Citation: Journal of Climate 32, 20; 10.1175/JCLI-D-18-0880.1
Regarding the vertical motion, the response in the AS is similar to that in the BOB, in that the western part is an upwelling anomaly and the eastern part is a downwelling anomaly. However, the magnitude in the AS is not as large as that in the BOB. This difference can be explained by the strength and direction of Ekman transport. First, the surface wind anomaly in the BOB is much larger than that in the AS (Figs. 8a,b), which results in a larger Ekman transport in the BOB (Figs. 7c,d). Second, although the corresponding surface wind anomaly is southerly, the direction is slightly different between the AS and the BOB. Typically, the surface wind anomaly over the BOB is southwesterly, but the southerly component is more dominant than that in the AS (Figs. 8a,b). This situation also results in water mass transport from northwest to southeast, with an eastward component more significant in the upper layer of the BOB (Fig. 7c) and a relatively significant southward transport in the eastern AS (Fig. 7d).
Surface wind speed (red vector; m s−1) and ocean current (black vector; cm s−1) differences between TP_S and TP_W averaged (a) in the AS (averaged over 10°–20°N, 55°–75°E; land excluded) and (b) in the BOB (averaged over 10°–20°N, 80°–100°E; land excluded). (c) The 5-m temperature evolution (°C) in the AS (red) and BOB (blue) in May, where dashed lines represent TP_W and solid lines TP_S. Shaded periods passed the two-sided t test at the 90% confidence level.
Citation: Journal of Climate 32, 20; 10.1175/JCLI-D-18-0880.1
The meridional current changes over the AS and BOB are further compared in Fig. 9. In general, both basins exhibit similar meridional current anomalies: there is upwelling in the north and southward drifting in the upper levels, and cold water anomalies appear at depths of 0–30 m to the north of 5°N. However, the penetration of the cold water anomaly is slightly deeper in the AS than in the BOB. This contrast is caused by a deeper climatological mixed layer over the AS. In the AS, upwelling is limited to north of 20°N, and southward water transport is most significant at approximately 25 m. The upper southward water transport anomaly mainly results from Ekman transport, as the surface wind anomaly over the AS tends to be westerly. In the BOB, however, the upwelling anomaly is observed from 5° to 20°N. Considering that the surface wind stress anomaly is much larger in the BOB and that the southeastern BOB is closed by the Malay Peninsula and Sumatra, it is unsurprising that the shape of the coast leads the water in the BOB to the south and even to merge with a southeastward sea current to cross the equator, as we describe in section 3d. In both the AS and BOB, the downwelling branch of the meridional anomaly circulation near the equator is not significant as it is just a compensating motion subjected to the mass continuity equation. We also observe that at depths of 60–110 m over the eastern equator, a maximum center of potential temperature is accompanied by downwelling motions (Fig. 9b), but in the western equatorial Indian Ocean this temperature increase is very limited (Fig. 9a). This phenomenon is mainly related to the thermocline depth change across the equator. In fact, the thermocline tends to be much deeper in the eastern tropical Indian Ocean than in the western part (figure not shown).
Difference fields of the potential temperature (shaded; K) and meridional circulation (vectors; cm s−1, with the vertical velocity multiplied by 500 for convenient visualization) between the TP_S and TP_W experiments in (a) the AS and (b) the BOB. Dotted areas and black vectors indicate the difference passed the 90% significance level in a two-sided t test.
Citation: Journal of Climate 32, 20; 10.1175/JCLI-D-18-0880.1
We attribute these differences between two basins in large part to their different positions relative to the TP. Located south of the TP, the BOB has a northward component of the surface wind that is larger than the eastward component. However, for the AS, which is located southwest of the TP, the surface wind has a larger eastward component. Being closer to the center of the TP heating-related atmospheric circulation response, the surface wind anomaly in the BOB is stronger than that in the AS (red vectors in Figs. 8a and 8b). Their different directions also contribute to the difference in the direction of water mass transport. Both the areal averages of the AS and BOB ocean current anomalies exhibit classic Ekman spirals (black arrows in Figs. 8a and 8b). In the AS, the surface wind anomaly is almost westerly; thus, the meridional component of Ekman transport is more significant because the mixed layer is a little deeper so the spiral can penetrate to 35 m. Meanwhile, in the BOB, the surface wind anomaly is more southerly, so the zonal Ekman transport is more significant. As the mixed layer is a little shallower, the Ekman spiral is limited to 0–25 m. We also note that the climatological southwestern monsoon current in the ocean along the western Indian peninsula [latest description by Schott et al. (2009)] is enhanced (along the western side of the Indian peninsula) even at a depth of 65 m (Fig. 7d; 70°–75°E).
Indeed, the significant responses in the two basins do not appear from the beginning, as ocean processes have large thermal and mass inertia and are relatively slow compared with atmospheric processes. We observed that the 5-m temperature in the BOB is higher than that in the AS in both experiments, and they all maintain a positive tendency in May (Fig. 8c). It takes only 15 days for the AS to exhibit a significant difference, but for the BOB it takes approximately 25 days. This is one of the reasons that we chose to analyze the difference field averaged from 15 to 31 May.
d. Cross-equatorial water transport
In a report by Godfrey and Meyers (1995), the following specific question was asked: “Is there a cross-equatorial transport?” Schott et al. (2002) reviewed observational evidence for such a cross-equatorial cell and found cross-equatorial Ekman transport to be part of its mechanism. Jensen (2003) tracked water floats in a model and found that the AS transports more water from the Northern Hemisphere to the Southern Hemisphere than the BOB in the model. Lutsko et al. (2019) discussed the relation between cross-equator ocean heat transport and the monsoonal circulation, and find that the atmosphere’s cross-equatorial meridional overturning circulation is weakened by the presence of southward heat transport in the Indian Ocean. We have explored the upper-layer response in the northern Indian Ocean to the TP thermal forcing. Now, let us consider whether this kind of response further impacts the tropical and the southern Indian Ocean through an “oceanic bridge” (Broecker 1991; Zhang et al. 1998).
The difference field along the equatorial cross section (Fig. 10) shows that there is significant southward water mass transport across the equator at depths of 30–75 m in the eastern Indian Ocean. This movement is the only horizontal water mass transport area that passes the 90% confidence level of a t test along the equator. As shown before, the Ekman transport in the AS is southeastward, which travels along the western coast of the Indian Peninsula and is directed toward this transport area. Meanwhile, limited by the Malay Peninsula and Sumatra, the water transport anomaly in the BOB also turns south in the mixed layer and is directed toward this area. These phenomena can be clearly observed at 45-m depth in Fig. 11a. As the mixed layer is deeper near the equator (Fig. 11d) and the southward transport occurs at the downwelling branch of the meridional circulation (Fig. 9b), the significant southward transport is not present at 0–30-m depth; instead, it is located at the depth of the bottom of the mixed layer at the equator. In addition, although a nonnegligible enhanced offshore current exists along the equator (Fig. 10a), it is just a peripheral motion of Ekman transport in the AS. Further, although the zonal sea current along 60°–90°E exhibits westward speed, all the zonal current changes along the equator are not significant at the 90% level, according to the Student’s t test.
Difference fields of (a) zonal and (b) meridional current speed (cm s−1), and (c) atmospheric and (d) oceanic heat transport (W m−2), across the equator. Dotted areas indicate the difference passed the 90% significance level in a two-sided t test.
Citation: Journal of Climate 32, 20; 10.1175/JCLI-D-18-0880.1
(a) Difference fields of potential temperature (shaded; K) and ocean current (vectors; cm s−1) at 45-m depth. (b) Difference fields of surface wind stress (vectors; dyne cm−2) and Sverdrup transport (shaded; t m−1 s−1). (c) Difference field of mixed layer depth (m). (d) Mean climate state of the mixed layer depth (m). Dotted areas and red vectors indicate the difference passed the 90% significance level in a two-sided t test.
Citation: Journal of Climate 32, 20; 10.1175/JCLI-D-18-0880.1
4. Conclusions and discussion
By conducting two groups of experiments based on an atmosphere–ocean coupled global climate model (CESM1.2.0) and data diagnosis, our research indicates that the influence of the TP surface heating is not limited to the atmosphere. The air–sea interaction processes are influenced in May so that the surface and subsurface of the Indian Ocean also exhibit significant responses within a month. Some main conclusions are summarized as follows:
Easterly anomalies appear in the upper troposphere due to the enhanced South Asian high. Precipitation is reduced and surface temperature increased in the Indian peninsula. At lower levels, the heat pumping effect of the TP surface heating produces a southwesterly anomaly that reinforces the background airflow, accelerates the sea surface evaporation process, and lowers the temperature of the mixed layer. Also, the vertical advection anomaly plays a significant cooling role along the western coast in the AS and BOB.
Ekman transport induced by sea surface wind stress anomalies causes the northern Indian Ocean mixed layer seawater to be transported toward the southeast. Specifically, as the AS and BOB are located in different positions relative to the TP thermal forcing, the water mass transport is similar but also shows some differences. In the AS, the surface wind anomaly is weaker than that in the BOB, and the eastward component is larger than the northward component, so the southward transport dominates the transport in the mixed layer. In addition, the transport magnitude in the AS is smaller than that in the BOB. In the BOB, the transport is more significant and is mainly from west to east, which leads to stronger upwelling in the northwestern BOB compared to the northwestern AS, and onshore downwellings in the southeastern BOB. The so-called southwest monsoon current in the northern Indian Ocean is also enhanced.
The response of meridional water mass transport to the TP thermal forcing is estimated using the Sverdrup relationship, and the results show that the northern Indian Ocean has significant enhancement of southward transport due to TP thermal forcing. We also note a “water leak” across the equator at depths of 45–85 m in the eastern Indian Ocean, together with significant heat transport from the Northern Hemisphere to the Southern Hemisphere.
Compared with the most closely related recent research (Table 1), this study focuses more on the thermal effects of the TP in May and its impacts on the mixed layer temperature and ocean currents in the Indian Ocean. So far, this is the only research using a “branch-run” technique to separate more direct, instantaneous, short-term influences from those in a state of equilibrium. Wang et al. (2019) and He et al. (2019) used regional models and global climate models, respectively, to study the relationship between summer plateau thermodynamic interactions and sea–air interactions in the Indian Ocean region, pointing out that sea–air coupling weakens the influence of TP thermal effects. At the same time, their conclusion echoes our finding that the SST cools down in the AS and BOB. The difference in this paper is that we use a fully coupled climate model and the starting point of each ensemble member is anchored (to compare the difference between strong and weak TP sensible heat situations in the same year). This allows the initial impulse of the atmosphere on the ocean to be examined before reaching equilibrium—later on in the simulation, separating the influence of the atmosphere from that of the ocean is harder to discern. The advantage of this scheme is to better control variates and avoid the interference of other factors in different years in the continuous integration process (i.e., after several years of integration, the control and the sensitivity experiments are located in the positive and negative phases of ENSO respectively, meaning the results will contain signals of ENSO). In addition, this experimental design allows for a more detailed diagnosis of the temperature equations. Since the temperature tendency corresponds to the difference between the two sets of experiments, the diagnosis is more convenient and the residual term is smaller. Although the seasons we investigated and the experimental design are different from those in Wang et al. (2019) and He et al. (2019), the final results have certain similarities, which implies that the situation in May may be closer to that in summer, and to some extent it shows that the short-term response has a certain consistency with the equilibrium response. Another difference is that they pay more attention to the feedback effects of air–sea coupling on the atmosphere, while we focus more on the direct response in the ocean. With an SST nudging technique in a CGCM, Baldwin et al. (2019) explored the response of Asian climate to Asian orography in the prebalanced state. They used a 5-day restoring time scale, effectively removing the influence of flattening Asian topography on climatological SSTs, and found that the AS and the BOB exhibit offshore upwellings during April–June because of the effects of Asian topography—a phenomenon very similar to that seen in Fig. 4d in this paper. Meanwhile, our branch-run results give a more detailed and continuous description to the evolution of this prebalanced state, which can barely be observed with the 5-day nudging.
Wang et al. (2009) pointed out that the interannual predictability of the ASM is still low, which may be due to the lack of insight and comprehension of interactions between its components. Therefore, the results of this study can be helpful insofar as, if we observe strong TP heating in May, we can infer that the northern Indian Ocean may be colder in less than one month and that the related ocean current may change the evolution of the Indian Ocean basin mode or Indian Ocean dipole mode, thus providing a better comprehension of the ASM in that year. However, the details of applying and validating this proposal deserve further study.
In this study, we discuss only the changes in sea temperature and currents. The thermohaline circulation, which is driven by global seawater density gradients, is a very slow process, so it is difficult to see a significant response to short-term atmospheric change. We note that the results of the CGCM may have large biases when the ocean is involved, but the short-term integration involves less nonlinear magnification of simulation error, reflecting a more direct response. Furthermore, to rule out being misled by random noises in the experiments as much as possible, all the results in this paper are accompanied by statistical tests. Although the model dependency problem cannot be completely excluded, the interference of uncertainty is minimized to some extent. In addition, we only focus on the TP surface heat source in May, during which it has the largest intensity and variance. The role it plays in the premonsoon background and the delayed effect of TP thermal forcing and the related ocean response in the following summer are also worthy of further exploration.
When exploring the relationship between the Indian Ocean SST and ENSO, not only the “atmospheric bridge” (Alexander et al. 2002; Behera et al. 2006) but also the related ocean processes and their persistence need to be considered. In addition, although the Indian Ocean basin mode behaves as a uniformly varying basin, it also has a certain northward SST gradient in the northern Indian Ocean. Whether this phenomenon is related to the process described in this study deserves further discussion. In addition, although we already know that ENSO can affect the ASM via the Indian Ocean basin mode’s capacitor effect across different seasons, in years without a strong ENSO signal, the process described in this paper might be helpful when investigating the interannual variability of the ASM system.
Acknowledgments
This study was jointly funded by the National Natural Science Foundation of China (Grants 91637312 and 41505049). The paper has also benefited from the comments of Dr. Hailong Liu, Dr. Yongxiang Huang, and Dr. Wei Zhuang. They provided many insightful suggestions about the analysis of the ocean.
REFERENCES
Abe, M., A. Kitoh, and T. Yasunari, 2003: An evolution of the Asian summer monsoon associated with mountain uplift—Simulation with the MRI atmosphere–ocean coupled GCM. J. Meteor. Soc. Japan, 81, 909–933, https://doi.org/10.2151/JMSJ.81.909.
Abe, M., T. Yasunari, and A. Kitoh, 2005: Sensitivity of the central Asian climate to uplift of the Tibetan Plateau in the coupled climate model (MRI-CGCM1). Isl. Arc, 14, 378–388, https://doi.org/10.1111/j.1440-1738.2005.00493.x.
Abe, M., M. Hori, T. Yasunari, and A. Kitoh, 2013: Effects of the Tibetan Plateau on the onset of the summer monsoon in South Asia: The role of the air–sea interaction. J. Geophys. Res. Atmos., 118, 1760–1776, https://doi.org/10.1002/jgrd.50210.
Alexander, M. A., I. Bladé, M. Newman, J. R. Lanzante, N.-C. Lau, and J. D. Scott, 2002: The atmospheric bridge: The influence of ENSO teleconnections on air–sea interaction over the global oceans. J. Climate, 15, 2205–2231, https://doi.org/10.1175/1520-0442(2002)015<2205:TABTIO>2.0.CO;2.
Baldwin, J. W., G. A. Vecchi, and S. Bordoni, 2019: The direct and ocean-mediated influence of Asian orography on tropical precipitation and cyclones. Climate Dyn., 53, 805–824, https://doi.org/10.1007/S00382-019-04615-5.
Behera, S. K., J. J. Luo, S. Masson, S. A. Rao, H. Sakuma, and T. Yamagata, 2006: A CGCM study on the interaction between IOD and ENSO. J. Climate, 19, 1688–1705, https://doi.org/10.1175/JCLI3797.1.
Behringer, D. W., and Y. Xue, 2004: Evaluation of the Global Ocean Data Assimilation System at NCEP: The Pacific Ocean. Eighth Symp. on Integrated Observing and Assimilation Systems for Atmosphere, Oceans, and Land, Seattle, WA, Amer. Meteor. Soc., 2.3.
Bjerknes, J., 1969: Atmospheric teleconnections from the equatorial Pacific. Mon. Wea. Rev., 97, 163–172, https://doi.org/10.1175/1520-0493(1969)097<0163:ATFTEP>2.3.CO;2.
Bolin, B., 1950: On the influence of the earth’s orography on the general character of the westerlies. Tellus, 2, 184–195, https://doi.org/10.3402/tellusa.v2i3.8547.
Boos, W. R., and Z. Kuang, 2010: Dominant control of the South Asian monsoon by orographic insulation versus plateau heating. Nature, 463, 218–222, https://doi.org/10.1038/nature08707.
Boos, W. R., and Z. Kuang, 2013: Sensitivity of the South Asian monsoon to elevated and non-elevated heating. Sci. Rep., 3, 1192, https://doi.org/10.1038/srep01192.
Broecker, W. S., 1991: The great ocean conveyor. Oceanography, 4, 79–89, https://doi.org/10.5670/oceanog.1991.07.
Cayan, D. R., 1992: Latent and sensible heat flux anomalies over the northern oceans: Driving the sea surface temperature. J. Phys. Oceanogr., 22, 859–881, https://doi.org/10.1175/1520-0485(1992)022<0859:LASHFA>2.0.CO;2.
Chang, C., and T. Li, 2000: A theory for the tropical tropospheric biennial oscillation. J. Atmos. Sci., 57, 2209–2224, https://doi.org/10.1175/1520-0469(2000)057<2209:ATFTTT>2.0.CO;2.
Chen, J., and S. Bordoni, 2014: Orographic effects of the Tibetan Plateau on the East Asian summer monsoon: An energetic perspective. J. Climate, 27, 3052–3072, https://doi.org/10.1175/JCLI-D-13-00479.1.
Du, Y., S.-P. Xie, G. Huang, and K. Hu, 2009: Role of air–sea interaction in the long persistence of El Niño–induced north Indian Ocean warming. J. Climate, 22, 2023–2038, https://doi.org/10.1175/2008JCLI2590.1.
Duan, A., and G. Wu, 2005: Role of the Tibetan Plateau thermal forcing in the summer climate patterns over subtropical Asia. Climate Dyn., 24, 793–807, https://doi.org/10.1007/s00382-004-0488-8.
Duan, A., and G. Wu, 2008: Weakening trend in the atmospheric heat source over the Tibetan Plateau during recent decades. Part I: Observations. J. Climate, 21, 3149–3164, https://doi.org/10.1175/2007JCLI1912.1.
Duan, A., C. Sui, and G. Wu, 2008: Simulation of local air–sea interaction in the great warm pool and its influence on Asian monsoon. J. Geophys. Res., 113, D22105, https://doi.org/10.1029/2008JD010520.
Duan, A., F. Li, M. Wang, and G. Wu, 2011: Persistent weakening trend in the spring sensible heat source over the Tibetan Plateau and its impact on the Asian summer monsoon. J. Climate, 24, 5671–5682, https://doi.org/10.1175/JCLI-D-11-00052.1.
Ebita, A., and Coauthors, 2011: The Japanese 55-year reanalysis “JRA-55”: An interim report. Sci. Online Lett. Atmos., 7, 149–152, https://doi.org/10.2151/SOLA.2011-038.
Fallah, B., U. Cubasch, K. Prömmel, and S. Sodoudi, 2016: A numerical model study on the behaviour of Asian summer monsoon and AMOC due to orographic forcing of Tibetan Plateau. Climate Dyn., 47, 1485–1495, https://doi.org/10.1007/s00382-015-2914-5.
Gent, P. R., and Coauthors, 2011: The Community Climate System Model version 4. J. Climate, 24, 4973–4991, https://doi.org/10.1175/2011JCLI4083.1.
Godfrey, J. S., and G. Meyers, 1995: The role of the Indian Ocean in the global climate system: Recommendations regarding the global ocean observing system. Report of the Ocean Observing System Development Panel, Texas A&M University, 88 pp.
Gualdi, S., E. Guilyardi, A. Navarra, S. Masina, and P. Delecluse, 2003: The interannual variability in the tropical Indian Ocean as simulated by a CGCM. Climate Dyn., 20, 567–582, https://doi.org/10.1007/s00382-002-0295-z.
Hahn, D. G., and S. Manabe, 1975: The role of mountains in the South Asian monsoon circulation. J. Atmos. Sci., 32, 1515–1541, https://doi.org/10.1175/1520-0469(1975)032<1515:TROMIT>2.0.CO;2.
Hastenrath, S., 2012: Climate Dynamics of the Tropics. Springer, 462 pp.
He, B., G. Wu, Y. Liu, and Q. Bao, 2015: Astronomical and hydrological perspective of mountain impacts on the Asian summer monsoon. Sci. Rep., 5, 17586, https://doi.org/10.1038/srep17586.
He, B., Y. Liu, G. Wu, Z. Wang, and Q. Bao, 2019: The role of air–sea interactions in regulating the thermal effect of the Tibetan–Iranian Plateau on the Asian summer monsoon. Climate Dyn., 52, 4227–4245, https://doi.org/10.1007/S00382-018-4377-Y.
Hoskins, B. J., and D. J. Karoly, 1981: The steady linear response of a spherical atmosphere to thermal and orographic forcing. J. Atmos. Sci., 38, 1179–1196, https://doi.org/10.1175/1520-0469(1981)038<1179:TSLROA>2.0.CO;2.
Hoskins, B. J., M. McIntyre, and A. W. Robertson, 1985: On the use and significance of isentropic potential vorticity maps. Quart. J. Roy. Meteor. Soc., 111, 877–946, https://doi.org/10.1002/qj.49711147002.
Hu, J., and A. Duan, 2015: Relative contributions of the Tibetan Plateau thermal forcing and the Indian Ocean Sea surface temperature basin mode to the interannual variability of the East Asian summer monsoon. Climate Dyn., 45, 2697–2711, https://doi.org/10.1007/s00382-015-2503-7.
IPCC, 2014: Climate Change 2013: The Physical Science Basis. T. F. Stocker et al., Eds., Cambridge University Press, 1535 pp.
Jensen, T. G., 2003: Cross-equatorial pathways of salt and tracers from the northern Indian Ocean: Modelling results. Deep-Sea Res. II, 50, 2111–2127, https://doi.org/10.1016/S0967-0645(03)00048-1.
Jin, R., Z. Wu, and P. Zhang, 2018: Tibetan Plateau capacitor effect during the summer preceding ENSO: From the Yellow River climate perspective. Climate Dyn., 51, 57–71, https://doi.org/10.1007/s00382-017-3906-4.
Kay, J., and Coauthors, 2015: The Community Earth System Model (CESM) large ensemble project: A community resource for studying climate change in the presence of internal climate variability. Bull. Amer. Meteor. Soc., 96, 1333–1349, https://doi.org/10.1175/BAMS-D-13-00255.1.
Kitoh, A., 1997: Mountain uplift and surface temperature changes. Geophys. Res. Lett., 24, 185–188, https://doi.org/10.1029/96GL03953.
Kitoh, A., 2004: Effects of mountain uplift on East Asian summerclimate investigated by a coupled atmosphere–ocean GCM. J. Climate., 17, 783–802, https://doi.org/10.1175/1520-0442(2004)017,0783:EOMUOE.2.0.CO;2.
Kitoh, A., 2007: ENSO modulation by mountain uplift. Climate Dyn., 28, 781–796, https://doi.org/10.1007/s00382-006-0209-6.
Kitoh, A., T. Motoi, and O. Arakawa, 2010: Climate modelling study on mountain uplift and Asian monsoon evolution. Geol. Soc. Lond. Spec. Publ., 342, 293–301, https://doi.org/10.1144/SP342.17.
Klein, S. A., B. J. Soden, and N.-C. Lau, 1999: Remote sea surface temperature variations during ENSO: Evidence for a tropical atmospheric bridge. J. Climate, 12, 917–932, https://doi.org/10.1175/1520-0442(1999)012<0917:RSSTVD>2.0.CO;2.
Kobayashi, S., and Coauthors, 2015: The JRA-55 reanalysis: General specifications and basic characteristics. J. Meteor. Soc. Japan, 93, 5–48, https://doi.org/10.2151/jmsj.2015-001.
Lau, N.-C., and M. J. Nath, 1996: The role of the “atmospheric bridge” in linking tropical Pacific ENSO events to extratropical SST anomalies. J. Climate, 9, 2036–2057, https://doi.org/10.1175/1520-0442(1996)009<2036:TROTBI>2.0.CO;2.
Li, C., and M. Yanai, 1996: The onset and interannual variability of the Asian summer monsoon in relation to land–sea thermal contrast. J. Climate, 9, 358–375, https://doi.org/10.1175/1520-0442(1996)009<0358:TOAIVO>2.0.CO;2.
Li, S., J. Lu, G. Huang, and K. Hu, 2008: Tropical Indian Ocean basin warming and East Asian summer monsoon: A multiple AGCM study. J. Climate, 21, 6080–6088, https://doi.org/10.1175/2008JCLI2433.1.
Li, T., B. Wang, C. Chang, and Y. Zhang, 2003: A theory for the Indian Ocean dipole–zonal mode. J. Atmos. Sci., 60, 2119–2135, https://doi.org/10.1175/1520-0469(2003)060<2119:ATFTIO>2.0.CO;2.
Liu, S., and A. Duan, 2017: Impacts of the leading modes of tropical Indian Ocean sea surface temperature anomaly on sub-seasonal evolution of the circulation and rainfall over East Asia during boreal spring and summer. J. Meteor. Res., 31, 171–186, https://doi.org/10.1007/s13351-016-6093-z.
Liu, Y., and G. Wu, 2000: Reviews on the study of the subtropical anticyclone and new insights on some fundamental problems (in Chinese). Acta Meteor. Sin., 58, 500–512.
Liu, Y., Z. Wang, H. Zhuo, and G. Wu, 2017: Two types of summertime heating over Asian large-scale orography and excitation of potential-vorticity forcing. II. Sensible heating over Tibetan-Iranian Plateau. Sci. China Earth Sci., 60, 733–744, https://doi.org/10.1007/s11430-016-9016-3.
Lutsko, N. J., J. Marshall, and B. Green, 2019: Modulation of monsoon circulations by cross-equatorial ocean heat transport. J. Climate, 32, 3471–3485, https://doi.org/10.1175/JCLI-D-18-0623.1.
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, 155–170, https://doi.org/10.1007/s00704-006-0265-1.
Neale, R. B., J. Richter, S. Park, P. H. Lauritzen, S. J. Vavrus, P. J. Rasch, and M. Zhang, 2013: The mean climate of the Community Atmosphere Model (CAM4) in forced SST and fully coupled experiments. J. Climate, 26, 5150–5168, https://doi.org/10.1175/JCLI-D-12-00236.1.
Okajima, H., and S.-P. Xie, 2007: Orographic effects on the northwestern Pacific monsoon: Role of air–sea interaction. Geophys. Res. Lett., 34, L21708, https://doi.org/10.1029/2007GL032206.
Paulson, C. A., and J. J. Simpson, 1977: Irradiance measurements in the upper ocean. J. Phys. Oceanogr., 7, 952–956, https://doi.org/10.1175/1520-0485(1977)007<0952:IMITUO>2.0.CO;2.
Qiu, B., 2000: Interannual variability of the Kuroshio Extension system and its impact on the wintertime SST field. J. Phys. Oceanogr., 30, 1486–1502, https://doi.org/10.1175/1520-0485(2000)030<1486:IVOTKE>2.0.CO;2.
Rayner, N., and Coauthors, 2003: Global analyses of sea surface temperature, sea ice, and night marine air temperature since the late nineteenth century. J. Geophys. Res., 108, 4407, https://doi.org/10.1029/2002JD002670.
Rodwell, M. J., and B. J. Hoskins, 1996: Monsoons and the dynamics of deserts. Quart. J. Roy. Meteor. Soc., 122, 1385–1404, https://doi.org/10.1002/qj.49712253408.
Rodwell, M. J., and B. J. Hoskins, 2001: Subtropical anticyclones and summer monsoons. J. Climate, 14, 3192–3211, https://doi.org/10.1175/1520-0442(2001)014<3192:SAASM>2.0.CO;2.
Saji, N., B. Goswami, P. Vinayachandran, and T. Yamagata, 1999: A dipole mode in the tropical Indian Ocean. Nature, 401, 360–363, https://doi.org/10.1038/43854.
Schott, F. A., and J. P. McCreary Jr., 2001: The monsoon circulation of the Indian Ocean. Prog. Oceanogr., 51, 1–123, https://doi.org/10.1016/S0079-6611(01)00083-0.
Schott, F. A., M. Dengler, and R. Schoenefeldt, 2002: The shallow overturning circulation of the Indian Ocean. Prog. Oceanogr., 53, 57–103, https://doi.org/10.1016/S0079-6611(02)00039-3.
Schott, F. A., S. P. Xie, and J. P. McCreary, 2009: Indian Ocean circulation and climate variability. Rev. Geophys., 47, RG1002, https://doi.org/10.1029/2007RG000245.
Sperber, K., and Coauthors, 2013: The Asian summer monsoon: An intercomparison of CMIP5 vs. CMIP3 simulations of the late 20th century. Climate Dyn., 41, 2711–2744, https://doi.org/10.1007/s00382-012-1607-6.
Stewart, R. H., 2008: Introduction to Physical Oceanography. Texas A & M University College Station, 352 pp.
Sugihara, G., R. May, H. Ye, C.-H. Hsieh, E. Deyle, M. Fogarty, and S. Munch, 2012: Detecting causality in complex ecosystems. Science, 338, 496–500, https://doi.org/10.1126/science.1227079.
Tang, H., A. Micheels, J. T. Eronen, B. Ahrens, and M. Fortelius, 2013: Asynchronous responses of East Asian and Indian summer monsoons to mountain uplift shown by regional climate modelling experiments. Climate Dyn., 40, 1531–1549, https://doi.org/10.1007/s00382-012-1603-x.
Wang, B., 2006: The Asian Monsoon. Springer, 683 pp.
Wang, B., and Coauthors, 2009: Advance and prospectus of seasonal prediction: Assessment of the APCC/CliPAS 14-model ensemble retrospective seasonal prediction (1980–2004). Climate Dyn., 33, 93–117, https://doi.org/10.1007/s00382-008-0460-0.
Wang, M., S. Zhou, and A. Duan, 2012: Trend in the atmospheric heat source over the central and eastern Tibetan Plateau during recent decades: Comparison of observations and reanalysis data. Chin. Sci. Bull., 57, 548–557, https://doi.org/10.1007/s11434-011-4838-8.
Wang, Z., A. Duan, M. Li, and B. He, 2016: Influences of thermal forcing over the slope/platform of the Tibetan Plateau on Asian summer monsoon: Numerical studies with the WRF model. Chin. J. Geophys., 59, 474–487, https://doi.org/10.1002/cjg2.30007.
Wang, Z., A. Duan, and S. Yang, 2019: Potential regulation on the climatic effect of Tibetan Plateau heating by tropical air–sea coupling in regional models. Climate Dyn., 52, 1685–1694, https://doi.org/10.1007/S00382-018-4218-Z.
Webster, P. J., A. M. Moore, J. P. Loschnigg, and R. R. Leben, 1999: Coupled ocean–atmosphere dynamics in the Indian Ocean during 1997–98. Nature, 401, 356–360, https://doi.org/10.1038/43848.
Wu, G., 1984: The nonlinear response of the atmosphere to large-scale mechanical and thermal forcing. J. Atmos. Sci., 41, 2456–2476, https://doi.org/10.1175/1520-0469(1984)041<2456:TNROTA>2.0.CO;2.
Wu, G., W. Li, H. Guo, and H. Liu, 1997: Sensible heat driven air-pump over the Tibetan Plateau and its impacts on the Asian summer monsoon. Collections in the Memory of Zhao Jiuzhang, Ye Duzheng, Ed., Chinese Science Press, 116–126.
Wu, G., and Coauthors, 2007: The influence of mechanical and thermal forcing by the Tibetan Plateau on Asian climate. J. Hydrometeor., 8, 770–789, https://doi.org/10.1175/JHM609.1.
Wu, G., Y. Liu, B. He, Q. Bao, A. Duan, and F.-F. Jin, 2012: Thermal controls on the Asian summer monsoon. Sci. Rep., 2, 404, https://doi.org/10.1038/srep00404.
Wu, G., and Coauthors, 2015: Tibetan Plateau climate dynamics: Recent research progress and outlook. Natl. Sci. Rev., 2, 100–116, https://doi.org/10.1093/nsr/nwu045.
Wu, Z., J. Li, Z. Jiang, and T. Ma, 2012: Modulation of the Tibetan Plateau snow cover on the ENSO teleconnections: From the East Asian summer monsoon perspective. J. Climate, 25, 2481–2489, https://doi.org/10.1175/JCLI-D-11-00135.1.
Wu, Z., P. Zhang, H. Chen, and Y. Li, 2016: Can the Tibetan Plateau snow cover influence the interannual variations of Eurasian heat wave frequency? Climate Dyn., 46, 3405–3417, https://doi.org/10.1007/s00382-015-2775-y.
Wyrtki, K., 1973: An equatorial jet in the Indian Ocean. Science, 181, 262–264, https://doi.org/10.1126/science.181.4096.262.
Xie, S.-P., H. Annamalai, F. A. Schott, and J. P. McCreary Jr., 2002: Structure and mechanisms of South Indian Ocean climate variability. J. Climate, 15, 864–878, https://doi.org/10.1175/1520-0442(2002)015<0864:SAMOSI>2.0.CO;2.
Xie, S.-P., K. Hu, J. Hafner, H. Tokinaga, Y. Du, G. Huang, and T. Sampe, 2009: Indian Ocean capacitor effect on Indo–western Pacific climate during the summer following El Niño. J. Climate, 22, 730–747, https://doi.org/10.1175/2008JCLI2544.1.
Xie, S.-P., Y. Kosaka, Y. Du, K. Hu, J. S. Chowdary, and G. Huang, 2016: Indo-western Pacific Ocean capacitor and coherent climate anomalies in post-ENSO summer: A review. Adv. Atmos. Sci., 33, 411–432, https://doi.org/10.1007/s00376-015-5192-6.
Yamagata, T., S. K. Behera, J.-J. Luo, S. Masson, M. R. Jury, and S. A. Rao, 2004: Coupled ocean–atmosphere variability in the tropical Indian Ocean. Earth’s Climate: The Ocean–Atmosphere Interaction, Geophys. Monogr., Vol. 147, Amer. Geophys. Union, 189–212, https://doi.org/10.1029/147GM12.
Yanai, M., and C. Li, 1994: Mechanism of heating and the boundary layer over the Tibetan Plateau. Mon. Wea. Rev., 122, 305–323, https://doi.org/10.1175/1520-0493(1994)122<0305:MOHATB>2.0.CO;2.
Yanai, M., C. Li, and Z. Song, 1992: Seasonal heating of the Tibetan Plateau and its effects on the evolution of the Asian summer monsoon. J. Meteor. Soc. Japan, 70, 319–351, https://doi.org/10.2151/JMSJ1965.70.1B_319.
Yang, J., Q. Liu, S.-P. Xie, Z. Liu, and L. Wu, 2007: Impact of the Indian Ocean SST basin mode on the Asian summer monsoon. Geophys. Res. Lett., 34, L02708, https://doi.org/10.1029/2006GL028571.
Yasunari, T., K. Saito, and K. Takata, 2006: Relative roles of large-scale orography and land surface processes in the global hydroclimate. Part I: Impacts on monsoon systems and the tropics. J. Hydrometeor., 7, 626–641, https://doi.org/10.1175/JHM515.1.
Ye, D., and Y.-X. Gao, 1979: The Meteo