Influence of Tibetan Plateau on the North American summer monsoon precipitation

It has been well known that the uplift of the Tibetan Plateau (TP) can significantly enhance the Asian monsoon. Here, by comparing the sensitivity experiments with and without the TP, we find that the TP uplift can also increase the precipitation of the North American Summer Monsoon (NASM), with atmosphere teleconnection accounting for 6% and oceanic dynamical process accounting for another 6%. Physically, the TP uplift generates a stationary Rossby wave train traveling from the Asian continent to the North Atlantic region, resulting in an high-pressure anomaly over the tropical-subtropical North Atlantic. This high pressure system enhances the low-level easterly winds, forcing an enhanced upward motion over the North American monsoon (NAM) region and then an increase in summer precipitation there. In addition, the TP uplift enhances the Atlantic meridional overturning circulation, which reduces the meridional temperature gradient and leads to a northward shift of Hadley Cell over eastern Pacific-Atlantic section. The latter shifts the convection center northward to 10°N and further increases the NASM precipitation. The enhanced NASM precipitation can also be understood by the northward shift of Intertropical Convergence Zone. Our study implies that the changes of NAM climate can be affected by not only local process but also remote forcing, including those from Asian highland region.


Introduction
The Tibetan Plateau (TP), with an average elevation of more than 4000 m above sea level and a total area of 2.3 million square kilometers, is one of the most prominent features on Earth. The uplift of the TP has been known to generate a great impact on regional climate change, especially on Asian monsoon climate (Harrison et al. 1992;Molnar et al. 2010). Until now, the TP uplift on Asian monsoon precipitation has been studied both by observations and model simulations. In observational studies, the documentation of planktonic foraminifer over the Arabian Sea indicates enhanced southwesterly winds and a strengthening of the Indian summer monsoon (Kroon et al. 1991;Prell et al. 1992). The aeolian "Red Clay" sediments over the Chinese Loess plateau suggest an onset of aeolian dust accumulation . The oxygen isotope composition of soil carbonates indicates changes in vegetation from forests to grasses in Pakistan (Quade et al. 1989), as well as a change from mixed needle-leaf or broad-leaf forests to grassland vegetation along the northeastern margin of the TP (Cerling et al. 1997). These widely distributed observations can be interpreted as signalling an environmental response to the TP uplift about 9-8 Myr ago . In model simulations, several studies investigate the Asian summer monsoon behavior under the TP elevation forcing (Boos and Kuang 2010;Park et al. 2012;Wu et al. 2012). Major results are that the uplift of the TP significantly increases the Asian monsoon intensity (Prell and Kutzbach 1992) and variability (Fallah et al. 2016). Considering different stages of TP uplift, the East Asian summer monsoon system similar to that of the present initially exists when the TP uplift 60% of its modern height and is gradually intensified with the continued plateau uplift (Jiang et al. 2008;Liu and Yin 2002;An et al. 2001). In addition, the uplift of different parts of the TP on distinct Asian monsoon climate has also been verified by recent numerical experiments (Zhang et al. 2010;Boos and Kuang 2010;Wu et al. 2012;Tang et al. 2013;Chen and Bordoni 2014).
However, few studies have focused on the impact of the TP topography on North American monsoon (NAM) climate. The NAM is one of the most complex and interesting atmospheric circulation features over the North America. Dominated by perennially dry conditions, the NAM contributes approximately 40 and 70% of the annual total precipitation for the southwest Unite States and northwest Mexico, respectively (Higgins et al. 1997;Adams and Comrie 1997). Paleo-climate research using fossil flowers to reconstruct Miocene climate over south Mexico shows that the environmental conditions of the Chiapas is warmer and drier during Miocene when the TP is not uplifted (Hernández et al. 2020). That is to say, the TP uplift is responsible for the wetter conditions over south Mexico in the present. However, the mechanisms are still unclear. The role of the TP on the NAM precipitation deserves an in-depth investigation, which is our focus in this study.
The drier conditions over NAM region during Miocene when the TP is not uplifted resembles the weakened NAM precipitation under global warming. The redistribution of NAM precipitation to global warming is usually conjectured to arise from two factors. One factor is the convection barrier related to the surface evapotranspiration and the tropospheric stability induced by greenhouse gas while the other is the moisture convergence related to monsoon circulation (Cook et al. 2013;Pascale et al. 2017). All these factors are related to atmospheric circulation and surface temperature. Previous works that study the NAM precipitation under global warming usually highlight the contributions from local regions, i.e., the Pacific and North Atlantic. For example, the sea surface temperature (SST) cooling in the North Atlantic can substantially alter the North Atlantic subtropical high, which may ultimately influence the NAM (Kushnir et al. 2010;Parsons et al. 2014;Wang et al. 2014). The NAM is also closely linked to tropical Pacific SST (Castro et al. 2001). The positive temperature anomaly over eastern-central Pacific can displace the North Pacific intertropical convergence zone and the South Pacific convergence zone equatorward, thereby directly reducing the NAM rainfall (Webster et al. 1998;Wang et al. 2012). However, the impacts from remote area, i.e., the Asian highland regions on the NAM climate have never been studied before. Our previous works suggest that the TP uplift may trigger a La Niña-like SST response over the tropical Pacific Ocean and an establishment of the Atlantic meridional overturning circulation (AMOC) Yang and Wen 2020). Building on these prior works, the TP uplift should be important for NAM climate change. How would the North American summer monsoon (NASM) precipitation response under the TP uplift? What is the role of the atmosphere and ocean? These questions have rarely been addressed in previous works. It is worth noting that the TP is reported to have experienced a faster warming than other regions in the Northern Hemisphere during past several decades (Duan et al. 2006) and will continue to warm more seriously in the future (Hu et al. 2015;Zhang et al. 2015). The TP uplift in our study includes the TP thermal forcing. The teleconnection between TP uplift and NAM precipitation in this study to some extent can be used to infer the future NAM change.
This paper is arranged as follows. An introduction to the model and experiments, as well as the methods is given in Sect. 2. Changes in NASM precipitation are illustrated in Sect. 3. Mechanisms for NASM precipitation change are analyzed in Sect. 4. Summary and discussion are given in Sect. 5.

Model and experiments
The Community Earth System Model (CESM1.0) is applied in this study. This model has been widely used to study the Earth's past, present, and future climate states (http:// www2. cesm. ucar. edu/). CESM was developed by the U.S. National Centre for Atmospheric Research (NCAR) and is composed of an atmosphere model (Community Atmosphere Model; CAM5), ocean model (Parallel Ocean Program; POP2), land surface model (Community Land Model; CLM4), sea ice model (Community Ice Code; CICE4) and one coupler . The model grid employed in this study is T31_gx3v7. The CAM5 has 26 vertical levels, with the finite-volume nominal 3.75° × 3.75° grid. The CLM4 has the same horizontal resolution as the CAM5. The POP2 has 60 vertical levels and a uniform 3.6° spacing in the zonal direction. In the meridional direction, the grid is nonuniformly spaced: It is 0.6° near the equator, gradually increasing to the maximum 3.4° at 35°N/S and then decreasing poleward. The CICE4 has the same horizontal grid as the POP2. No flux adjustments are used in CESM1.0.
To fully assess the TP uplift on NASM precipitation, we separate the atmospheric and oceanic dynamical process by conducting two groups of simulations. The first set of simulations are performed using fully coupled (CPL) runs, including a 2400-year simulation with no modifications made to topography which we will refer to as "CTRL" and a 400-year no-mountain run called "NoTP". In CTRL, the model geometry, topography, and continents are realistic (Fig. 1a). The CTRL run reaches equilibrium state around 2000 years (Yang et al. 2015). The NoTP simulation starts from the year 2001 of CTRL, and is integrated for 400 years with the topography around the TP set to 50 m above the sea level (Fig. 1b). Except the topography elevation, all other boundary conditions, such as vegetation and albedo remain unmodified and they are free adjusted. The second set of simulations considers a slab ocean model (SOM), in which the ocean dynamical process has been shut-down and replaced by a mixed layer from the climatology of CTRL simulation. Two experiments in this group are named "CTRL_SOM" and "NoTP_SOM", in which the topography is set as that of "CTRL" and "NoTP", respectively. "CTRL_SOM"is integrated for 400 years and "NoTP_SOM" starts from the year 201 of "CTRL_SOM" and is integrated for 200 years. The equilibrium stages are deduced by using data from the last 50 years of integrations in SOM runs and last 100 years in CPL runs (outlined by light blue shadings in Fig. 2b, c). The difference between the "CTRL" and "NoTP" is for the sum of atmospheric and oceanic dynamical processes, and the difference between the "CTRL_SOM" and "NoTP_SOM" can be treated as atmospheric effect.

North American monsoon region
According to Wang and Ding (2008), the NAM domain is delineated by the region in which the local summer precipitation minus local winter precipitation exceeding 2 mm/day and the local summer precipitation exceeds 55% of annual rainfall. Based on this definition, the global monsoon regions are clearly shown in Fig. 2a by using the GPCP data (outlined by purple contours) and our CTRL simulation (outlined by black contours). It is clearly shown that our CTRL simulation can well capture the spatial pattern of global monsoon domains, which is consistent with the findings of Liu et al (2016). Note that the NAM domain obtained in this study is primarily based on precipitation contrast in the solstice seasons and is larger than that traditionally recognized by many scientists working on the NAM (Wang et al. 2021). The NASM precipitation can be obtained by the weightedarea average local summer precipitation over the NAM area.

Moisture budget
To reveal the mechanisms that govern changes in NASM precipitation related to the TP uplift, we analyse changes in the moisture budget according to Chou et al. (2009): Here, P is precipitation, E is surface evaporation, is vertical velocity, q is specific humidity, p is pressure, V is horizontal wind vector, R is residual. Overbars represent monthly means in CTRL, and primes denote the difference between CTRL and NoTP. Based on this equation, the monsoon precipitation change can be decomposed into evaporation change ( E ′ ), the thermodynamic and dynamic changes of vertical moisture advection ( − − q � p and − � q p ), and thermodynamic and dynamic changes of horizontal moisture advection ( −⟨v • ∇q � ⟩ and −⟨V � • ∇q⟩ ) and the residual R. (1)

Three-pattern decomposition of global atmospheric circulation
The monsoon circulation is the component of large-scale atmospheric motion. The large-scale atmospheric motion to first order is consist of Rossby wave at mid-high latitudes (Rossby 1939), Hadley and Walker circulation at low latitudes (Trenberth and Solomon 1994;Julian and Chervin 1978). However, these components interact with each other and are hard to be distinguished in the real world. Fortunately, a new method named the threepattern decomposition of global atmospheric circulation (3P-DGAC) has been introduced by Hu et al. (2017), by which the global atmospheric circulation �� ⃗ in which, are the wind vector components of horizonal circulation, meridional circulation, and zonal circulation, respectively. R , H , and W are the stream functions which can be obtained from following equation: 2 a Monsoon domains in GPCP data (purple contours) and in CTRL simulation (black contours). According to Wang and Ding (2008), the monsoon domains are defined as the region where the local summer precipitation minus local winter precipitation exceeding 2 mm/ day and the local summer precipitation exceeds 55% of annual rainfall. b Time evolution of changes in North American summer monsoon (NASM) precipitation (mm/day) in SOM runs. Blue curve is for precipitation change while green curve is for its percentage change. c Is the same as b but for fully coupled (CPL) simulations

Changes in NASM precipitation
The time evolution of changes in NASM precipitation under the TP uplift are first shown in Fig. 2b, c. The TP uplift immediately triggers NASM precipitation increase by 0.5 mm/day (6%). This can be seen from the SOM runs ( Fig. 2b), in which the precipitation change occurs rapidly right after the TP uplift and remains stable during the following integrations. The fast response of NASM precipitation to the TP uplift can be also seen from CPL runs during the first several decades when the oceanic dynamical process is not fully responded (Fig. 2c). The oceanic dynamical feedbacks further double the NASM precipitation change and lead to precipitation increase to 0.9 mm/day (12%) during the equilibrium stage (Fig. 2c). The spatial distributions of monsoon precipitation change in SOM runs and CPL runs are shown in Fig. 3a, b. The NAM region is characterized by a large extent of significant summer precipitation increase via atmospheric processes (Fig. 3a) and is further amplified by oceanic dynamical feedbacks (Fig. 3b). It is interesting to note that the precipitation is consistently increased over the NAM region by atmospheric processes (Fig. 3a) while the precipitation change induced by oceanic dynamical processes exhibits an out-of-phase pattern with large precipitation increase occurred over central to north NAM region and decrease over south NAM region (Fig. 3b). The changes in surface humidity are also checked here ( Fig. 3c, d). The surface air humidity is slightly increased over NAM region in SOM runs ( Fig. 3c) while that shows an increase over central to north NAM region and decrease over south NAM region in CPL runs (Fig. 3d). Our finds indicate that the TP Equilibrium changes in precipitation and humidity due to the TP uplift: a precipitation for SOM runs and b precipitation for CPL runs. Units: mm/day. c Humidity (shading) for SOM runs and (d) humidity (shading) for CPL runs. In c and d, the purple contours denote the climatological humidity from CTRL_SOM and CTRL, respectively. Units: g/kg. All these values are extracted from boreal summer. The stippling areas indicate the difference exceeding 95% confidence level, determined by a two-tailed Student's t test. The black contours denote the North American monsoon (NAM) region uplift can increase the NASM precipitation and moisten the most NAM regions.

Changes in mean climate
To better understand the response of NASM precipitation to the TP uplift, the mean climate changes are also examined here. The TP uplift immediately induces anticyclonic geopotential height anomaly to the north of the TP area and extending from the tropical-subtropical North Atlantic to NAM region, and cyclonic to the south of the TP area, extending to subpolar Pacific and subpolar Atlantic (Fig. 4b).
The atmospheric circulation changes are roughly barotropic since the wave pattern at 500 hPa shows resemblance to that at 850 hPa . The adjustment of planetary wave pattern is very fast and is also confirmed by our CAM5 model simulation (Figure not shown). More importantly, the atmospheric responses do not change much from SOM runs to CPL runs ( Fig. 4b vs d), indicating that the oceanic dynamical processes in the hundreds of years later do not have considerable feedback to the atmospheric circulation. Detailed atmospheric circulation changes are also discussed in , which states that the teleconnection patterns in Fig. 4b, d agree well with those in previous studies (Zhao et al. 2007(Zhao et al. , 2012 and can be well understood by the classic planetary wave theory in a linear quasigeostrophic system (Hoskins and Karoly 1981).
The high pressure anomaly extending from the tropical-subtropical North Atlantic to NAM region enhances the easterlies over the central American continent, which promotes the evaporation and thus the surface latent heat loss from ocean to atmosphere (Figure not shown), resulting in slightly colder surface air temperature (SAT) around the NAM region in SOM runs (Fig. 4c). In CPL runs, the strengthened trade winds over the eastern Pacific cause poleward Ekman transport and The subsurface cold water then upwells to compensate the surface water loss, which results in eastern Pacific (180°W-80°W, 10°S-10°N) cooling (Fig. 4e). In addition, the uplift of the TP can also lead to the establishment of the AMOC (Fig. 4a) by inducing more water vapor transport from the North Atlantic to Pacific Ocean . The establishment of the AMOC warms the North Atlantic, western Eurasian continent, as well as the North American continent by generating pronounced heat transport from Southern Hemisphere (SH) to Northern Hemisphere (NH) (Fig. 4e).

Moisture budget
To understand the processes that drive the NASM precipitation change, the projected changes in precipitation and the first five terms on the right-hand side of Eq.
(1) are shown in Figs. 5 and 6 for the SOM and CPL runs, respectively. Note that since we use 50 years of data in SOM runs and 100 years of data in CPL runs, all variables with color shadings are exceeding 95% confidence level determined by a two-tailed Student's t test. For simplifying the figure, we do not add stippling. It is obvious that the dynamic change of vertical moisture advection is the largest contributor to the increased NASM precipitation both in SOM runs and CPL runs, with the magnitude in CPL runs much bigger than that in SOM runs (Figs. 5d, 6d). The enhanced vertical moisture  In (a), the purple contours denote the climatological precipitation in CTRL_SOM simulation. In (e), the vectors (units: m/s) denote the climatological winds at 850 hPa in CTRL_SOM simulation. In (f), the vectors (units: m/s) denote the wind difference between CTRL_SOM and NoTP_SOM advection is due to the increased upward motion, a point to be returned later. The evaporation term is hardly changed because the TP-uplift induced surface temperature change over this region is small (Fig. 4c, e). The weak temperature change also contributes to the hardly changed thermodynamic component of vertical moisture advection (Figs. 5c,  6c). The thermodynamic and dynamic changes in horizontal moisture advection are both uncertain and small over the NAM region (Figs. 5e, f, 6e, f), consistent with previous studies that the horizontal moisture advection is usually less important than vertical moisture advection to precipitation (Chou et al. 2009;Zhang et al. 2017). The contribution of these terms to NASM precipitation change are summarized in Fig. 7a. The dynamic change of vertical moisture advection term nearly contributes 100% precipitation increase in SOM runs and CPL runs, respectively, which is undoubtedly the dominant term for NASM precipitation change. Based on Eqs.
(2)-(7), the dynamic change of vertical moisture advection term can be further broken into vertical moisture advection due to meridional circulation and zonal circulation change (Fig. 7b). The increased dynamic change of vertical moisture advection term ( −⟨ � q⟩ ) largely comes from meridional circulation change ( −⟨ H � q⟩ ). The zonal circulation change appears to suppress the NASM precipitation increase ( −⟨ W � q⟩).

Mean vertical velocity and 3P-DGAC
To understand the atmospheric circulation change, the mean atmospheric vertical velocity and precipitation during boreal summer are first shown in Fig. 8. In the annual mean state of observations, the large-scale upward motion is expected in the tropics, with centers near equatorial Africa, the Indian Ocean, the western Pacific, the eastern Pacific, the South America, and the Atlantic Ocean (Fig. 8a) Cheng et al. 2020). During boreal summer, the upward motion gets stronger in the NH and weaker in the SH (shading in Fig. 8b compared with that in Fig. 8a), indicating the large-scale northward shift of Intertropical Convergence Zone (ITCZ) (contours in Fig. 8b compared with that in  Fig. 8a). These features are also captured in our CTRL and CTRL_SOM simulation (Fig. 8c, d). Strong upward motion in the tropics is associated with deep convection and corresponds to above-normal precipitation in these regions while strong descending motion in the subtropics suppresses the convection and corresponds to the aridity (contours in Fig. 8a-d) ). Comparing CTRL_SOM and NoTP_SOM, the upward motion is strengthened over the NAM region (shading in Fig. 8h), resulting in more precipitation over there (contours in Fig. 8h). Comparing CTRL and NoTP, the TP uplift leads to a northward shift of the maximum upward motion over the NAM region, results in enhanced upward motion and thus the increased precipitation in the northern area of NAM region, and weakened upward motion and thus the decreased precipitation in the southern area of NAM region (shading and contours in Fig. 8g). Previous works suggest that the tropical overturning circulation consists of a couple of orthogonal overturning circulation, that is, meridional and zonal circulations (Hu et al. 2018). Based on 3P-DGAC method, the vertical wind can be decomposed into two parts, the vertical winds related to meridional circulation and zonal circulation. i.e., = H + W , where , H and W represent the total vertical wind, the vertical wind of meridional circulation and that of zonal circulation, respectively (Hu et al. 2018). The vertical velocity and its meridional and zonal components are shown in Fig. 9. In the real world, the maximum upward motion of meridional circulation are located over equatorial Africa, the Indian Ocean, the western Pacific, the eastern Pacific, and the tropical North Atlantic Ocean, indicating the ascending branch of regional Hadley circulations (Fig. 9b, f). All these regional Hadley circulations are characterized by two circulations with rising branch in the tropical regions and sinking branch in the subtropics of both hemispheres (Fig. 9b, f) (Cheng et al. 2020). For zonal circulation, there are three main centers over the Indian Ocean, the western Pacific Ocean and the western Atlantic Ocean, representing the three rising branches of Walker circulation (Fig. 9c, g). The sinking branches of Walker circulation are located in the western Indian Ocean, the eastern Pacific Ocean and the eastern Atlantic Ocean. The residual term is very small and can be negligible in latter discussion (Fig. 9d, h). The decomposition results of vertical velocity are similar to previous works (Hu et al. 2017;Cheng et al. 2020 (Wen et al. 2018). However, it is incorrect to use regional total vertical velocity, i.e., the vertical wind between 80 • W and 10 • W to represent the regional Hadley circulation since the W is not zero (Fig. 9c, g). In addition, the traditional definition of the Walker circulation is restricted to the tropical region and ⟨ ⟩ 5 • N 5 • S is often used to calculate the Walker circulation. In this case, the ⟨ H ⟩ 5 • N 5 • S ≠ 0 (Fig. 7b, f), which means that the contribution of the meridional circulation is included in the vertical velocity of the Walker circulation. So, when analyzing the meridional circulation (zonal circulation), the vertical wind of meridional component (zonal component) should be used. By using this method, we can analyze the regional meridional circulation and zonal circulation.

Mechanisms for NASM precipitation change
The changes in vertical velocity and its meridional component and zonal component during boreal summer are shown in Fig. 10 to illustrate the mechanisms that govern the NASM precipitation response. As discussed in Sect. 4.3, the NAM region is characterized by an enhanced ascending motion (shading in Fig. 8g, h), which is dominated by the meridional circulation change both in SOM runs and CPL runs (Fig. 10b, e). However, the mechanisms may be different. In SOM runs, the TP uplift generates enhanced upward motion of meridional component over the NAM region and its surrounding regions, which dominates the total vertical  Fig. 7 Bar char for the mean changes in a precipitation and its contributors and b dynamic component of vertical moisture advection and its contributors. Orange bar is for SOM runs and blue bar is for CPL runs. In a, The meaning of labels at x-axis is the same as that in Fig. 4. In b, −< � W q > denotes the dynamic term due to zonal circulation change and −< � H q > denotes the dynamic term due to meridional circulation change. Units: mm/day. All values are from boreal summer wind change (Fig. 10b) and results in increased NASM precipitation (Fig. 7b). Compared to SOM runs, the enhanced upward motion in CPL runs is located further north with abnormal descending motion at the south tip of it, indicating that the meridional circulation shifts northward (Fig. 10e). The changes in upward motion of zonal component over the NAM region is very small and uncertain compared with that of meridional component.

Mechanisms in SOM runs: the atmospheric teleconnection response
The change in meridional component of vertical velocity over the NAM region refers to the regional Hadley circulation (HC) response. Here, we define the regional HC over eastern Pacific-Atlantic sector as spanning 120°W-40°W (red box in Fig. 10b, e) and plot it in Fig. 11. In SOM runs, the summer HC gets strengthened in both hemispheres under the TP uplift with the magnitude in NH much stronger than that in SH (Fig. 11a). The strengthened HC in NH results in enhanced convection over the NAM region and thus the increased summer precipitation (Fig. 7b). The strengthened regional HC in SOM runs is related to the adjustment of planetary waves (Fig. 4b). The abnormal positive geopotential height extending from the tropical-subtropical North Atlantic to the NAM region is accompanied by the northeasterly wind at the south tip of it, which results in stronger regional HC over NAM region by enhancing the horizontal momentum flux from the surface into the atmosphere (Cook et al. 2003). Actually, the negative-positive geopotential

Mechanisms in fully coupled runs: the indirect impact from the altered AMOC
In CPL runs, the TP uplift induces a northward shift of HC in NH, resulting in increased upward motion over the NAM region (Fig. 11c, d). The northward shift of HC in CPL runs can be understood by the meridional temperature gradient change. The uplift of the TP induces the establishment of AMOC (Fig. 4a), which brings substantial heat northward to warm the NH, especially over the North Atlantic and North American continent. The profound warming over mid-high latitudes reduces the meridional temperature gradient, which leads to a northward shift of HC (Bush and Philander, 1999;Yang et al. 2017;D'Agostino et al. 2017;Liu and Zhou, 2017). Actually, there is close coupling between sea surface temperature and precipitation in the tropics (Xie et al. 2010). Over the eastern Pacific-Atlantic section in CPL runs, the SST warming occurs north of 10°N while cooling occurs south of 10°N (Fig. 12c). The asymmetric SST change and thus the weakened meridional temperature gradient lead to a profound upward motion around 10°N in the real world (Fig. 8c), otherwise the upward motion is located further south (Fig. 8e). This is consistent with the northward shift of precipitation (Fig. 12c). However, in SOM runs, the The decomposition are based on 3P-DGAC method. Units: Pa/s. The value of vertical velocity have been multiplied by -1 from its original value, so the positive value is for upward motion meridional temperature change is relatively small, so the HC hardly shifts. The northward shift of precipitation in CPL runs can also be understood by regional ITCZ shift. The ITCZ positions are shown in Fig. 12. The ITCZ over the eastern Pacific-Atlantic Sect. (120°W-40°W) is related to the NASM precipitation. In this section, the mean position of ITCZ in CTRL_SOM is at 6.7°N, with no shift in NoTibet_SOM (Table 1 or red and orange dots in Fig. 12b). We further split the eastern Pacific-Atlantic section into eastern Pacific (120°W-80°W) and Atlantic Sect. (80°W-40°W) and find that the ITCZ is no shift in these two regions in SOM runs under the TP uplift (Table 1). However, in CPL runs, the ITCZ shifts northward by 2.2°N over the eastern Pacific-Atlantic section, with 1.3°N northward shift over the eastern Pacific and 2.7°N northward shift over the western Atlantic. The northward shift of ITCZ in CPL runs results in substantial precipitation increase over the central to north NAM region and decrease over the southern tip of the NAM region (Fig. 12d).
The northward shift of ITCZ over the eastern Pacific-Atlantic section can be explained by regional atmospheric energy budget (Boos and Korty 2016;Lintner and Boos 2019). The energy input into the atmosphere is shown in Figs. 13,14,15. In the TP region, the radiation fluxes at the TOA are remarkable reduced (Fig. 13a). The TP uplift generates anomalous low pressure over the TP region, which promotes the convection and moisture convergence, and thus the clouds formation (Fig. 13e, f). The increased clouds increase the planetary albedo (Fig. 13d), which hinder the incoming solar radiation (SW) and outgoing longwave radiation (LW) (Fig. 13b, c). In contrast to the heat fluxes reduction over the TP, the North Atlantic experiences a heat flux increase, with net incoming SW and outgoing LW all enhanced (Fig. 13a-c). The increased net downward SW is attributed to the low cloud reduction while the increased outgoing LW is due to surface warming instead of high cloud change (Fig. 13f). These processes are detailed described in . At the surface, the net heat fluxes into the atmosphere over the TP region are nearly unchanged, with the SW increase perfectly compensating the LW decrease ( Fig. 14a-c). The SW increase is due to enhanced surface albedo, which can reflect more SW from surface to the atmosphere. The LW reduction is caused by surface cooling The ITCZ position is defined as the location of the maximum precipitation at low latitudes (Mamalakis et al. 2021). The red rectangle outlines the eastern Pacific-Atlantic section (Fig. 4e). In the North Atlantic, the surface net heat fluxes into the atmosphere are remarkable increased (Fig. 14a), which is predominately caused by increased LW and latent heat flux (Fig. 14c, f). The AMOC establishment generates profound warming over the North Atlantic, which allows more LW loss from ocean to atmosphere (Fig. 14c). The warming can also enhance the ocean surface evaporation and thus the latent heat flux into the atmosphere (Fig. 14f). The energy flux change at the TOA and surface leads to energy loss over the Asian continent and energy gain over the North Atlantic (Fig. 15a) while the energy flux change over the SH is very small (Fig. 15). The inter-hemispheric energy asymmetry over the eastern Pacific-Atlantic section results in an increase of southward atmospheric energy transport at the equator (Fig. 15b), which is consistent with the revealed northward shift of the ITCZ (Fig. 12d). In addition, the atmospheric energy transport over tropics largely depends on the mean circulation change, i.e., the HC. Thus, the abnormal southward energy transport is completed by the northward shift the HC (Fig. 11c, d).
Actually, The eastern Pacific exhibits a significant cooling in our study (Fig. 4e), which may also contribute to the northward shift of eastern Pacific ITCZ. The tropical eastern Pacific cooling is not unlike the La Niña conditions. Previous studies demonstrate that the interannual variations of the ITCZ associated with ENSO are most pronounced over the Pacific (Dai and Wigley 2000;Adams et al. 2016a). These variations are characterized by an equatorward shift of the ITCZ during El Niño episodes and a poleward shift during La Niña episodes (Adams et al. 2016a, b). Our results are quite consistent with these prior works, in which we show that the La Niña-like SST change over tropical Pacific is followed by northward ITCZ shift over the eastern Pacific.

Summary and discussion
In this study, the impact of the TP topography on NASM precipitation is investigated in SOM runs and CPL runs. The TP uplift is found to enhance the NASM both directly via the atmospheric teleconnection and indirectly via the impact of the altered AMOC. First, the TP uplift alters the planetary wave patterns and generates an enhanced Atlantic  subtropical high, which strengthens the northeasterlies over tropical eastern Pacific-Atlantic section and thus the enhanced regional HC there. The strengthened upward motion leads to enhanced convection and thus the increased NASM precipitation. These processes are completed within several decades due to atmospheric adjustment. Second, the TP uplift can also enhance the NASM rainfall indirectly by triggering the AMOC establishment, which reduces the meridional temperature gradient, leading to a northward shift of the HC. The northward shift of the HC shifts the center of the ascending motion northward to 10°N and substantially enhance the convection over the NAM region, and then, the NASM precipitation. Actually, the pattern of SAT response over the tropical Atlantic in CPL runs is not unlike the Atlantic meridional mode, which is characterized by meridional SST gradients over the tropical Atlantic, and is reported to regulate the position of the ITCZ and Hadley circulation (Chang et al. 1997;Chiang et al. 2002;Chiang and Vimont 2004). This study shows a robust relationship between the topography of the TP and NASM precipitation, complimentary to previous perspective that TP uplift can substantially change rainfall over the Asian monsoon region. The topography of the TP in shaping the NASM precipitation is helpful for our understanding of the TP's role in the global climate system. Previous studies mostly focus on the TP impact on Asian monsoon precipitation, while we highlight the connection between the TP topography and North American monsoon precipitation and show that the existence of the TP leads to more humid NAM climate. The results   Fig. 15 a Equilibrium changes in net energy input into the atmosphere during boreal summer. Units: W/m 2 . Positive value represents heat flux incoming into the atmosphere. b Changes in the divergent meridional component of the atmospheric energy transport (AET) during boreal summer. Units: 10 7 W/m. The red arrow represents the southward transport of AET. The calculation for AET follows Mamalakis et al. (2021) obtained in this study may be model dependent. For example, the model resolution is quite low in this study, which cannot well capture the realistic topography, synoptic circulation, and mesoscale circulation that have been reported to be crucial for adequately representing the NAM (Adams et al. 1997;Pascale et al. 2016Pascale et al. , 2019Varuolo-Clarke et al. 2019). The NAM response to external forcings is also sensitive to SST biases (Pascale et al. 2016(Pascale et al. , 2017. However, the excessive cold tongue is a common feature in general circulation model , which may alter the response of NASM precipitation to TP uplift. In addition, we only consider the total precipitation change over NAM region in this study. Actually, changes in the timing and seasonal distribution of precipitation may also have significant ecological and societal consequences (Cook et al. 2013), which deserves more investigation in the future. Although this is a highly idealized modeling study with some model limitations, this work helps explain the quantitative role of the TP in the real world. The evolution of tropical American climate during the geological time period is not only related to regional circulation change, but also links to Asian high land regions.
Our modeling results may have applications for paleoclimate studies. For example, previous works use fossil flowers to reconstruct the Miocene climate over south Mexico and show that the environmental conditions of the Chiapas is warmer and drier than in the present (Hernández et al. 2020), which is consistent with our study. Su et al. (2018) use climate models and show that water vapor is divergent over tropical American continent without the TP, indicating that the tropical American continent is much drier in a world without the TP. Huber and Goldner (2012) analyze the Eocene monsoons and show that a high TP can generate precipitation increase over the central American region, consistent with this study. As suggested in previous works, the elevated TP heating can affect the Asian-Pacific Oscillation intensity, with positive tropospheric temperature deviation over the Eurasian continent and negative tropospheric temperature deviation over the central and eastern North Pacific, as well as the Atlantic Ocean (Nan et al. 2009;Duan et al. 2012). The teleconnection pattern is also found in our studies.
This work may also have some implications on modern climate. The TP uplift can induce thermal heating at middle troposphere by lapse rate relationship, which is similar to the rapid warming over the TP in the past decades (Duan and Xiao 2015). There are many studies show that the TP heating could enhance East Asian summer rainfall , the monsoon rainfall variability in Pakistan (Wang et al. 2019). In addition to the climate change over the Asian continent, the TP heating may also trigger the warming and high-pressure anomaly over the North Atlantic (Zhao et al. 2012;Lu et al. 2018). The high pressure anomaly over the North Atlantic may enhance the upward motion over the NAM region and result in enhanced NAM precipitation. However, The role of TP heating on NAM climate lacks direct investigation, more model studies and observations should be considered to unravel the role of TP in shaping the NAM climate.