上新世青藏高原北部隆升事件的模拟检验

石正国，刘晓东，沙莹莹

（1.中国科学院地球环境研究所 黄土与第四纪地质国家重点实验室，西安 710061；

2.中国科学院青藏高原地球科学卓越创新中心，北京 100101；

3.中国科学院大学，北京 100049）

doi：10.7515/JEE201502001

上新世青藏高原北部隆升事件的模拟检验

石正国1，2，刘晓东1，2，沙莹莹1，3

（1.中国科学院地球环境研究所 黄土与第四纪地质国家重点实验室，西安 710061；

2.中国科学院青藏高原地球科学卓越创新中心，北京 100101；

3.中国科学院大学，北京 100049）

关于上新世阶段青藏高原北部是否存在显著隆升事件这一问题目前依然存在较大的争议。本研究尝试利用沙尘气候模拟方法对上新世高原隆升的可能性做一评估。北太平洋风尘记录显示亚洲内陆干旱化在三到四百万年存在非常显著的增强，这可能是由于全球气候变冷和/或青藏高原隆升引起的。基于此，我们利用一个耦合了沙尘模块的气候模式，通过一系列敏感性试验测试大气沙尘沉积通量对于上新世以来主要气候强迫，包括冰盖变化，海表温度，大气二氧化碳浓度和青藏高原北部地形的响应。可能引起粉尘通量变化的大气环流和沙尘源区面积两个因素的改变在试验中都被考虑。模拟结果表明，冰盖扩张、海表温度下降和大气二氧化碳浓度下降这些全球变冷因素仅能解释中上新世和末次盛冰期北太地区粉尘沉积通量改变量的约三分之二，而剩下的三分之一需要归因于青藏高原北部的隆升。高原北部的隆升可以显著恶化内陆干旱状况从而提升大气粉尘浓度，支持了上新世高原北部存在明显隆升事件这一观点。

青藏高原；构造隆升；上新世；沉积速率；数值模拟

1 Introduction

Known as "the roof of the world" or "the third pole on earth"，the Tibetan Plateau（TP） is an immense upland with an averaged elevation of more than 4500 meters，covering an area of 2.5 million km2.A large number of numerical modeling studies have shown that the high and broad plateau exerts a significant role on forming the atmospheric circulations in northern hemisphere，especially on the monsoonal and arid environments in Asia（e.g.，Manabe and Terpstra，1974; Kutzbach et al，1989; Liu and Yin，2002; Zhang et al，2012）.Generally，the uplift of terrain can interact with the atmosphere via posing a physical obstacle for air currents，and heating the surface like a pump and intensifying the thermal contrasts，thus altering the wind fields.Hence，the reconstruction of the history of the Tibetan uplift is of great significance for us to better understand its effects on climate change during Cenozoic and has attracted a great deal of attention（e.g.，An et al，2001; Molnar and Stock，2009; Molnar et al，2010）.

The TP is widely accepted to be formed as a result of the collision of Indian subcontinent with southern Eurasia since the early Eocene（ca.55— 45 Myr，Myr：million years ago，Jain et al，2009）.The TP uplift is considered as an episodic process with several periods of rapid uplift.After northward penetration of India，a large area of central Tibet was substantially built and reached to 3000～4000 m approximately during ～25 Myr（DeCelles et al，2007; Wang et al，2008），probably causing the onset of the Asian desertif cation（Guo et al，2002）.Afterwards，the major construction of the Himalayas towards its present-day height underwent rapid uplift in mid-Miocene as a result of the underthrusting of slices of India's northern margin（Molnar et al，2010; Tapponnier et al，2001）.

In this study，our main focus is restricted to the Pliocene，especially during the period of 4—3 Myr.In previous geological studies，the TP is considered to have already reached its maximal height at 16— 14 Myr（Turner et al，1993; Coleman and Hodges，1995; Blisniuk，2001; Spicer，2003），or at 10—8 Myr（Harrison，1992; Burbank，1993; Molnar et al，1993）.They argued that there were no significant uplift events on the northern Tibetan Plateau（NTP） during the Pliocene epoch（Molnar and Stock，2009）.However，other researchers hold distinctly different views that significant uplift occurred in the NTP in the Pliocene（Shackleton and Chengfa，1988; Zheng，2000; An et al，2001）.Hence，debates still exist on the Pliocene uplift of the NTP.

Not only can the uplift history of TP be obtained from local records of tectonic activities，but also be indirectly deduced from the eolian deposits on the periphery.The dust mass accumulation rate（MAR） variations，which reflect the levels of the inland Asian aridity，have been closely related to tectonic events besides climate change，hence becoming a good indicator for the possible tectonic activities.One widely-used MAR record，covering the past several million years，is obtained by Rea et al（1998） from ODP 885/886 in the North Pacific.A very remarkable MAR increase is found during 4—3 Myr in the atmospheric dust level and they proposed that it is associated with the phased TP uplift at that time（Rea et al，1998）.Reconstructions of paleo-dust activities from the Chinese loess deposits located in another downwind sites of the inland Asian deserts also show similar increases in the Pliocene（Sun and An，2002）.

In fact，besides the NTP uplift，the significant global cooling from the Pliocene to the Quaternary glacial periods is also considered as one candidate responsible for increasing dust concentrations（Lu，2010）.In the Pliocene，a remarkable warming is suggested for nearly all around the globe（e.g.，Crowley，1996; Dowsett et al，1994）.Compared to the Pliocene warming，the dramatic glacial cooling is closely associated with the appearance of the Northern Hemisphere continental ice sheets，the significant decrease in sea surface temperature（SST） and reduced atmospheric carbon dioxide concentration.In thePliocene，no land ices existed on the Eurasian and North American continents except for Greenland，very much like today's distribution，while the ice volume and areal coverage on Greenland were reduced by 50% and the Antarctic ice was also slightly modified.The ocean temperature originated from a number of marine sites and areas（Dowsett et al，2009） indicate that it was much warmer at that time，especially in the high latitudes，and this would have led to a weakened meridional SST gradient（Brierley et al，2009）.Thus，one can not simply ascribe the remarkable dust increase to the Tibetan uplift.

In previous studies，the timing and locations of uplift episodes were usually deduced by geological evidence.However，these estimates are mostly originated from one cross-section or one site，which somewhat limits us to systematically assess the uplift process.Modeling approach can provide new insights to verify whether these geological hypotheses are reasonable from a more physically-based perspective.In the past 20 years，coupled dust-climate models have been rapidly developed（e.g.，Joussaume，1990; Mahowald et al，2006）.Now the dust emissiontransport-deposition cycle can be simulated quite well for both present-day and geological periods，e.g.，the Last Glacial Maximum（LGM，Mahowald et al，2006） and the Pliocene（Shi et al，2011），which helps to directly compare the simulated results with the reconstructed eolian dust activities.Thus，it is possible for us to match the simulated MARs to the observed values by designing different numerical experiments to examine whether certain uplift scenarios are reasonable in various geological periods corresponding to different heights of plateau boundary conditions.

Based on the above reasoning，we present the results from the attempt in this study to test whether significant Pliocene uplift occurred in the NTP from a modeling perspective.The effects of global cooling，related to the ice sheets expansion，the decrease in atmospheric CO2concentration and lowering of SST，are respectively evaluated in order to determine whether the cooling factors can account for the Pliocene increases in the MARs in the North Pacific.In section 2，the model basics and experimental design are introduced.The contributions of various climatic forcings，including the possible NTP uplift，on the eolian dust deposition fluxes，are analyzed and discussed in section 3.The main points are finally concluded in section 4.

2 Model description and experimental design

The model used in our study is the National Center for Atmospheric Research's（NCAR's） Community Atmosphere Model version 3（CAM3，Collins et al，2004） coupled with an on-line dust module（Mahowald et al，2006）.CAM3 is the atmospheric component of Community Climate System Model version 3（CCSM3），a coupled atmosphere-ocean-land-sea ice model（Collins et al，2006）.The dust mechanism generally follows the Dust Entrainment and Deposition Module（DEAD，Zender et al，2003）.In the module，dust can be entrained into the atmosphere when the friction velocity exceeds the threshold.The threshold friction velocity is calculated following a semi-empirical parameterization（Iversen and White，1982） and the inhibition of dust saltation by soil moisture is effectively considered（Fécan et al，1999）.The horizontal saltation flux is then calculated and converted to a vertical mass f ux with the sandblasting mass efficiency（Marticorena and Bergametti，1995; Alfaro et al，1997）.Four transport bins（0.1～1.0μm，1.0～2.5μm，2.5～5.0μm，and 5.0～10.0μm） are established in this model.The calculation of dry deposition processes is based on the parameterizations of gravitational settling and turbulent mix-out，and wet deposition is treated using a modif ed method of Rasch et al（2006）.

In the default version of the model，the satellitebased vegetation climatology used for the land surface calculations is employed to determine dust source distributions（Bonan et al，2002）.For our purpose，we employed a new dust source scheme（Shi et al，2011）.The distributions and fractions of dust source in inland Asia are obtained based on the simulated annual precipitation levels.To avoid the model's systemic errors，the annual mean precipitationoriginating from the Climate Prediction Center（CPC） merged analysis of precipitation（CMAP） datasets from 1979 to 2000（Xie and Arkin，1997） are used as the standards and only the simulated anomalies in each experiment are added to the present-day precipitation，to obtain more robust precipitation estimates.The land grids where precipitation is less than 200 mm·a-1are defined as arid regions（with the desert fraction equal to 1）.Those grids with annual precipitation between 200 mm and 300 mm are defined as semi-arid regions（with the desert fraction equal to 0.5）.Only the dust sources north of 25°N and east of 40°E on the inner Asian continent are considered and those out of these regions are excluded since all our focus is the inland Asian deserts and their downwind regions.Consistent results with previous studies（Mahowald et al，2006） show that our dust sources are quite reasonable.

Two series of simulations，respectively named standard and sensitivity runs，are conducted in our study.All experiments are listed in Table 1.First of all，two offline simulations（i.e.，PDn and LGn），representing present-day（Interglacial period） and LGM（Glacial period），without dust module，are employed to calculate the dust source distributions.Then the corresponding atmospheric dust cycles and MARs during two periods are simulated（i.e.，PDc and LGc，respectively） to represent the MAR levels during the Quaternary glacial cycles.The PD cases in the PDn and PDc experiments are simulated with the historical sea surface temperature data（AMIP SST） averaged between 1950—1993.For the LGM，including the LGn and LGc experiments，we generally follow the PMIP framework（Joussaume and Tayloy，1995）.The atmospheric CO2concentration is set to 180 ppmv and the orbital parameters are kept as present.

Table 1 Experimental design.The experiments are divided into two series：standard and sensitivity runs.The standard runs contain PD and LG cases respectively for the present day and LGM conditions.Contributions from three cooling factors（ICE，SST and CO2） and different NTP heights（NT0，3 and 6） are quantif ed in the sensitivity runs，respectively.The ending letters of the names indicate the different dustsources（i.e.，n，f and c for none，f xed and changed dust sources，respectively）.The names of the dust sources follow thecorresponding experiments（e.g.，the dust source PDn is calculated from the experiment PDn.）

In the sensitivity runs，we have evaluated the contributions on the MARs from various factors related to the global cooling and TP uplift，respectively.Since the atmospheric dust cycles are affected both by the changes in atmospheric circulations and source areas，simulations with fixed and varied dust sources are employed so as to distinguish the contributions from the two effects.The climatic boundary conditions for the Pliocene are from the PRISM3D datasets（Dowsett et al，2009），which are reconstructed for mid-Pliocene，about 3.3—3 Myr before.We use the PRISM3D land covers and monthly SST as the approximates and the CO2concentration is set to 405 ppmv.Compared to Colder LGM，the individual contributions from three climatic forcings are respectively calculated（ICE，SST，CO2） once each of them is updated to the warmer Pliocene values.On the other hand，we conduct several experiments for the possible tectonic forcing to test the sensitivity of MAR changes to different NTP heights.In these experiments，we keep the elevation of the southern TP（including the Himalayas） unchanged but reduce the elevation of the NTP by different extents（Fig.1; NT6，NT3，and NT0 denoting the experiments with 66%，33% of the modern height and 200 m in the NTP，respectively） since the uplift events and northward expansions are considered to only occur on the northern TP during this period.All experiments are performed at T42，corresponding to a horizontal resolution of 2.8°×2.8°.No feedbacks from dust aerosols are included in our experiments.After a 15-year spin-up，each simulation is integrated for another 15 years and the results are then averaged and analyzed.

Our focus is the North Pacific regions（170°E～200°E，40°N～50°N）.Since the marine location is very far away from the source regions，only suff ciently fine dust particles（usually smaller than 10μm） can be delivered long distances in the atmosphere to reach this region.Therefore，the particle range of the reconstructed MAR can be directly compared to our simulated results with negligible errors.The MAR variation from ODP 885/886 in North Pacif c（Rea et al，1998） is chosen for comparison.

3 Contributions from climatic and tectonic forcings to the MARs

3.1 MARs in the present day and LGM

As the standard runs，the simulated annual mean MARs over the East Asian regions in the present day and LGM are shown in Fig.2.The basic patterns of deposition fluxes in these two stages are consistent with the largest deposition fluxes being usually found close to the source regions，particularly in central Asia and northern China.A remarkable increase in the highlatitude dust sources of central Asia during the LGM has contributed to the larger deposition fluxes than during the present day.Dust originated from inland Asia is mainly delivered east towards the Pacific primarily due to the dominance of the westerlies，which is in good agreement with the modern observations（Mahowald et al，2009）.For the LGM，an annual MAR of 13.3 g·m-2·yr-1is averaged over our focus region in the North Pacif c，appromximately 53% higher than the modern time（8.7 g·m-2·yr-1）.Comparisons between the simulated MARs and the Dust Indicators and Records of Terrestrial and Marine Palaeoenvironments（DIRTMAP，Kohfeld and Harrison，2001） observations show that their values are within the same order of magnitude over the North Pacific（Fig.3）.Hence，our model can reconstruct the MARs quite well and this allows us to further test the sensitivities of MARs to different forcings.

3.2 Effects of global cooling factors

The inland Asian aridity，which can be traced back to about 22 Myr ago，was becoming more and more severe and achieved its peak after the Quaternary glacial periods（Guo et al，2002）.Global cooling is no doubt a possible factor of the intensification of aridity.Compared to the warm Pliocene，the remarkable glacial（e.g.，LGM） cooling is resulted from three primary factors：the appearance of the Northern Hemisphere continental ice sheets，the signif cant decrease in SST and reduced atmospheric carbon dioxide concentration.There is no doubt that these factors are highly dependent and can interact with each other，but in this study，we assume them as independent forcings and evaluate individually their impacts on the Asian aridity.

Fig.1 Tibetan Plateau elevations（m，shaded） and the assumed reductions in the northern Tibetan Plateau（m，contour） in our study.NT6f/c（a），NT3f/c（b），NT0f/c（c）

The differences in the responses to climatic forcings over Asia are shown in Fig.4.Significant temperature decreases can be induced by all three forcings，totally achieving a global average of about 3.50°C by the combined effects.The ice sheets significantly reduce the temperature in the northern high latitudes directly by changing the surface albedo and the local reduction of temperature on glacial ice can even reach to more than 20°C（not shown）.However，in the tropics/subtropics of Asia，the regional cooling is not very obvious（Fig.4a）.Furthermore，the temperature might be even raised in certain areas，e.g.，in parts of the Tibetan Plateau and the marine regions of the West Pacif c.Over most source regions，the surface wind speeds are signif cantly enhanced by more than 0.5 m·s-1，facilitating the dust mobilization.The SST variation is of the greatest importance among the three forcings and cools the atmosphere by 2.36°C.The enhanced meridional contrasts in SST have also intensified the tropical-polar meridional temperature difference.A cooling of up to 2°C can be found almost across the entire Eurasian continent excluding Indiaand southeastern Asia，together with an intensified wind f eld（Fig.4c）.The contribution of CO2reduction is much less compared to the other two forcings but it might play a very significant role in the limited cooling in the tropical regions since its effect is much more globally uniform（Fig.4e）.

Fig.2 The simulated MARs（g·m-2·yr-1） over East Asia.present-day（a），LGM（b）

The cooling generally results in a significant decrease in the precipitation.Globally averaged，the annual precipitation differences respectively reach to -0.04 mm·d-1，-0.17 mm·d-1and +0.05 mm·d-1due to the three climatic forcings and the influences are regionally distinct.In Asia，the ice sheets and SST changes have largely suppressed precipitation over the continent，particularly in East Asia，while in tropical marine regions，rainfall is significantly enhanced by more than 0.6 mm·d-1（Fig.4b and Fig.4d）.However，annual precipitation across Asia is not very sensitive to the CO2changes with the greatest responses no more than 0.3 mm·d-1.Over the inland areas，the decreasing CO2might have lead to increased rainfall and slightly alleviate the arid conditions（Fig.4f）.The differences in summer show similar patterns as those in the annual average ones，particularly in the Asian monsoon regions where summer rainfall dominates annual total precipitation，indicating that dramatic continental cooling in summer can reduce the ocean-land thermal contrasts and thus weaken the Asian monsoon system during the LGM.Hence，the monsoon-related rainfall is suppressed over coastal regions in Asia and the inland aridity is more or less strengthened.

Based on the simulated annual precipitation，the dust sources for each experiment are calculated.During the LGM，deserts cover most mid-latitude regions in the inland Asian continent，extending fromthe Middle East to East Asia（Fig.5a）.Compared to the present-day condition，the source areas have been signif cantly enlarged，especially in the north of 50°N in central Asia，which is consistent with previous studies（Mahowald et al，2006）.Without the inf uences from the ice sheets，the sources become quite similar with the present-day distribution（Fig.5b），indicating that the higher-latitude arid regions are sensitive to the ice sheets expansion.SST variations have also affected the Asian aridity.If no cooler SST forcing，more rainfall can obviously reduce the sources in inland Asia，by changing the land cover either from arid to semi-arid or from semi-arid to sub-humid（Fig.5c）.Nearly no changes are found in the responses of source areas to the CO2variation（Fig.5d） since total precipitation is not very sensitive to this factor as indicated earlier.Hence，the inland aridity in Asia and the corresponding dust sources have been signif cantly impacted by global cooling，especially by the expansion of ice sheets and SST changes.

3.3 Effects of the NTP uplift

The NTP uplift during the Pliocene，although still controversial in the scientif c community，is considered as another candidate for explaining such MAR increases from the Pliocene level to LGM.Although this uplift is regional，it can significantly influence the local thermal distributions，thus modulating not only the Asian monsoon system and inland aridity，but also the global climate（Broccoli and Manabe，1992; Kutzbach et al，1989）.In this section，we intend to evaluate the possible effects from the Tibetan uplift and to test the sensitivity of MARs over the North Pacific to different NTP heights.Compared to present-day，the NTP uplift significantly suppresses precipitation from 1.09 mm·d-1to 0.87 mm·d-1over the inland regions（Table 2） after the NTP has already reached to a relative high altitude（NT6c），together with a temperature reduction of 3.7°C over source regions（Table 2）.In the NT6c experiment，the dust sources only exist in certain parts of central Asia and northwestern China.Decreased precipitation at the present-day stage has intensified the inland aridity and a remarkable increase in the source areas is found when comparing with the NT6c results（Fig.6a and Fig.6b）.However，the surface wind speeds are reduced（Table 2），which exerts a negative effect on dust emission.Furthermore，we notice that these significant climatic responses to the uplift only occur when the NTP has reached to a relatively high altitude.

Fig.3 Comparisons between the simulated MARs and DIRTMAP observations over the North Pacif c in the presentday and LGM stages

When the NTP is relatively low，the inland aridity across Asia is not very sensitive to the different NTP heights.The dust source regions do not significantly change responding to the uplift，so that very similar source distributions are found in these three scenarios（Fig.6b，Fig.6c and Fig.6d）.In fact，the source areas have even slightly decreased along the northern edges of the TP from NT3c to NT6c experiment，which indicates that the dry conditions over the NTP has been largely alleviated（as indicated by the changes from source to non-source） after the NTP height reaches to the threshold value.Similarly，no significant changes are indicated in the other climatic responses，including the soil water content，precipitation and wind speed（Table 2）.Although the role of the southern TP is not evaluated in this study，our results still indicate that the TP uplift there might only exert a very minor effect on the inner Asian aridity during specific geological periods when the TP is still low.Therefore，global cooling might have played a much more important role during the Cenozoic than expected.

Fig.4 Climate response to global cooling forcings.Ice expansion（a and b），SST decrease（c and d），and CO2reduction（e and f）.The left column is the differences in mean annual temperature（℃，shaded） and in surface wind speed（mm·s-1，contours）.The right column is the anomalies of mean annual precipitation rate（mm·d-1，shaded） and in summer precipitation rate（mm·d-1，contour）.The global mean values for the differences in mean annual temperature and precipitation are shown in brackets

Table 2 The climatic responses to different northern TP heights：surface temperature（°C），precipitation rate（mm·d-1），soil water content（mm3·mm-3），and surface wind speed（m·s-1） over the present-day source regions

Fig.6 Precipitation-controlled dust source distributions including the arid（fraction=1，yellow） and semi-arid regions（fraction=0.5，light yellow） in our experiments corresponding to the NTP uplift.PDc（a），NT6c（b），NT3c（c），NT0c（d）.

3.4 Explanation for Pliocene MAR increase

Employing the different source areas，we further examined the responses of eolian dust deposition to global cooling and the Tibetan uplift（Fig.7）.As mentioned above，global cooling can affect the atmospheric dust levels through two primary ways，by altering the dust source areas and modulating circulation patterns.With the LGM sources unchanged，the simulated MARs are respectively reduced by 0.9 g·m-2·yr-1and 3.4 g·m-2·yr-1without the inf uences of ice sheets and SST，which indicates that these climatic forcings can signif cantly increase the dust MARs by changing the atmospheric circulations alone.However，the CO2decrease might have reduced the dust depositions by 1.1 g·m-2·yr-1.After considering the retreat in the source areas，the effects of ice sheet，SST and CO2on the MAR become 2.7 g·m-2·yr-1，4.8 g·m-2·yr-1and -0.5 g·m-2·yr-1，respectively，indicating that the MAR variation is no doubt benef ted from both the changes in circulations and source areas.Based on our results，SST decrease is the factor of the great importance in global cooling and induces the largest increase for dust depositions.On the other hand，dust variation is not very sensitive to the CO2changes and the responses might be even reversed（-0.6 g·m-2·yr-1）.The dust deposition has also responded well to the NTP uplift after the NTP has reached to certain elevation（66% of its modern height）.Although the surface wind speed is signif cantly suppressed（Table 2），the MAR over the North Pacif c is still raised from 4.1 g·m-2·yr-1（NT6c） to 8.7 g·m-2·yr-1（PDc） as a result of the increase in the inland sources（Fig.6a and Fig.6b）.Due to the unchanged sources，the MAR variation is not very sensitive to the NTP uplift when the NTP is low.The MAR over the North Pacific has already reached to a level of 3.8 g·m-2·yr-1even when the NTP does not exist.

Fig.7 Simulated annual mean MARs（g·m-2·yr-1） over the North Pacific in the sensitivity experiments.The gray line shows the levels during the LGM and present-day stages

From the reconstructed MAR at ODP885/886 in the North Pacif c，a remarkable increase by approximately 10 folds is suggested from 1.3 g·m-2·yr-1（averaged for 4.2—3.6 Myr） in Pliocene to 12.1 g·m-2·yr-1（2.60 Myr） in the glacial cycles（Fig.8）.Due to its coarse time resolution，it is diff cult for us to distinguish the differences between the glacial and interglacial periods，but a mean value of the simulated MARs of 11.0 g·m-2·yr-1in the preset day and LGM is quite consistent with the observed value of 12.1 g·m-2·yr-1.As calculated above，the combined effects of global cooling forcings have raised the MARs by approximately 7 g·m-2·yr-1to 13.3 g·m-2·yr-1in the LGM（Fig.8），which cover about 2/3 of the total change.We will see whether the TP uplift，as the only factor other than global cooling，can explain the remaining portion of this MAR change.In our simulations，the NTP uplift can increase the source areas and lead to a further enhancement of 4.6 g·m-2·yr-1（NT6c，Fig.7），just about to cover the remaining 1/3 fraction.After considering the uplift together，the simulated MAR has achieved 1.7 g·m-2·yr-1in the Pliocene（Fig.8），in good agreement with the ODP885/886 value.Hence，we can argue that the main reasons for the increase in the atmospheric dust MARs during Pliocene are actually the combined effects of global cooling and the Tibetan uplift.In previous studies，the Tibetan uplift has been proposed to be one of the triggering mechanisms of the Quaternary Northern Hemisphere glaciations through its climatic and geochemical effects（Raymo，1994）.If that is the case，then the Tibetan uplift might be the ultimate reason for all the MAR changes during Pliocene.

Fig.8 Simulated MAR levels averaged over the North Pacific（red dots） and the reconstructed MAR variation during the last 4.2 Myr（black curve） at ODP 885/886.The referenced values of ODP885/886 for glacial cycles and Pliocene（blue lines） are respectively averaged from 2.6—0 Myr，and from 4.2—3.6 Myr.The contributions by the cooling and uplift are also marked.The period of Quaternary（～2.6 Myr） is shaded by yellow.

4 Conclusions

A modeling approach is used in this study as the first attempt to explore the possibility of significant uplift events over the NTP during the Pliocene based on the simulated eolian dust MARs.To explain the remarkable increase in the dust depositions over North Pacific，the influences from global cooling and the northern Tibetan uplift are respectively quantified using a climate-dust model，considering both the changes in the dust source areas and atmospheric circulations.The major findings are concluded as follows:

Global cooling has increased the eolian dust MARs over North Pacific by approximately 7 g·m-2·yr-1，which could not account for all the observed MAR changes.The enlarged dust source areas and enhanced surface wind velocities are found in the sensitivity tests，indicating that the cooling can indeed affect the atmospheric dust emissions via these two ways.The SST reduction and ice sheet expansion are proposed to be the most important factors，nearly explaining all the effects of the cooling，while the dust deposition might not be very sensitive to the CO2changes.Overall，the combined effects of global cooling factors can explain approximately 2/3 of the observed MAR increases since the mid-Pliocene.

As the other candidate，the Tibetan uplift has also remarkably impacted the atmospheric dust MARs through increasing the inland Asian aridity.In our simulations，the uplift of the NTP from 66% of its present height to the present elevation has increased the MARs over the North Pacific by 4.6 g·m-2·yr-1，large enough to cover the remaining variation unexplained by the global cooling forcings.Hence，we propose that considerable NTP uplift events should exist during the Pliocene and that the MAR increase is associated with both global cooling and the NTP uplift.However，the North Pacific MAR is not very sensitive to different NTP heights when the NTP is relative low.

Our analysis has provided useful supplemental evidence to prove the Pliocene uplift of the NTP from a new modeling point of view.However，our conclusions might be model-dependent and need to be verified by other climate models.Hence，further geological records and simulation studies are still eagerly desirable to clarify the history of the Tibetan uplift.

Acknowledgments: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 41105060）.The authors would like to thank Prof.Zhi-Sheng An for insightful suggestions.

Alfaro S C，Gaudichet A，Gomes L，et al.1997.Modeling the size distribution of a soil aerosol produced by sandblasting［J］.Journal of Geophysical Research，102（D10）：11239-11249.

An Z S，Kutzbach J E，Prell W L，et al.2001.Evolution of Asian monsoons and phased uplift of the Himalaya-Tibetan plateau since Late Miocene times［J］.Nature，411（6833）：62-66.

Blisniuk P M，Hacker B R，Glodny J，et al.2001.Normal faulting in central Tibet since at least 13.5 Myr ago［J］.Nature，412：628-631.

Bonan G B，Levis S，Kergoat L，et al.2002.Landscapes as patches of plant functional types：An integrating concept for climate and ecosystem models［J］.Global Biogeochemical Cycles，16（2）：5-1-5-23.

Brierley C M，Fedorov A V，Liu Z，et al.2009.Greatly expanded tropical warm pool and weakened Hadley circulation in the early Pliocene［J］.Science，323（5922）：1714-1718.

Broccoli A，Manabe S.1992.The effects of orography on midlatitude Northern Hemisphere dry climates［J］.Journal of Climate，5（11）：1181-1201.

Burbank D W，Derry L A，France-Lanord C.1993.Reduced Himalayan sediment production 8 Myr ago despite an intensif ed monsoon［J］.Nature，364（6432）：48-50.

Coleman M，Hodges K.1995.Evidence for Tibetan plateau uplift before 14 Myr ago from a new minimum age for east-west extension［J］.Nature，374（6517）：49-52.

Collins W D，Bitz C M，Blackmon M L，et al.2006.Thecommunity climate system model version 3（CCSM3）［J］.Journal of Climate，19（11）：2122-2143.

Collins W D，Rasch P J，Boville B A，et al.2004.Description of the NCAR community atmosphere model（CAM 3.0）［M］.Technical Note TN-464+STR，National Center for Atmospheric Research，Boulder，CO.

Crowley T J.1996.Pliocene climates：the nature of the problem［J］.Marine Micropaleontology，27（1）：3-12.

DeCelles P G，Quade J，Kapp P，et al.2007.High and dry in central Tibet during the Late Oligocene［J］.Earth and Planetary Science Letters，253（3）：389-401.

Dowsett H，Robinson M，Foley K.2009.Pliocene threedimensional global ocean temperature reconstruction［J］.Climate of the Past，5（4）：769-783.

Dowsett H，Thompson R，Barron J，et al.1994.Joint investigations of the Middle Pliocene climate I：PRISM paleoenvironmental reconstructions［J］.Global and Planetary Change，9（3）：169-195.

Fécan F，Marticorena B，Bergametti G.1999.Parametrization of the increase of the aeolian erosion threshold wind friction velocity due to soil moisture for arid and semi-arid areas［J］.Annales Geophysicae，17：149-157.

Guo Z，Ruddiman W F，Hao Q，et al.2002.Onset of Asian desertif cation by 22 Myr ago inferred from loess deposits in China［J］.Nature，416（6877）：159-163.

Harrison T M，Copeland P，Kidd W，et al.1992.Raising tibet［J］.Science，255（5052）：1663-1670.

Iversen J，White B.1982.Saltation threshold on earth，mars and venus［J］.Sedimentology，29（1）：111-119.

Jain A，Lal N，Sulemani B，et al.2009.Detrital-zircon f ssiontrack ages from the Lower Cenozoic sediments，NW Himalayan foreland basin：Clues for exhumation and denudation of the Himalaya during the India-Asia collision［J］.Geological Society of America Bulletin，121（3-4）：519-535.

Joussaume S.1990.Three-dimensional simulations of the atmospheric cycle of desert dust particles using a general circulation model［J］.Journal of Geophysical Research 95（D2）：1909-1941.

Joussaume S，Taylor K.1995.Status of the paleoclimate modeling intercomparison project（PMIP）［J］.World Meteorological Organization-Publications-WMO TD：425-430.

Kohfeld K E，Harrison S P.2001.DIRTMAP：the geological record of dust［J］.Earth-Science Reviews，54（1）：81-114.

Kutzbach J，Guetter P，Ruddiman W，et al.1989.Sensitivity of climate to late Cenozoic uplift in Southern Asia and the American West：Numerical experiments［J］.Journal of Geophysical Research，94（D15）：18393-18407.

Liu X，Yin Z Y.2002.Sensitivity of East Asian monsoon climate to the uplift of the Tibetan Plateau［J］.Palaeogeography，Palaeoclimatology，Palaeoecology，183（3）：223-245.

Lu H，Wang X，Li L.2010.Aeolian sediment evidence that global cooling has driven late Cenozoic stepwise aridification in central Asia［J］.Geological Society，London，Special Publications，342（1）：29-44.

Mahowald N M，Engelstaedter S，Luo C，et al.2009.Atmospheric Iron Deposition：Global Distribution，Variability，and Human Perturbations［J］.Annual Review of Marine Science，1：245-278.

Mahowald N M，Muhs D R，Levis S，et al.2006.Change in atmospheric mineral aerosols in response to climate：Last glacial period，preindustrial，modern，and doubled carbon dioxide climates［J］.Journal of Geophysical Research，111（D10），doi:10.1029/2005JD006653.

Manabe S，Terpstra T B.1974.The effects of mountains on the general circulation of the atmosphere as identif ed by numerical experiments［J］.Journal of the Atmospheric Sciences，31（1）：3-42.

Marticorena B，Bergametti G.1995.Modeling the atmospheric dust cycle：1.Design of a soil-derived dust emission scheme［J］.Journal of Geophysical Research，100（D8）：16415-16430.

Molnar P，Boos W R，Battisti D S.2010.Orographic controls on climate and paleoclimate of Asia：thermal and mechanical roles for the Tibetan Plateau［J］.Annual Review of Earth and Planetary Sciences，38（1）：77.

Molnar P，England P，Martinod J.1993.Mantle dynamics，uplift of the Tibetan Plateau，and the Indian monsoon［J］.Reviews of Geophysics，31（4）：357-396.

Molnar P，Stock J M.2009.Slowing of India's convergence with Eurasia since 20 Ma and its implications for Tibetan mantle dynamics［J］.Tectonics，28（3），doi:10.1029/2008TC002271.

Rasch P J，Coleman D B，Mahowald N，et al.2006.Characteristics of atmospheric transport using three numerical formulations for atmospheric dynamics in a single GCM framework［J］.Journal of Climate，19（11）：2243-2266.

Raymo M.1994.The initiation of Northern Hemisphere glaciation［J］.Annual Review of Earth and Planetary Sciences，22：353-383.

Rea D K，Snoeckx H，Joseph L H.1998.Late Cenozoic eolian deposition in the North Pacific：Asian drying，Tibetan uplift，and cooling of the northern hemisphere［J］.Paleoceanography，13（3）：215-224.

Shackleton R，Chengfa C.1988.Cenozoic uplift and deformation of the Tibetan Plateau：the geomorphological evidence［J］.Philosophical Transactions of the Royal Society of London，Series A，Mathematical and Physical Sciences，327（1594）：365-377.

Shi Z，Liu X，An Z，et al.2011.Simulated variations of eolian dust from inner Asian deserts at the mid-Pliocene，last glacial maximum，and present day：contributions from the regional tectonic uplift and global climate change［J］.Climate Dynamics，37（11-12）：2289-2301.

Spicer R A，Harris N B，Widdowson M，et al.2003.Constant elevation of southern Tibet over the past 15 million years［J］.Nature，421（6923）：622-624.

Sun Y，An Z.2002.History and variability of Asian interior aridity recorded by eolian flux in the Chinese Loess Plateau during the past 7 Ma［J］.Science in China Series D：Earth Sciences，45（5）：420-429.

Tapponnier P，Zhiqin X，Roger F，et al.2001.Oblique stepwise rise and growth of the Tibet Plateau［J］.Science，294（5547）：1671-1677.

Turner S，Hawkesworth C，Liu J，et al.1993.Timing of Tibetan uplift constrained by analysis of volcanic rocks［J］.Nature，364（6432）：50-54.

Wang C，Zhao X，Liu Z，et al.2008.Constraints on the early uplift history of the Tibetan Plateau［J］.Proceedings of the National Academy of Sciences，105（13）：4987-4992.

Xie P，Arkin P A.1997.Global precipitation：A 17-year monthly analysis based on gauge observations，satellite estimates，and numerical model outputs［J］.Bulletin of the American Meteorological Society，78（11）：2539-2558.

Zender C S，Bian H，Newman D.2003.Mineral Dust Entrainment and Deposition（DEAD） model：Description and 1990s dust climatology［J］.Journal of Geophysical Research，108（D14），doi:10.1029/2002JD002775.

Zheng H，Powell C M，An Z，et al.2000.Pliocene uplift of the northern Tibetan Plateau［J］.Geology，28（8）：715-718.

Zhang R，Jiang D，Liu X，et al.2012.Modeling the climate effects of different subregional uplifts within the Himalaya-Tibetan Plateau on Asian summer monsoon evolution［J］.Chinese Science Bulletin，57：4617-4626.

Did northern Tibetan Plateau uplift during Pliocene? A modeling test

SHI Zheng-guo1，2，LIU Xiao-dong1，2，SHA Ying-ying1，3
（1.State Key Laboratory of Loess and Quaternary Geology，Institute of Earth Environment，Chinese Academy of Sciences，Xi'an 710061，China; 2.CAS Center for Excellence in Tibetan Plateau Earth Sciences，Beijing 100101，China; 3.University of Chinese Academy of Sciences，Beijing 100049，China）

Different opinions have been derived from geological evidence on whether there was signif cant tectonic uplift in the northern Tibetan Plateau（NTP） during the Pliocene.We made the f rst attempt in this paper to explore the possibility of the Pliocene uplift by climate modeling.Previous studies on eolian sediment records show that the inland Asian aridity was largely intensified 4—3 million years ago，which was proposed to be induced by global cooling and/or the Tibetan uplift.Employing a coupled climate-dust model，we conducted a series of experiments in order to test the sensitivity of atmospheric dust deposition fluxes to different forcings including the ice sheets，sea surface temperature，atmospheric CO2concentration and the NTP topographic height.Both changes in atmospheric circulations and dust source areas are considered.The results show that，global cooling induced by all the above-mentioned factors except the NTP uplift can only explain approximately 2/3 of the increase in the dust deposition f uxes in the downwind North Pacif c from mid-Pliocene to the last glacial maximum and the remaining 1/3 of the increase can be attributed to NTP uplift.The NTP uplift has remarkably strengthened the inland aridity and raised the dust levels，suggesting that considerabletectonic uplift events should have occurred in the NTP during Pliocene.

Tibetan Plateau; tectonic uplift; Pliocene; mass accumulation rate; climate modeling

P467；P534.6

A

1674-9901（2015）02-0067-14

2014-11-13

中国科学院战略性先导科技专项（XDB03020601）； 国家自然科学基金（41290255，41105060）

石正国，E-mail：shizg@ieecas.cn