Topographic insulation is one of the primary origins for the influence of the Tibetan Plateau (TP) on Asian climate. The Yunnan–Guizhou (YG) Plateau, at the southeastern margin of the TP, is known to block the northern branch of the Indian monsoon circulation in summer. However, it is an open question whether this blocking feeds back to the monsoon. In this study, the effect of the YG topography on the Indian monsoon and its comparison with that of the TP were evaluated using general circulation model experiments. The results showed that the TP strengthens the monsoon precipitation, especially during the onset. However, the YG topography significantly weakens the monsoon. With the YG topography, strengthened low-level airflow around the YG Plateau induces anomalous anticyclonic winds to the southwest, and the changes remodulate the whole circulation structure over Asia. As a result, the Indian monsoon becomes weakened from the Bay of Bengal to the Indian subcontinent and Arabian Sea, as does the associated precipitation. In addition, the YG topography affects the anomalous warming center over the TP and the precipitation during the monsoon onset. The YG-reduced summer precipitation occupied approximately one-third of the total increment compared to the entire TP. The Indian monsoon weakened by YG topography distinctly opposes the traditional paleoclimatic viewpoint that all of the TP topography contributes to the monsoon strengthening. In fact, the climatic effect of the TP depends closely upon both its central and marginal topography, and the topography of its subterrains does not necessarily play a similar role.
Mountains, produced by Earth’s tectonic activity, are one of the primary factors modulating the formation of modern climate. The Tibetan Plateau (TP), the largest plateau on Earth, has long been thought to play an important role in Asian and even global climate change (e.g., Hahn and Manabe 1975; Kutzbach et al. 1989; Manabe and Broccoli 1990; Prell and Kutzbach 1992; An et al. 2001; Molnar et al. 2010; An et al. 2015). Via its thermal and mechanical effects, the TP can intensify the general circulation pattern over the Northern Hemisphere (Manabe and Terpstra 1974; Kutzbach et al. 1989; Shi et al. 2015), as well as the coupled arid–monsoon climate over Asia (Manabe and Broccoli 1990; Kutzbach et al. 1989; Liu and Yin 2002; Boos and Kuang 2010; Shi et al. 2011; Ma et al. 2014).
In summer, the TP is a heat source; it heats the atmosphere and amplifies the thermal gradient between itself and ocean to the south. One school of thought attributes the TP-induced strengthening of the Indian monsoon circulation and precipitation to this summer heating (Hahn and Manabe 1975; Kutzbach et al. 1989; Wu and Zhang 1998; Wu et al. 2012). In these studies, the changes in the thermal conditions over the TP and the importance of sensible heating on the Indian summer monsoon have formed the major focus (e.g., Yanai et al. 1992; Yanai and Li 1994; Rajagopalan and Molnar 2013). Latent heat from the Bay of Bengal also contributes to the anomalous high-tropospheric warming center around the TP (Tamura et al. 2010).
From a mechanical perspective, the TP forces the westerly jet into two branches, and it also blocks the northward transport of moisture from the Indian Ocean (Kutzbach et al. 1989; Manabe and Broccoli 1990; Broccoli and Manabe 1992). In contrast to the thermal effect, it is difficult to quantify the mechanical effect of the TP precisely. Based on numerical experiments, the Himalaya Mountains are believed to act as a barrier in isolating the northern cold dry air and thus potential thermal influence of the TP on the Indian monsoon circulation (Boos and Kuang 2010), emphasizing that the blocking effect of the Himalayas or the southern TP may be more significant than the thermal effect. Except for the mature period of the Indian monsoon, the mechanical impact of the TP is also evident in its onset and seasonal evolution (Park et al. 2012).
Although the thermal and mechanical effects of the TP on the climate are widely acknowledged, most theoretical studies have concentrated on the main TP–Himalaya topography. Recent studies have raised the idea that the evolutions of different climate systems over Asia respond sensitively to the uplift of different parts of the TP (Boos and Kuang 2010; Liu and Dong 2013). This suggests that we should ascribe the reconstructed Asian climate change to the subregional TP uplift and not simply to the whole TP. The intensification of the Indian summer monsoon may be closely associated with the uplift of the Himalayas and southern TP (Boos and Kuang 2010), while the uplift of the northern TP is more likely responsible for the formation of the East Asian monsoon, especially the northern part (Zhang et al. 2012; Tang et al. 2013), and the inland arid climate (Shi et al. 2011; Liu et al. 2015).
In most numerical experiments, the defined topography is so coarse that only the main TP can be resolved. The real TP topography is simplified and idealized, ignoring the small mountains at the margin of, and around, the main TP, which might lead to an inaccurate estimate of the real effect of the TP. For instance, the Mongolian Plateau, a smaller area of terrain located to the north of the TP, can have a marked effect on the stationary planetary wave pattern and the winter monsoon system over East Asia (Shi et al. 2015; Sha et al. 2015). Facilitated by its more northerly location, the Mongolian Plateau forces the westerly winds flowing around it to shift more northward and ultimately exerts a larger influence on the westerly jet than the TP (Shi et al. 2015). These analyses tell us that the effects of smaller terrains at sensitive locations for specific climate systems may be unexpectedly important. Given that it would be difficult for these mountains to exert similar thermal influences as the main TP because of their smaller size, we can speculate that any noticeable response of the climate system might come primarily from the mechanical effect.
Following this idea, it seems important to examine the climatic effect of the Yunnan–Guizhou (YG) Plateau, which seems to be located in a sensitive zone. The YG Plateau is a stretch of terrain belonging to the southeastern area of the TP. Compared to the main TP, the YG topography is much smaller in size, with an average height of less than 2000 m. However, the YG Plateau is located within the transition region between the Indian and East Asian summer monsoon circulations. In summer, the Indian monsoon winds flow eastward over the Indian subcontinent (Fig. 1a). When passing the eastern Bay of Bengal, they are divided into two branches; the southern branch continues to move eastward to the South China Sea, while the northern branch turns northeastward and is then blocked by the YG Plateau. From the circulation pattern, it can be speculated that the Indian monsoon winds would flow directly to southeastern China if the YG Plateau did not exist. In other words, the presence of the YG Plateau forces the winds to flow around the terrain.
In this study, the effect of the YG Plateau on the Indian summer monsoon circulation and associated precipitation, as well as its comparison to the entire TP, was evaluated using climate model experiments. The response of the anomalous heating center over the TP during the monsoon onset is also examined. The experimental design is introduced in section 2. The results are presented and discussed in sections 3 and 4, respectively. Section 5 provides a summary of the key findings.
2. Model experiments and evaluation
Three numerical experiments using a general circulation model were conducted to evaluate the orographic effect of the YG Plateau on the Indian monsoon and its relative contribution compared to the TP. The model used was the Community Atmosphere Model, version 3 (CAM3), developed by the National Center for Atmospheric Research. CAM3 is an atmospheric model in which a land surface model (Community Land Model, version 3) is coupled. It performs well in its simulation of the Asian monsoon and is widely employed in monsoon research, including the examination of monsoon sensitivity to mountain topography (Shi et al. 2011; Sha et al. 2015). However, CAM3, like most climate models, is biased in its simulation of the thermodynamic structure of the Indian summer monsoon, including both the upper-tropospheric temperature and surface air moist static energy. This bias has been shown in other models to be associated with the overly smoothed topography west of the TP (Boos and Hurley 2013).
In the control experiment (TP1YG1), the modern global topography was employed. Then, in the other two runs, the entire TP topography including the YG (TP0YG0) and only the YG topography, not Tibet (TP1YG0), were removed from the global topography in sensitivity experiments. In TP0YG0, we flattened all the TP-related high terrains in the model to 400 m. In TP1YG0, we flattened only the southeastern margin of the TP where the height is greater than 800 m and less than 2000 m, flattening a maximum height of 800 m (Fig. 2). All boundary conditions except the elevation of topography and the atmospheric carbon dioxide concentration were set to present-day values in the three experiments. The Hadley Centre global sea surface temperature (SST) data (Rayner et al. 2003) are employed as the climatologically averaged SST. The atmospheric carbon dioxide concentration was kept at 280 ppmv, the preindustrial value. The changes in the vegetation cover and the boundary layer roughness associated with the modified topography and their potential climatic feedbacks are not taken into consideration. The gravity wave drag is also not altered. The horizontal resolution was T85, corresponding to approximately 1.4° × 1.4°, which resolves the smaller YG Plateau but does not capture its details. All experiments were integrated for 15 model years after a 5-yr spinup period. The 15-yr-mean outputs were calculated and analyzed in order to evaluate the response of the Indian monsoon to the YG topography.
The control experiment performance with respect to the Indian summer monsoon, including the monsoonal circulation and precipitation, was examined first using ERA-Interim with a resolution of 0.75° × 0.75° (Dee et al. 2011; Fig. 1). In the TP1YG1 experiment, the simulated 850-hPa wind vectors presented a similar circulation pattern as observed over the Indian monsoon region, with a few notable differences. In the northern Bay of Bengal, the simulated westerly winds turned northward to the YG Plateau (Fig. 1b), but they turn too far northward such that they are southerly with almost no branched westerly component. These southerly winds appear to intersect the TP and get branched toward the YG rather than flowing straight to the YG. This bias might somewhat affect our evaluation of the YG topography. Another bias is that the main westerly winds observed around the southwestern TP were not captured by the model; instead it simulates easterly winds in this region. However, overall the model can successfully capture the relevant circulation phenomena, in particular that the Indian monsoonal winds flow around the YG Plateau when facing it, allowing us to evaluate the climatic effect of this topography. The modeled and observed distribution of summer precipitation rates, associated with the Indian monsoon circulation, were also in general qualitative agreement. The modeled absolute value of the simulated precipitation maximum was larger over the eastern Arabian Sea, and the precipitation center over the Bay of Bengal is located too far to the southwest.
The responses of seasonal precipitation over the Indian monsoon region to the YG topography and the entire TP were examined first (Fig. 3). Compared to that in the TP1YG0 experiment, April to August precipitation is consistently suppressed in the TP1YG1 experiment (Fig. 3a), indicating that the YG topography weakens the Indian monsoon precipitation during both the mature monsoon period and the monsoon onset. However, the precipitation anomalies between the TP1YG1 and TP1YG0 experiments become positive in the non-summer-monsoon months (September to November and January to March) other than December. As a result, the seasonality of the Indian monsoon climate is amplified in the TP1YG0 experiment, compared with TP1YG1. This effect of the YG topography is different from that of the entire TP, shown as the difference between the TP1YG1 and TP0YG0 experiments, which tends to intensify the precipitation throughout the whole year. The most sensitive response of precipitation to the TP occurs mainly during the monsoon onset (May and June), which supports the view that the onset of the Indian monsoon is closely associated with the thermal and mechanical change of the TP (Yanai et al. 1992; Wu and Zhang 1998; Park et al. 2012; Tamura et al. 2010).
The ratios of the differences between TP1YG1 and TP1YG0 to those between TP1YG1 and TP0YG0 were then calculated (Fig. 3b) as the contribution of the YG Plateau to the Indian monsoon rainfall relative to the TP as a whole. From April to August, the proportions are negative, indicating that the YG Plateau exerts a negative effect on the TP-induced increase in precipitation during the onset and mature periods of the monsoon. As a summertime average, the YG-induced precipitation decrease equals to approximately 38% of the total increment by the entire TP. In autumn and winter, the influence of the YG Plateau on Indian precipitation is generally similar to that of the TP and occupies a nonnegligible proportion of the TP’s overall effect. This analysis indicates that the YG Plateau exerts an important role in the Indian monsoonal precipitation change, although it is much smaller in height and size.
To evaluate the role of the YG Plateau in detail, the Indian monsoon season in the following analysis is divided into two stages: the mature period [June–August (JJA)] and the onset (May).
b. Monsoon mature period (JJA)
From the TP1YG1–TP0YG0 difference, it can be seen that the response of monsoon precipitation during the mature period is quite sensitive to the topography change, and the precipitation anomaly induced by the entire TP is found positive over most of the Indian monsoon areas and also the TP itself (Fig. 4a). This precipitation change is closely associated with strengthened southwesterly wind over the Arabian Sea and the cyclonic winds over the northern Bay of Bengal, which results in significant anomalous ascending motion over the Indian monsoon region (Fig. 4c). Consistent with previous studies (Kutzbach et al. 1989; An et al. 2001), our results also demonstrate that the TP intensifies the Indian summer monsoon. However, the YG topography does not increase the rainfall even though it is a subregion within the TP. On the contrary, the precipitation is significantly weakened over the Indian continent and Bay of Bengal by the YG (Fig. 4b). According to 850-hPa wind vectors, the YG topography in the model forces the surface winds to turn southward and an anticyclone anomaly (center at approximately 25° N, 95° E) is induced to the southwest of the YG Plateau (Fig. 4d). Across the Bay of Bengal, Indian subcontinent, and Arabian Sea, the westerly monsoon winds and the ascending motion are decreased, which indicates that the YG Plateau plays an opposite role with the entire TP in the Indian monsoon formation. In addition, obvious but complicated changes in the circulation and rainfall over subtropical East Asia are apparent (Figs. 4b,d), highlighting a possible effect of the YG on the East Asian summer monsoon; this will be pursued in future work.
Because of the TP, the 850-hPa temperature is decreased over the Indian Ocean and subcontinent but increased over the midlatitude Eurasian continent, especially over arid central Asia (Fig. 5a), intensifying the cross-equatorial thermal gradient. This continental warming leads to a strengthening of a low pressure cell to the southwest of the TP (Fig. 5c), which facilitates the northward development of the summer monsoon. For the YG Plateau, the response of temperature to the topography is quite different. Compared with experiment TP1YG0, without the YG topography, the YG Plateau in TP1YG1 causes an obvious cooling over the Arabian Peninsula but a warming over the equatorial western Indian Ocean (Fig. 5b); this temperature change corresponds to weakening of low pressure (Fig. 5d) and subsequently the cross-equatorial airstream, as seen in Fig. 4d.
In the upper atmosphere, an anomalous anticyclone–cyclone pattern is simulated over central Asia and south of the TP between the TP1YG1 and TP0YG0 experiments (Fig. 6a). The upper-atmosphere meridional thermal gradient between midlatitude Asia and the southern Indian Ocean is intensified in the upper atmosphere (not shown), in agreement with low-level circulation change. With the YG topography, the high pressure center over the TP is significantly weakened (Fig. 6b), which facilitates and contributes to the suppression of the Indian summer monsoon.
The effect of the YG topography on the Indian summer monsoon primarily originates from its important location on the route of monsoon propagation. From the change in surface wind vectors, the existence of the YG Plateau hinders the monsoon winds and forces them to turn southward when the Indian monsoon goes through (Fig. 7). The southward turning of the winds produces anticyclone anomalies to the southwest of the YG Plateau and over the Bay of Bengal, consistent with the response detected in the 850-hPa wind vectors (Fig. 4d). The mechanically forced northerly winds along the western slope of the YG Plateau extend from the surface through the midtroposphere and become southerly wind anomalies in the upper troposphere (Fig. 8a). The anticyclonic circulation corresponds to anomalous downward motion (Fig. 8b), which extends through nearly the whole troposphere and suppresses the convection activities over the Bay of Bengal. Although the thermal forcing of the YG Plateau on the Indian monsoon cannot be totally excluded since the sensible heating change over the YG Plateau is considered in the experiments, this mechanical effect of the YG Plateau in blocking monsoon circulation seems more direct and important.
c. Monsoon onset (May)
During the onset of the Indian monsoon, the responses of monsoon circulation and precipitation to the TP and YG Plateau are similar to those during the mature period. For the entire TP, obvious southwesterly or westerly wind anomalies are simulated over the whole Indian monsoon region in May, which are more significant than those in JJA, and the corresponding precipitation increase is greater over the Indian subcontinent and Bay of Bengal (Fig. 9a). These spatial distributions indicate that the response of the Indian monsoon to the TP topography is more sensitive during the onset than the mature period, consistent with the results in Park et al. (2012). The forced southward turning of the air current by the YG Plateau also occurs once the summer monsoon circulation establishes during May (Fig. 9b), although this turning is not so significant as that in JJA. Over the Bay of Bengal, an anticyclone anomaly is presented and the precipitation is suppressed (Fig. 9b). However, the response of monsoon circulation and rainfall is not statistically significant over most of the monsoon areas.
In the upper troposphere, an anomalous warming center, which may originate from the sensible heating of the TP (Yanai et al. 1992) and the delivery of water vapor from the oceans (Tamura et al. 2010), emerges to the southwest of the TP in May (Figs. 10a,c). Because of its importance in the Indian monsoon establishment (Wu and Zhang 1998), the possible effect of the YG topography on this anomalous heating was also evaluated. In response to the TP, a warming anomaly is apparent across the southern TP in the 300-hPa temperature field, which spreads to most of arid central Asia and southeastern China (Fig. 10a). This warming significantly intensifies the mean climate warm center southwest of the TP, making the warm region extend from the lower to the upper troposphere. (Fig. 10c). This warming thus strengthens the meridional thermal gradient in the upper atmosphere and becomes one of the primary reasons for the sensitive response of the Indian monsoon during the onset (Yanai et al. 1992). In contrast, to the northeast of the TP the temperature is decreased by the TP (Fig. 10a). This cooling by the TP may result from the blocking of the westerly winds by the Tibetan and Mongolian Plateaus. Over the downwind region, the forced northerly anomalies bring cold air from high latitudes, which affects the trough over Japan and intensifies the cold advection in the upper levels (Shi et al. 2015).
Between TP1YG1 and TP1YG0, a YG-induced cooling is simulated over the western and central TP, and this cooling is apparent in the meridional belt spanning from the midlatitude continent to the southern oceans, which remarkably weakens the anomalous heating center (Fig. 10b). In addition to this widespread cooling, the TP1GY − TP1YG0 difference creates statistically significant warming over southeastern China and the Indochina Peninsula. Vertically, the decreased warming center is significant at all levels of the mid- to high troposphere, while in contrast the surface warming is strengthened (Fig. 10d). The simulated weakened warming center in May is consistent with less precipitation over Indian monsoon areas in the TP1YG1 experiment compared with TP1YG0 (Fig. 9b).
A heat budget analysis based on Yanai and Li (1994) was conducted to quantify the contributions of different heating sources in the weaker anomalous warming center in the upper troposphere (Fig. 11). A strong adiabatic warming anomaly occurs to the southwest of the TP (Fig. 11a), corresponding to the intensified downdraft over this region (not shown). This adiabatic warming is accompanied by diabatic warming over this area with a comparable amplitude (Fig. 11b). In contrast, a diabatic cooling occurs over the southern TP, which is due to decreased latent heat from the Bay of Bengal linked to the weakened upward air motion. Besides the diabatic cooling, the decreasing upper-level temperature is also caused by horizontal advection, which presents a strong negative anomaly from the western TP to its west (Fig. 11c). The stronger northerly winds (Fig. 9b) bring cold air from the high-latitude region and significantly decrease the temperature in the upper troposphere around the TP in the TP1YG1 experiment, compared with TP1YG0. Thus, the YG-forced circulation not only directly affects the development of the Indian monsoon circulation but also modulates the atmospheric thermal structure, providing an additional influence on the monsoon.
a. Link between the YG topography and Indian monsoon
The physical link between the YG topography and Indian summer monsoon revealed in this study is summarized and discussed. As the results show, the Indian monsoon is significantly influenced by the increase in elevation over the YG region during both its onset and mature period. During the onset, the terrain-forced redirection of wind flow occurs once the Indian monsoon westerly establishes. The decreased upward air motion over the Bay of Bengal reduces the delivery of water vapor and latent heat from the ocean to the TP. Together with the cold advection, the decreased latent heat weakens the anomalous warming center around the TP in the upper troposphere. This mechanism referring to latent heat was previously emphasized (Tamura and Koike 2010; Tamura et al. 2010), and our results support its role in the YG-induced changes in the upper-tropospheric warming. In Boos et al. (2015), they argued that most of the monsoon precipitation arrives from low pressure systems over the Bay of Bengal, which propagate in the northwest direction over the mainland. These low pressure systems over the Bay of Bengal are of importance in the Indian monsoon development, and YG topography modulates them significantly.
As the monsoon develops, the already-weakened upper-tropospheric warming around the TP, and the weaker high pressure cell over the TP, directly contributes to the weaker mature Indian monsoon. In the low-level atmosphere, the mature Indian monsoon continues to be blocked by the YG topography, which produces anticyclonic circulations over the Bay of Bengal and eastern India by the conservation of potential vorticity. The weakened updrafts over these regions, as well as the weaker monsoon westerly wind, are both responsible for the decreased precipitation over the Indian monsoon areas, primarily by suppressing the local convection activities and the remote transport of water vapor, respectively. Certainly, the importance of the sensible heat from surface TP (Yanai et al. 1992; Wu and Zhang 1998) on the Indian monsoon onset cannot be totally excluded. But based on this study, it is speculated that the YG impact is mainly realized by the mechanically forced changes in the latent heat transport.
b. Influence of small-scale topography on Asian climate
Compared to the main TP, the climatic effect of surrounding small-scale or marginal topography is seldom focused upon in theoretical studies (e.g., Kutzbach et al. 1989; Manabe and Broccoli 1990; Liu and Yin 2002), meaning changes in the evolution of the Asian climate are largely ascribed to the uplift of the main TP (e.g., An et al. 2001; Molnar et al. 2010; Miao et al. 2012). To better understand mountain-induced Asian climate change, it is necessary to evaluate in detail the relative contributions of small-scale topography. In fact, previous studies have already indicated an influence of small-scale topography on local climate. The mechanical blocking of the narrow Himalaya Mountains can isolate the potential thermal influence of the main TP on the Indian monsoon circulation (Boos and Kuang 2010). Orographic rainbands over narrow mountains in India and the Indochina Peninsula form the cores of convection over the Bay of Bengal and South China Sea (Xie et al. 2006). A regional climate model based on boundary conditions for the Late Miocene, which reduces the elevation of the Pamir Plateau and Tian Shan, indicates that the local precipitation over these mountains was likely remarkably decreased compared to the present day (Liu et al. 2015). The Tian Shan are further considered to result in climatological regional differentiation of arid extratropical Asia (Baldwin and Vecchi 2016).
Beyond local effects, some small-scale topography may exert far-reaching effects on the climate. For example, the Mongolian Plateau blocks the westerly flow, induces strong cold advection in the downwind region, and has a comparable effect to that of the TP on the westerly jet over the North Pacific and the East Asian winter monsoon area (Shi et al. 2015; Sha et al. 2015). Elsewhere, the existence of the Iranian Plateau generates a cyclonic circulation that encircles it and results in increased precipitation over Pakistan and northern India (Wu et al. 2012). In this study, the remote effect of the YG Plateau was also found to be important, with the magnitude of the YG-induced precipitation decrease over the Indian monsoon area reaching about 1/3 of the magnitude of the total increase induced by the main TP. This indicates that the climatic effect of the TP on the Indian summer monsoon relative to that of other topography has been to some extent overestimated in prior literature. Although our results depend on a specific model, they nonetheless highlight that the small-scale topography around the main TP, like the TP itself, can also substantially influence climate systems, both locally and remotely.
c. Mechanical vs thermal effects of the Tibetan Plateau
The mechanical and thermal effects of the main TP have long been accepted by the scientific community; however, it remains controversial as to which makes the relatively larger climatic contribution. The traditional view is that the mechanical effect of the main TP is more significant for the westerly flow and interior Asian aridity (Manabe and Broccoli 1990; Broccoli and Manabe 1992), while the thermal effect contributes more to the Asian monsoon (Yanai et al. 1992; Wu et al. 2012). Such a viewpoint was challenged by a recent modeling study (Boos and Kuang 2010) that emphasized the thermal effect of the TP on the Indian monsoon is not as significant as expected. The study found that the Indian summer monsoon circulation does not vary with the emergence of the TP if the Himalaya Mountains already exist to its south, and thus the Himalayas could act to isolate the potential elevated heating of the TP. The importance of the mechanical effect of the TP has also been highlighted in terms of the seasonal transition of the Indian monsoon (Park et al. 2012). Another viewpoint is that the thermal forcing of the TP still controls the eastern part of the South Asian monsoon and the East Asian monsoon, and the work of Boos and Kuang (2010) may underestimate the lateral heating of the south slope of the Himalayas (Wu et al. 2012). Regional climate model experiments, however, indicate that the elevated surface TP heating can induce a low-level cyclonic anomaly that reduces the Indian monsoon by suppressing the lower-tropospheric monsoon vorticity (Tang et al. 2013).
Our study did not focus on the main TP but instead discussed the climatic effect of marginal mountains. It is difficult for marginal mountains to exert a thermal effect comparable to that of the main TP because of their limited heights and sizes. However, marginal mountains located in sensitive locations for certain climate systems can significantly hinder the circulation or winds. The remarkable impact of the YG topography on the Indian monsoon in this study mainly originates from its location. The YG Plateau is located between the Indian and East Asian monsoon regions, and when the northern branch of the Indian monsoon southwesterly flows toward the YG terrain, it is impeded and forced to become northerly. Subsequently, anticyclonic circulations are produced, which leads to weakened Indian monsoon circulation and precipitation over the Bay of Bengal, Indian subcontinent, and Arabian Sea. Such a blocking effect from marginal mountains has also been reported for the Mongolian Plateau, with respect to the East Asian winter climate, which is located along the route of the westerly jet (Shi et al. 2015; Sha et al. 2015). Thus, the mechanical effect of marginal mountains may be as important as that of the main TP. That said, the thermal effect of small mountains cannot be completely ignored. For example, the thermal forcing of the Iranian Plateau on Asian precipitation change, although perhaps not as strong as that of the main TP, has been previously emphasized (Wu et al. 2012).
d. Possible influence of Yunnan–Guizhou Plateau uplift on the Indian monsoon’s evolution
The evolution of the Indian monsoon and its relationship with the TP–Himalaya uplift is a highly debated topic. On the basis of modeling results (Hahn and Manabe 1975; Kutzbach et al. 1989; An et al. 2001), the uplift of the TP–Himalaya complex has long been employed to explain reconstructed Indian monsoon intensification on the tectonic time scale (An et al. 2001; Clark et al. 2005). However, this link remains hypothetical and simplified since the intensification of the whole TP topography on the Indian monsoon in numerical experiments does not mean that all the subterrains ought to strengthen the monsoon. As a result of the eastward expansion of the main TP, the southeastern TP (including the YG Plateau) rose thousands of meters 13–9 Ma ago as suggested by the rapid incision of deep river valleys in this area (Clark et al. 2005, 2006; Tian et al. 2015). The cooling histories of exposed rocks show pulses of rapid exhumation at 30–25 and 15–10 Ma ago, supporting the occurrence of mountain building during these episodes (Wang et al. 2012). In brief, the Cenozoic topography of the YG Plateau experienced significant uplift and attained its present height after the Late Miocene (Clark et al. 2006; Hoke et al. 2014) and following its initiation as early as the Oligocene (Clift and Sun 2006; Liu-Zeng et al. 2008).
If the suggested timing of the YG uplift is correct, the height change from TP1YG0 to TP1YG1 in this study is most likely to be associated with the uplift stages during the Late Miocene and later. Marine records from the Arabian Sea, although they may not be sound (Rodriguez et al. 2014), indicate a remarkable strengthening of southwest monsoon winds between 10 and 8 Ma ago (Kroon et al. 1991; Gupta et al. 2015) and a major transition of continental vegetation ~7 Ma ago has also been detected (Quade et al. 1989). In previous studies (e.g., Clark et al. 2005), the Late Miocene intensification of the Indian monsoon has even been correlated with the rapid increase in elevation of the southeastern TP. But the appearance of the YG topography in the model significantly weakens the southwesterly winds over the Arabian Sea, which completely opposes the traditional viewpoint that the uplift of the southeastern TP can intensify the Indian monsoon. Hence, based on our results, the uplift of the YG Plateau or the southeastern TP ought to be linked to the weakening stages of Indian monsoon.
Based on numerical experiments, the effect of the YG Plateau, at the southeastern margin of the TP, on Indian summer monsoon climate was evaluated in this study. The analysis showed that the YG topography is of great importance in the formation of Indian summer monsoon during both the onset and the mature period. From the modeling estimates, the YG-induced summer precipitation reduction over the Indian monsoon region is approximately 38% of the TP-induced increase. This remarkable effect of the YG Plateau is primarily facilitated by its sensitive location for the Indian monsoon. During the monsoon onset when the monsoonal westerly winds form, the YG Plateau redirects the wind flow and an anticyclone emerges over the Bay of Bengal. This anticyclone reduces the latent heat transport and the anomalous upper-troposphere warming around the TP, which weakens the development of the monsoon circulation. When the Indian monsoon matures in JJA, the YG Plateau continues to block the southwesterly Indian monsoon winds and forces them to turn southward. The anticyclonic winds to the southwest of the YG Plateau and the Bay of Bengal result in a weakening of the westerly winds over southern areas and in turn monsoon precipitation. Although the YG Plateau is much smaller than the main TP, it can exert a comparable but negative influence on the Indian monsoon. Hence, not all of the subterrain of the TP topography exerts similar impacts on the climate system, and the traditional view regarding the effect of the main TP, especially its thermal effect, might be partly overestimated relative to other topography. The evaluation of this study has focused on the effect of change in mountain elevation and other factors (e.g., vegetation and sea surface temperature) not explored in this work. Further assessments of this marginal topography using fully coupled Earth system models are required to deepen our understanding of the climatic effect of the mountains.
The authors thank two anonymous reviewers and Dr. Naftali Cohen for their insightful comments and Dr. Hong Chang for the helpful discussions on the timing of Tibetan uplift. This work was jointly supported by the Strategic Priority Research Program of the Chinese Academy of Sciences (XDB03020601), the National Natural Science Foundation of China (41290255 and 41572160), and the Chinese Academy of Sciences (QYZDY-SSW-DQC001 and ZDBS-SSW-DQC001).