JIANG Dabang,YU Ge,ZHAO Ping,CHEN Xing,LIU Jian,LIU Xiaodong,WANG Shaowu,ZHANG ZhongshiYU Yongqiang,LI Yuefeng0,JIN Liya,XU Ying2,JU Lixia2,,ZHOU Tianjun,2,and YAN Xiaodong

1Nansen-Zhu International Research Centre,Institute of Atmospheric Physics,Chinese Academy of Sciences,Beijing100029

2Climate Change Research Center,Chinese Academy of Sciences,Beijing100029

3State Key Laboratory of Lake Science and Environment,Nanjing Institute of Geography and Limnology, Chinese Academy of Sciences,Nanjing210008

4State Key Laboratory of Severe Weather,Chinese Academy of Meteorological Sciences,Beijing100081

5School of Atmospheric Sciences,Nanjing University,Nanjing210093

6Key Laboratory for Virtual Geographic Environment of Ministry of Education,School of Geography Science, Nanjing Normal University,Nanjing210023

7State Key Laboratory of Loess and Quaternary Geology,Institute of Earth Environment,Chinese Academy of Sciences,Xi’an710075

8Department of Atmospheric and Oceanic Sciences,School of Physics,Peking University,Beijing100871

9State Key Laboratory of Numerical Modeling for Atmospheric Sciences and Geophysical Fluid Dynamics, Institute of Atmospheric Physics,Chinese Academy of Sciences,Beijing100029

10China Meteorological Administration Training Center,Beijing100081

11Key Laboratory of Western China’s Environmental Systems,Lanzhou University,Lanzhou730000

12National Climate Center,China Meteorological Administration,Beijing100081

13State Key Laboratory of Earth Surface Processes and Resource Ecology,Beijing Normal University,Beijing100875

Paleoclimate Modeling in China:A Review

JIANG Dabang?1,2,YU Ge3,ZHAO Ping4,CHEN Xing5,LIU Jian6,LIU Xiaodong7,WANG Shaowu8,ZHANG Zhongshi1YU Yongqiang9,LI Yuefeng10,JIN Liya11,XU Ying12,JU Lixia2,1,ZHOU Tianjun9,2,and YAN Xiaodong13

1Nansen-Zhu International Research Centre,Institute of Atmospheric Physics,Chinese Academy of Sciences,Beijing100029

2Climate Change Research Center,Chinese Academy of Sciences,Beijing100029

3State Key Laboratory of Lake Science and Environment,Nanjing Institute of Geography and Limnology, Chinese Academy of Sciences,Nanjing210008

4State Key Laboratory of Severe Weather,Chinese Academy of Meteorological Sciences,Beijing100081

5School of Atmospheric Sciences,Nanjing University,Nanjing210093

6Key Laboratory for Virtual Geographic Environment of Ministry of Education,School of Geography Science, Nanjing Normal University,Nanjing210023

7State Key Laboratory of Loess and Quaternary Geology,Institute of Earth Environment,Chinese Academy of Sciences,Xi’an710075

8Department of Atmospheric and Oceanic Sciences,School of Physics,Peking University,Beijing100871

9State Key Laboratory of Numerical Modeling for Atmospheric Sciences and Geophysical Fluid Dynamics, Institute of Atmospheric Physics,Chinese Academy of Sciences,Beijing100029

10China Meteorological Administration Training Center,Beijing100081

11Key Laboratory of Western China’s Environmental Systems,Lanzhou University,Lanzhou730000

12National Climate Center,China Meteorological Administration,Beijing100081

13State Key Laboratory of Earth Surface Processes and Resource Ecology,Beijing Normal University,Beijing100875

This paper provides a review of paleoclimate modeling activities in China.Rather than attempt to cover all topics,we have chosen a few climatic intervals and events judged to be particularly informative to the international community.In historical climate simulations,changes in solar radiation and volcanic activity explain most parts of reconstructions over the last millennium prior to the industrial era,while atmospheric greenhouse gas concentrations play the most important role in the 20th century warming over China.There is a considerable model-data mismatch in the annual and boreal winter temperature change over China during the mid-Holocene[6000 years before present(ka BP)],while coupled models with an interactive ocean generally perform better than atmospheric models.For the Last Glacial Maximum(21 ka BP),climate models successfully reproduce the surface cooling trend over China but fail to reproduce its magnitude,with a better performance for coupled models.At that time,reconstructed vegetation and western Pacif c sea surface temperatures could have signif cantly affected the East Asian climate,and environmental conditions on the Qinghai-Tibetan Plateau were most likely very different to the present day.During the late Marine Isotope Stage 3(30-40 ka BP),orbital forcing and Northern Hemisphere glaciation,as well as vegetation change in China,were likely responsible for East Asian climate change.On the tectonic scale, the Qinghai-Tibetan Plateau uplift,the Tethys Sea retreat,and the South China Sea expansion played important roles in the formation of the East Asian monsoon-dominant environment pattern during the late Cenozoic.

paleoclimate modeling,China,millennium,orbital scale,tectonic scale

1. Introduction

Physically-based climate models have been widely applied in the geosciences in recent decades.One of the hottesttopics by far is the projection of anthropogenic climate change on the basis of a series of scenarios of atmospheric greenhouse gas and aerosol concentrations.However,high levels of uncertainty still exist in current climate models. In an attempt to objectively evaluate the eff cacy of climate models under various boundary conditions,and to scientifically understand the mechanism of climate change over awide range of timescales,more and more attention has been paid worldwide to paleoclimate modeling since the 1990s (Joussaume and Taylor,1995).By comparing simulations with reconstructions,our knowledge on past climate change has been greatly advanced in many respects(Cane et al., 2006;Jansen et al.,2007).

As an interdisciplinary subject,studies on past climate changeconsistofbothreconstructionsandsimulations.Since the pioneering research of Chu(1973),a great deal of climate reconstruction has been carried out in China through the use of historical documents,tree rings,ice cores,stalagmites,peat,pollen,f uvial and marine sediments,lacustrine sediments,loesses,and paleosols.The regional climate of China has been found to have undergone large changes on varying timescales.These proxy data also form a solid foundationuponwhichpaleoclimatemodelingcanbebuilt,allowing qualitative and/or quantitative model-data comparisons to be made.

A variety of climatic states and events in Earth’s history have beenrecorded.Paleoclimatemodelers tendmainly to be interested in abnormal climatic intervals,which feature signif cantly different climates from the present day,and have abundant proxy data to be used alongside.For instance,the mid-Holocene is an ideal period for evaluating the response ofclimatemodelstodifferentseasonaldistributionsofinsolation due to changes in the Earth’s orbital parameters,and the Last Glacial Maximum(LGM)is equally suitable for evaluating the ability of climate models to reproduce extremely cold climates and for understanding how massive ice sheets and lower atmospheric CO2concentrations affect the climate system(Jansen et al.,2007).The last millennium bridges the periods containing proxy data as well as modern instrumental observations and thus provides a unique opportunity for seeking the nature and cause of climate change on decadal to centennial scales,and for distinguishing the effect of anthropogenic activity from natural climate variability,which is of course an issue of great concern for both the social and scientif c communities.Additionally,simulations of the climate of the mid-Pliocene,ca.3 million years before present(Ma BP),knownas themostrecentgeologicalerawhenglobalclimate was signif cantly warmer than at present,will improve our knowledge of the operation of a warmer-than-presentclimate regime,which is likely to be very helpful in evaluating man-made global warming in the future(Jansen et al.,2007).

China is one of the most densely populated areas on Earth.Its natural environments,which are inf uenced greatly by geographic location,topography,and geomorphology,are susceptible and vulnerable to climate change.Statistically, annual meteorological disasters during the 1990s accounted for as much as 3%-6%of the Gross Domestic Product of China,with larger percentages in years that featured significant climate anomalies(Huang et al.,2003).Accordingly, more and more attention is being paid to investigate past, present,and future climate changes in China.It should be noted,however,that although numerous paleoclimate simulations have been reported worldwide in the scientif c literature,few have focused on past climate change in China and adjacent regions,as compared to the North Atlantic,Europe, North America,etc.

Based on reconstructions,considerable effort has been made by Chinese scientists to simulate past climate change and events in East Asia and across the globe for more than two decades. This began with Wang and Zeng(1992a, 1992b),with their simulations of the climate of the LGM and that of 9000 years before present(ka BP).The motivation behind this paper is to present a review of those scientif c activities,with the aim being to facilitate forthcoming paleoclimate simulations,particularly for the East Asian monsoon region.In section 2 we discuss historical climate modeling. Section 3 reports on the progress made in the area of mid-Holocene climate modeling and abrupt climate change over China at 4 ka BP.Simulations of the LGM and the late Marine Isotope Stage 3(MIS3)are reviewed in sections 4 and 5,respectively.Section 6 is devoted to pre-Quaternary climate modeling;specif cally,the mid-Pliocene,the impact of tectonic changes on the late Cenozoic climate,the climatic consequences of the Qinghai-Tibetan Plateau(QTP)uplift, and the East Asian climate transition during the Cenozoic. Finally,some conclusions and perspective are presented in section 7.

2. Historical climate modeling

2.1.Reconstructed historical climate in China

The last millennium is a key period for linking proxy records and instrumental observations,during which time the Earth’s environment and ecology have become increasingly changeable due to the increasing inf uence of human activity. It provides us with a unique opportunity to evaluate the variability induced by anthropogenic and natural factors,which is particularly important for understanding global warming over recent and coming decades.

Based on historical documents,stalagmites,tree rings, lacustrine sediments,ice cores,and so on,effort has been made to reconstruct historical climate change over China (e.g.,Chu,1973;Zhang,1980;Shao et al.,2005;Gou et al., 2008;Zhang et al.,2008;Tan et al.,2009).For the last two millennia,Yang et al.(2002)used multiple proxy records to establish three China-wide temperature series.Five temperature phases were identif ed:a warm stage during 0-240;a cold interval during240-800;a returnto a warm stage during 800-1400;a cold interval during 1400-1920;and the present warm stage from 1920.For the last millennium,Wang et al. (2007)used several kinds of records to reconstruct regionally averaged temperature series for 10 regions of China.After that,temperature series for the whole country were obtained by averaging the regional series in terms of the size of the regions.On the decadal scale,temperature change in eastern China was different from that of western China.In the f rst 400 years,temperature was above normal in the east but near or lower than the normal in the west.Temperature increased by almost 1 K in the west,but only～0.5 K in the east from the 17thto the20thcentury.Warminginthe 20thcenturywasstrongest in western China.Both the Medieval Warm Period (MWP)and Little Ice Age(LIA)occurred in the east,but probably not in the west,although the 17th century was also cold in the west.In addition,the reconstructed thermal contrast in East Asia was strongest(weakest)in the MWP(early LIA),with more(less)precipitation in North China and less (more)precipitation in southern China(Zhou et al.,2011b).

2.2.Last millennium climate modeling in China

There are uncertainties in reconstructions due to the sparse coverage of proxy data and their translation into climate. More importantly,proxy records themselves cannot explain the mechanisms underpinning historical climate change.There are a variety of issues calling for investigation into climate change over the last millennium from the perspective of climate modeling.For example,what was the climateduringtheMWP,LIA,andpresentwarmperiod?Which period had the more signif cant warming:the MWP or the 20th century?What were the mechanisms responsible for historical climate change over China on decadal to centennial scales?

2.2.1.The last millennium climate forcing

For the last millennium,solar variation and volcanic activity are likely to be leading reasons for climate change before the start of the industrial era,while anthropogenicgreenhouse gases and aerosols become important factors for climate change thereafter(Jansen et al.,2007).Solar variation is usually estimated by a combination of observed sunspot numbers and cosmogenic isotope production as recorded in ice cores and tree rings(e.g.,Crowley,2000).Volcanic histories are based on analyses of polar ice cores containing minor dating uncertainty and obvious geographical bias.Meanwhile,there are some differences in the way that models implement records of volcanic activity(Jansen et al.,2007).For atmospheric concentrations of greenhouse gases,both reconstructed and simulated series have been used in simulations (e.g.,Crowley,2000;Jooset al.,2004).Besides thosefactors, orbital insolation is computed with the algorithm of Berger (1978).The effect of land-use and land-cover change and atmospheric aerosol concentrations has also been included in several simulations,although their spatiotemporal evolutions are highly uncertain.Overall,these two anthropogenic factors cause a negative forcing,which tends to offset the effect of greenhouse gas warming(Hansen et al.,1998).

2.2.2.Time-slice simulations for the MWP,LIA,and 20th century

Using solar radiation,volcanic aerosols,and reconstructed vegetation,Liu et al.(2004)performed a set of experiments for the LIA using an atmospheric general circulationmodel(AGCM).Inresponsetodecreasedsolarradiation, annual temperature reduced in China.Temperature decrease was moreobviousin summer(Juneto August throughoutthis paper)than in winter(December to February throughout this paper),which was due to the larger changes in net solar radiation at the top of the atmosphere in summer than in winter. Volcanic aerosols reduced winter temperature,but to less an extent than solar radiation.The synergistic effect of the reduced solar radiation and increased volcanic aerosols had a superposed strengthening impact on temperature decrease in large regions.An increase in vegetation cover gave rise to temperature increase,and vice versa.Meanwhile,reduced solar radiation increased summer precipitation in East Asia, while increasedvolcanicaerosolshadlittle ornoeffectonannual precipitation in most parts of Eurasia.The combined effect of solar radiationand volcanic aerosols led to an increase in summer precipitation averaged over eastern China(20°-40°N,105°-120°E),but a decrease in South Asia.In addition,there was a weak anti-correlation between the Indian monsoonandEast Asian subtropicalmonsoon,becausewhen the Indian monsoon trough enhanced,the western North Pacif c Subtropical High extended westward,reducing the rainfall along the East Asian subtropical front.Precipitation in eastern China increased(decreased)when vegetation cover increased(decreased).Further study is required to establish the reason behind this feature.

Later,six sets of time-slice and equilibrium simulations for the MWP during 1100-1200,the LIA during 1650-1750, andthe20thcenturywereconductedwiththe FlexibleGlobal Ocean-Atmosphere-Land System Model(FGOALS)(Zhou et al.,2008;Zhang et al.,2009;Man et al.,2010;Zhou et al.,2011a;Zhou et al.,2011b).The effect of solar radiation and volcanic activity was found to largely contribute to the warming(cooling)in the MWP(LIA),while an increase in atmospheric greenhouse gas concentrations played more important roles in the 20th century warming,which was consistent with the results of phase 3 of the Coupled Model Intercomparison Project(Zhou and Yu,2006).The MWP warming was evident on a global scale,except for the mid-latitude North Pacif c,and was weaker in magnitude than that in the 20th century.The LIA cooling was also evident on a global scale,with a larger magnitude in the Northern Hemisphere (NH)than in the Southern Hemisphere(SH)and in the high latitudes than in the lower latitudes.Global model-datacomparisonsindicatedthat FGOALS’s performancein simulating the temperature change during the warm periods was better than during the LIA,while model-data consistency in lower latitudes was better than in high latitudes.A comparison of the simulated LIA temperature with proxy data in eastern China showed a high level of consistency.The interannual variabilitymodeoftheEastAsiansummermonsoon(EASM) rainfall during the MWP,LIA and 20th century displayed a consistent pattern.On the centennial scale,the externalmode of the EASM variability driven by effective solar radiation was determined by the change of large-scale land-sea thermal contrast.The EASM was strongest in the MWP but weakest in the LIA.When the EASM was weaker,the monsoon rain belt in eastern China was generally located more southward,with less precipitation in North China and more precipitationin the Yangtze River valleys;namely,a southern f ood/northerndroughtpattern.Globally,therewas moreprecipitation in the MWP and 20th century but less in the LIA. The results corresponded well to the synchronous evolutionof global temperature,which resembled simulations with the ECHO-G(ECHAM4 and the global Hamburg Ocean Primitive Equation coupled ocean-atmosphere model)(Liu et al., 2009a).However,the EASM and precipitation did not vary synchronously with the trend of global temperature.During the last 150 years,for example,although the temperature around the world and in China increased,the EASM and precipitation possessed no detectable trend.

2.2.3.Transient simulations for the last millennium

The Earth system model of intermediate complexity,the McGill Paleoclimate Model-2,was used to simulate climate change during 1000-1800 by Yin et al.(2007).The joint effect of solar variability and volcanic eruptions was found to form the basic pattern of temperature evolution and explain the major characteristics of climate change at the global and northern hemispheric scales,where solar variability was responsible for the long-term trend,with volcanism possibly strengthening or weakening this trend.Based on the same model,Shi et al.(2007)conducted further experiments to evaluate the effect of anthropogenic land-cover change. The biogeophysicaleffect of historical land-coverchange decreasedglobalannualtemperatureby0.09-0.16K(0.14-0.22 K in the NH during the last 300 years),indicating the importance of this factor in climate over the last millennium.

Control and transient simulations for the last millennium using the ECHO-G atmosphere-ocean general circulation model(AOGCM)have also been used to examine regional climate in China.Liu et al.(2005)compared these simulations with reconstructed winter half-year temperature in central eastern China(Ge et al.,2003).The correlation coeff cient between the simulated and reconstructed time series was0.37,whichwasstatisticallysignif cantatthe97.5%conf dence level.The MWP during 1000-1300,the LIA during 1300-1850,and the modern warm period after 1900 all appeared in both the simulated and reconstructed temperature.The simulations for the LIA and the 20th century were in good agreement with reconstructions and/or observations. In particular,both the simulated and reconstructed temperature reached their minima in the Maunder sunspot minimum during 1670-1710.For the MWP,however,signif cant discrepancies existed between the simulation and reconstruction,which might partly ref ect the degrading quality of reconstructions(Ge et al.,2003)and the model’s def ciency in initialization.Overall,variations in solar radiation and volcanic activity were found to be the main factors for temperature change over China before the 20th century,while variations in atmospheric greenhouse gas concentrations played the most important role in the 20th century warming,which was in line with the aforementioned time-slice simulations.

Monsoon precipitation affects about two thirds of the world’s population.Its response to external and anthropogenic forcings during the last millennium has also been examined based on ECHO-G simulations(Liu et al.,2009a). The monsoon precipitation domain over the globe was def ned by the regions in which the annual range of precipitation exceeds 2 mm d-1and the local summer precipitation exceeds 55%of annual rainfall(Wang and Ding,2008). The strength of global monsoon precipitation was found to undergo a signif cant variation with a prominent quasibicentennialoscillation.It was weak in the LIA,but strong in the MWP.Before the industrial period,effective solar radiation variations reinforced the thermal contrasts both between the oceanandlandandbetweenthe NH andSH,resultingin a millennium-scalevariation and quasi-bicentennialoscillation in the global monsoon index.The prominent upward trend in global monsoon precipitation in the last century and the remarkable strengthening of the global monsoon during 1961-90 appeared unprecedented and were due possibly in part to the increase of atmospheric greenhouse gas concentrations. The global monsoon in the last 30 years had a different spatial pattern from that in the MWP,suggestingthat greenhouse gas and solar/volcanic forcing might have different impacts on global monsoon precipitation.Global monsoon strength was closely related to the temperaturedifference between the NH and SH,andon thecentennialscale it respondedmoredirectly to the effectivesolar forcing than the concurrentforced response in global temperature.

Based on the ECHO-G simulations,it was also found that the centennial-millennial variation of the EASM precipitation was essentially a forced response to the external radiative forcing over the past millennium(Liu et al.,2011). The strength of the response depended on latitude,and the spatial structure of the centennial-millennial variation differed from the interannual variability that arose primarily from the internal feedback of the climate system.On the millennial scale,extratropical and subtropical precipitation was generally strong(weak)in the MWP(LIA).Tropical rainfall was insensitive to the effective solar radiation forcing but responded signif cantly to modern anthropogenic radiative forcing.On the centennial scale,the variation of extratropical and subtropical rainfall also tended to closely follow the effective solar radiation forcing.The forced response featured in-phase rainfall variability between the extratropics and subtropics,which was in contrast to the anti-correlation on the interannualscale.As such,the proxydata in extratropical East Asia could more sensitively ref ect EASM rainfall, andtheMei-yuandnorthernChinarainfallprovideda consistent measure for EASM strength on the millennial scale.Further simulations indicated that EASM circulation during the MWP was stronger than during the LIA as a result of landsea thermal contrast change caused by the effective radiative forcing(Man et al.,2012;Man and Zhou,2014).There was a coherent cooling over East Asian continent and the tropical ocean after large volcanic eruptions,and stronger cooling over the mid-high latitudes of the East Asian continent than overthetropicaloceanled toa reducedland-seathermalcontrast and hence a weak EASM circulation(Man et al.,2014).

In addition to analysis of available simulations,Peng et al.(2009)used the Community Climate System Model (CCSM)version 2.0.1 AOGCM to simulate climate change over the last millennium.The simulated temperature across the whole of China and in eastern China correlated to someextent with reconstructions,while simulated precipitation in eastern China and in the middle and lower reaches of the YangtzeRivervalleysdisplayedsomesimilaritieswith reconstructions for certain periods of time.Both simulations and reconstructions indicated that the 20th century warming was anomalous in a long-term context.The wet and dry conditions appeared alternately in eastern China in the MWP.Dry conditionsdominatedintheLIA,whereaswetconditionsprevailed after 1890.The correlation between the simulated and reconstructedprecipitationwas betterinthemiddleandlower reaches of the Yangtze River valleys than in eastern China, especially before 1850.Regional climate differences were present in eastern China in the last millennium,and there were no f xed modes of climate change,such as warm-wet or cold-dry.Temperature and precipitation in eastern China were controlled mainly by the changes in effective solar radiation and volcanic activity,while atmospheric greenhouse gas concentrationsplayed a leadingrole in the rapidwarming of the past 150 years.Shen et al.(2009)used these simulations to further examine summer precipitation variability in eastern China.Model-data comparisons suggested that the centennial oscillation might be linked to the f uctuation of solar forcing,and the decadal oscillation could be associated with internal variability of the climate system.The increased frequency of the southern f ood/northern drought pattern in eastern China over the last two decades was unusual over the past fve centuries.

FGOALS reproduced the MWP,LIA,and 20th century warming reasonably,with enhanced warming over northern high-latitude continents(Man and Zhou,2014).Model-data consistencywas loweronregionalscales thanonhemispheric scales.Different from signif cant global signals in the 20th century,climate changesduringthenatural-forcing-dominant periods were mainly manifested in the NH;and total external forcings explained about half of the climate variance and signif cantly impacted the evolution of atmospheric oscillations during the last millennium,especially at decadal scales(Man and Zhou,2011;Zhang et al.,2013a).FGOALS’s sensitivity to naturalforcingsis generallyweak,leadingto a weakMWP (GuoandZhou,2013).The modelsensitivityin the industrial era is higher than that of the pre-industrial period.Both the weaker negative net cloud feedback and stronger water vapor feedbackin the industrial era than in the pre-industrialperiod favor higher model sensitivity and thus a reasonable simulation of the 20th century warming.

2.3.Perspective

The aforementioned simulations generally showed that solar radiation and volcanic activity accounted for large parts of the MWP and LIA climate,while atmospheric greenhouse gas concentrationsplayed the most important role in the 20th century warming both in China and across the globe.The warming in the MWP was likely weaker than that in the 20th century over China.The effect of land-cover change could also be important for historical climate change,particularly in the last 300 years.Since both similarities and differences have been noticeable when simulations were compared to reconstructions in China in the context of the last millennium (e.g.,Liu et al.,2005;Peng et al.,2009;Man et al.,2012; Man and Zhou,2014),in future work it is necessary to use climate models to perform transient simulations and evaluate the effect of not only solar radiation,volcanic eruptions,and atmospheric greenhouse gas concentrations,but also anthropogenicaerosolsandland-useandland-coverchange.Uncertainties of last millennialclimate simulationareresultedfrom both the specif ed external forcing data and the model sensitivity to natural/anthropogenic forcings.The millennial climate simulation driven by different external forcing data including effective solar radiation and volcanic aerosol should be compared.Multi-model inter-comparison should also be performed,andthesensitivities ofclimatemodelstothenatural/anthropogenic forcing should be studied.Where regional climate is concerned,the horizontal resolution of global climate models is too coarse to perform a transparent modeldata comparison.Regional climate models should be emphasized in that area,particularly for historical climate events such as the MWP and LIA.

3. Mid-Holocene climate modeling and abrupt climate change in China at 4 ka BP

3.1.Mid-Holocene climate modeling in China

3.1.1.Reconstructed mid-Holocene climate in China

The mid-Holocene was a typical interglacial period atca.6 ka BP,and many efforts have been devoted worldwide to investigating the response of climate models to the different seasonal distributionsof incominginsolation for that time (Jansen et al.,2007).Based on records from pollen,fossil remains of plants and animals,paleosols,lacustrine sediments, ice cores,stalagmites,andNeolithicarchaeologicalevidence, the mid-Holocene megathermal was inferred to occur at 8.5-3.0 ka BP,with stable warmer and wetter conditions duringca.7.2-6.0 ka BP over China(Shi et al.,1993;Jiang et al., 2012,2013b).For that period,the deviation of annual temperaturefromthepresentdaywas estimatedroughlyas 1K in South China,2 K in the Yangtze River valleys,3 K in North China and Northeast China,and 4-5 K on the southern QTP. Moreover,winter warming was stronger than annual warming;summermonsoonintensif ed;wintermonsoonweakened in East Asia;vegetation zones shifted northwestward;and higherlevelsofinlandlakesoccurredinInnerMongolia,Xinjiang,Qinghai,and Tibet,implying wetter climates(e.g.,Shi et al.,1993;Qin and Yu,1998;Yu et al.,2003a).These proxy data provide a benchmark for mid-Holocene climate modeling and an opportunity for examining the dynamic mechanisms underpinningthe changes.

3.1.2.Mid-Holocene boundary conditions

Under the framework of the Paleoclimate Modeling Intercomparison Project(PMIP;Joussaume and Taylor,1995), boundaryconditionsfor AGCMs are composedof changes in the Earth’s orbital parameters and atmospheric CO2concentrations.The former led to an enhanced(reduced)seasonalcycle of insolation in the NH(SH),by about 5%(Berger, 1978).The latter were set to 280 ppmv from the present value of 345 ppmv.For AGCMs coupled with a slab ocean model in PMIP Phase 1(PMIP1),sea surface temperatures (SSTs)were computed.Within PMIP Phase 2(PMIP2)and Phase 3(PMIP3),apart from the Earth’s orbital parameters beingthesameas inPMIP1,atmosphericCO2concentrations were held at 280 ppmv both for the pre-industrial period and mid-Holocene.Atmospheric CH4concentrations were set to 650 ppbv for the mid-Holocene and 760 ppbv for the preindustrial period.SSTs and sea ice extent were computed in PMIP2/3.Other aspects of PMIP2/3 boundary conditions were kept the same as in PMIP1.Besides the aboveboundary conditions recommended by PMIP,reconstructed vegetation (e.g.,Shi et al.,1993;Yu et al.,2000),rather than present day vegetation,has been used in several simulations in order to evaluate the effect of vegetation on mid-Holocene climate in China.

3.1.3.Mid-Holocene climate modeling

Since the work of Wang and Zeng(1992a),considerable effort has been made to simulate the mid-Holocene climate in China.Wang(2000)indicated that,during mid-Holocene summers,temperatures rose by 1-4 K in much of the northern continents,the African and Asian monsoons intensif ed signif cantly,and precipitation increased by 10%-20%over China.When changes in the mid-Holocene Earth’s orbital parameters alone were considered,both summer warming and winter cooling were signif cant in East Asia(Chen et al., 2002).In this region,the mid-Holocene summer temperature was～2 K warmer in areas south of 40°N,whereas winter temperature was～2 K colder than at present.When viewed from multiple climate models,36 PMIP1/2 models reproduced colder-than-baselineannual temperature,with an average cooling of 0.4 K,over China,while seasonal temperature changescloselyfollowedchangesin incomingsolarradiation at the top of the atmosphere over the country,with a summer warming but a winter and spring cooling(Jiang et al.,2012). Thirty-six PMIP1/2/3 models indicated that mid-Holocene annual precipitation,evaporation,and net precipitation were 3%,1%,and 7%more than the baseline period,respectively; and seasonally,all three variables decreased in boreal winter and spring but increased in boreal summer and autumn on the national scale(Jiang et al.,2013b).For CCSM3 AOGCM simulations,both the East Asian winter and summermonsoonsstrengthenedinresponsetoanincreasedlandsea thermal contrast,while the changes of boreal spring and summer tropospheric thermal contrasts between Asia and the North Pacif c played crucial roles in atmospheric circulation and precipitation changes over the Asian-Pacif c region during the mid-Holocene(Zhou and Zhao,2009,2010,2013). The mid-Holocene EASM strengthened by 32%across 28 PMIP1/2/3 models with a demonstrable ability to simulate the modern EASM climatology(Jiang et al.,2013a).

Regional climate models with high horizontal resolutions have also been used to examinethe mid-HoloceneEast Asian climate(e.g.,Zheng et al.,2004,2007;Liu et al.,2009b). For example,Zheng et al.(2007)investigated mid-Holocene changes in hydrological processes in eastern and western China.It was found that wetter and warmer climates dominated on the QTP during that period.The increased amount of water vapor arriving on the plateau came mostly from its western boundary,and theincrease inrunoffstemmedmainly from increased precipitation.The mid-Holocene increase in precipitation and runoff in eastern China was closely related to strengthened Asian summer monsoon,which led to increased vapor coming into the area through its southern boundary.

In response to warmer and wetter climates,the mid-Holocene vegetation conditions differed largely from the presentdayinChina.Ingeneral,tropicalbroadleaf-evergreen trees extended northward and a large area was covered by forests in eastern China,while forests on the QTP extended towards higher altitudes(Shi et al.,1993;Yu et al.,2000). Wang(1999a)revealed that changes in vegetation and associated soil characteristics further enhanced monsoon precipitation in China during mid-Holocene summers,as they decreased surface albedo and,in turn,increased land surface temperature,which reinforcedmonsoonvia an increased land-sea thermal contrast.With the same mechanism,reconstructedvegetationwas foundto lead to a warmingof 1.0-2.0 K,0.5-1.0 K,and 0.5-1.5 K for the mid-Holocene summer, winter,and annual mean temperatures in East Asia,respectively(Chen et al.,2002).Accordingly,the thermal contrast between the East Asian continent and western North Pacif c enlarged and caused a stronger EASM as described by Wang (1999a).The effect of vegetation in the Asian and African monsoon areas was also corroborated by Wang(2002),in which a coupled atmosphere-vegetation model was used to simulate the mid-Holocene climate.By contrast,averaged across six pairs of PMIP2 coupledmodels with and without a dynamic vegetation model,interactive vegetation was found to have little effect on mid-Holocene annual and seasonal temperatures in China(Jiang et al.,2012),which was also supportedby recent simulations(Tian and Jiang,2013).Note that their spread in vegetation-induced temperature changes between each of the six pairs of models implied a level of uncertainty in the mid-Holocene vegetation effect over the country.Recently,interactive vegetation was found to affect the interannual and interdecadal variability of the Asian summermonsoonin the mid-Holoceneandin the present day (Li et al.,2009).In strong interannual or interdecadal South Asian summer monsoon years,dynamic vegetation tended to keep the intensity of westerly wind over South Asia in the lower troposphere for both periods.However,in strong interannual or interdecadal western North Pacif c monsoon years,dynamicvegetationtendedto reducethewesterly wind and the south-north cross-equator transport over the tropical western Pacif c in the lower troposphere for both periods.This suggested that the impact of dynamic vegetation was moreobviousonthewesternNorthPacif c monsoonthan on the South Asian monsoon.In other words,it implies the impact of dynamic vegetation on the intensity of interannual circulations is region-dependent.

Since the effect of ocean dynamics was neglected in PMIP1,an asynchronously coupled AGCM and an oceanic general circulation model(OGCM)were used to quantify the role of orbital forcing and the ocean in forming the mid-Holocene East Asian climate(Wei and Wang,2004).With referenceto the simulationby theAGCM alone,moreprecipitation and stronger monsoon were reproduced in East Asia during summer,while winter temperatures rose over China due to the large thermal inertia of the ocean.It was revealed that solar radiation changes increased the convergence of atmospheretowardtheland,andSSTchangestransportedmore moisture to the sea surface atmosphere during mid-Holocene summers.Their synergisticeffect on East Asian precipitation was much stronger than the sum of their respective effects. Later,FGOALS simulations reproduced an enhancement of the Asian monsoon,which resulted from an increased landsea thermal contrast during mid-Holocene summers(Zheng and Yu,2009).In the East Asian monsoon region,the vertical and horizontal temperature gradient changes gave rise to a weakening and southward shifting of the subtropical westerly jet,which favored convergence in the upper troposphere and divergence in the mid-troposphere in North China.As a result,monsoonal rainfall was suppressed in North China but enhanced in South China.In earlier simulations with the Fast Ocean Atmosphere Model,SSTs were found to increase in the western North Pacif c due to orbitally induced insolation changes,which reduced land-sea thermal contrast and,hence,monsoon circulation in East Asia during mid-Holocene summers(Liu et al.,2003b).Based on the same model,Li and Harrison(2008)revealed that ocean feedback dampened orbitally induced increases of summer precipitation in southeastern China,consistent with Liu et al.(2003b). These results were opposite to the aforementioned positive feedback of the ocean on the EASM(Wei and Wang,2004; Zheng and Yu,2009),implying an uncertainty in the effect of the ocean on the Asian climate during the Holocene, as discussed by Liu et al.(2003b)and Tian and Jiang (2013).How,and to what extent,the East Asian monsoon responded to mid-Holocene orbital forcing calls for further studies.

Change in the amplitude of El Ni?no and its link to the mean climatology of the mid-Holocenewere examined using PMIP2 AOGCM simulations by Zheng et al.(2008).Most simulations reproduced the modern large-scale features of the tropical Pacif c and ENSO variability.El Ni?no amplitude was shown to be an inverse function of the mean trade wind within the Ni?no4 region and the seasonal cycle relative strength,andto have a linearrelationshipwith seasonal phase locking.All the AOGCMs reproduced a consistent El Ni?no weakeningin the mid-Holocene,consistent with previous experiments(Liu et al.,2000).The associated mechanism was that,while the NH received more insolation in summer,the Asian summer monsoon strengthened and then led to an enhancement in Walker circulation,as discussed by Liu et al. (2000).Easterlies prevailing in the central eastern Pacif c inducedan equatorialupwelling,which dampenedthe development of El Ni?no.

3.1.4.Mid-Holocene model-data comparisons in China

Responding faithfully to the imposed mid-Holocene negative radiative forcing in China as derived from changes in the Earth’s orbital parameters and atmospheric greenhouse gas concentrations,35 of the 36 PMIP1/2 models reproduced colder-than-baselineannual and winter temperatures over the country(Jiang et al.,2012).By contrast,as discussed above, a variety of proxy data indicated warmer annual and winter temperatures in the mid-Holocene.Taken together,the results of 36 PMIP models were opposite to the multi-proxy records.On the whole,interactive vegetation had little effect on mid-Holocene temperature over China according to the six pairs of PMIP2 coupled models.Interestingly,an AGCM simulation with reconstructed vegetation was closer to proxy data.In particular,the simulated winter temperature increase of～0.5-1 K was more consistent with the value of～3 K suggested by proxy data in the mid-Holocene(Chen et al., 2002).On the other hand,36 PMIP1/2 models indicated that an interactive ocean gave rise to an additional warmingof 0.5 K in winter and 0.7 K in boreal autumn in China,and hence the annual and winter temperatures of coupled models were inbetteragreementwithproxydatathanthoseofatmospheric models during the mid-Holocene(Jiang et al.,2012).In summer,the results of 12 PMIP2 AOGCMs were consistent with reconstructed temperature in eastern China,but they failed to capture the strongest warming on the southern QTP(Wang et al.,2010).Comparedwith the wetter-than-presentclimates derivedfromtherecordsat 64 outof 69sites across China,36 PMIP1/2/3 models agreed qualitatively with the multi-proxy data in most parts of China,except Xinjiang and the areas between the middle and lower reaches of the Yangtze and Yellow River valleys,where drier-than-baseline climates were obtained from the models(Jiang et al.,2013b).

3.2.Abrupt climate change in China at 4 ka BP

3.2.1.Evidence of abrupt climate change in China at 4 ka BP

A signif cant cold and dry abrupt climate change event has been reported to have occurred atca.4 ka BP,which was possibly related to the collapse of ancient civilizations in the alternation of the ancient cultures in China(Wang et al.,2004).Proxy records at 80 sites,covering almost the entire Chinese territory,illustrate that humidity reduced along a band stretching from Southwest to Northeast China at that time,whereasarchaeologicalevidenceindicatedthatf ooding was prominent in the lower reaches of the Yangtze River valleys(Wang et al.,2009a).Meanwhile,dramatic environmental changes occurred in the western Chinese Loess Plateau and correspondedwith substantial changes in human demography atca.4 ka BP(An et al.,2005).A rapid climate transition from wet to dry led to a period of ecological devastation between 4.1 and 3.6 ka BP.The sudden reduction in the number of archaeological sites during that period—a reduction in the total number of sites and a contraction of the areal distribution of sites—pointed to a declining agricultural productivity associated with widespread aridif cation beginningatca.4 ka BP.

3.2.2.Possible cause of abrupt climate change in China at 4 ka BP

Holocene abrupt climate changes were mostly characterizedbycoldeventsin theNorthAtlantic.Sedimentrecordsin the North Atlantic have proventhe occurrenceof a cold event at 4 ka BP(Wang,2009).It was therefore hypothesized that the abruptdroughteventin China at 4 ka BP may relate to the cold event in the North Atlantic(Wang et al.,2009a).Wang et al.(2004)conducted sensitivity experiments to simulate global climate responses to the SST forcing in the North Atlantic,the geographical distribution of which stood for a typical weakeningof the thermohalinecirculationin the Atlantic Ocean.Results showed a dropof temperaturein northernEurope,the northern central East Asia,and northern East Asia, and a signif cant reduction of precipitation in East Africa,the Middle East,the Indian Peninsula,the Yellow River valleys, and North China.These results seem to support the hypothesis that coldness and aridif cation in China at 4 ka BP was, at least partly,caused by the weakening of the thermohaline circulation.

Meanwhile,a set of numerical experiments revealed that changes in Earth’s orbital precession could also have signif cantly affected summer precipitation in China during the Holocene(Wang et al.,2008b).More specif cally,summer precipitationincreasedin the middle and lower reaches of the Yangtze and Yellow River valleys but deceased signif cantly in most parts of the rest of the mainland China,the pattern of which resembled the reconstructed environment at 4 ka BP. Taken together,the drought event in North China at 4 ka BP may have resulted from both abrupt climate change in the North Atlantic and the precession effect.

3.3.Perspective

The reconstruction and simulation of the Holocene climate are important for exploring natural variability and external forcing of the climate system on the orbital scale,and have become a central theme both in climate modeling and proxydatacommunities.Asis wellknown,changesinorbital forcing are the main reason for the mid-Holocene climate having been signif cantly different from that of the present day.Based on records and simulations in China,we would like to emphasize that an interactive ocean appears to modify the response of climate models to mid-Holocene orbital forcing and give rise to more reasonable results in China as a whole.However,ocean feedback on the Asian monsoon climate still remains an open question,particularly in terms of theunderlyingdynamicmechanism.Inaddition,basedonthe earlier AGCM simulations,reconstructed vegetation appears to decrease surface albedo and lead to a surface warming in China;whereas,based on the six pairs of PMIP2 coupled models,interactive vegetationhas little effect overall on mid-Holocene climate over the country.The extent to which vegetation varied and how it interacted with climate during that period should be specif cally investigated in the context of cause and effect.Furthermore,it should be kept in mind that the results of the 36 PMIP models were opposite to warmerthan-present annual and winter climate conditions as derived from proxy data.At the moment it is unclear whether this inconsistency arises from the models,from the proxy data,or from both sides.If the interpretations of those proxy data are correct,a big question is why the PMIP models failed to reproduce the mid-Holocene East Asian climate.On the other hand,the PMIP models seem to do what is asked of them in termsofnegativeradiativeforcingin China.Is it possiblethat the paleoarchives are not actually able to record the information that is equivalent to temperature in the models?Interestingly,mid-Holocene temperature was reconstructed to be colder in part of China in the works of Guiot et al.(2008) and Bartlein et al.(2011).More reconstruction work using multiple proxies and methods is therefore required to reduce the uncertainty of proxy data.Comprehensive comparisons between multiple climate models and multiple proxy records will ultimately reveal the nature and underlying mechanisms of mid-Holocene climate change in China.

The abrupt climate change that took place in China atca.4 ka BP was likely synchronous with that which occurred in the North Atlantic and the change in Earth’s orbital precession.More simulations are needed to explore the reasons behind the event.In addition,transient simulations of the Holocene climate,together with its comparison with proxy data in China,remain an open f eld for climate modelers.Such studies will greatly improve our knowledge on Holocene climate change and events,such as those that occurred atca.8.2 ka BP andca.4 ka BP,and on the dynamic mechanisms operating at the regional and global scales during the Earth’s recent history(e.g.,Jin et al.,2005,2009;Liu et al.,2009c).

4. LGM climate modeling in China

4.1.Reconstructed LGM climate

The LGM refers to the time when the ice sheets during the last glacial period were at their maximum extent,approximately 21 ka BP.This extreme period persisted for several thousand years,during which time global climate was very different to that of today(Jansen et al.,2007).Based on various proxies over China,the LGM generally featured cold and dry climates,with a spatial variability(e.g.,Qin and Yu, 1998;Jiang et al.,2011).In western China,higher lake levels and fresher water than today ref ected wetter conditions, whereas the opposite situation was recorded in most parts of eastern China at that time(Yu et al.,2003a).On the QTP,the LGM temperature was～7 K colder than at present,while precipitation was only 30%-70%of the current level(Shi et al.,1997).In tropical areas of China,temperatures were 5-8 K colder than the present day,while precipitation was higher(Zheng and Guiot,1999).Clearly,these reconstructions lay f rm foundations for LGM East Asian climate modeling.By comparing simulations with reconstructions,it is possible to uncover the realistic characteristics of the LGM climate.

4.2.LGM climate modeling

4.2.1.LGM boundary conditions

The LGM was characterized by great changes in surface boundary conditions and atmospheric greenhouse gas concentrations,but minor changes in orbital forcing.Among these include changes in ice sheet extent and topography (Peltier,1994,2004),atmospheric CO2concentrations,and the Earth’s orbital parameters(Berger,1978).SSTs and sea ice extent are usually prescribed as in CLIMAP(Climate: Long range Investigation,Mapping,and Prediction)project members(1981)for AGCM experiments,but they are computed by slab ocean models or OGCMs in coupled model experiments.Variations in atmospheric CH4and N2O concentrations were considered in part of the PMIP experiments. In East Asia,LGM vegetation compiled from pollen records has also been used in several experiments.

Infact,thereareuncertaintiesintheearlierboundaryconditions recommended by PMIP1.In particular,the SSTs established through the CLIMAP project were considered too high for the LGM,especially in the tropical Pacif c and Atlantic oceans(e.g.,Farrera et al.,1999;Mix et al.,1999).For example,the LGM SSTs compiled by Wang(1999b)differed largely from those of the CLIMAP project in the western Pacif c.The former described a colder tropical western Pacif c west of130°E,with anannualdecreasein SSTs of～2K,and a warmer ocean current near Japan.In particular,the LGM annual mean SSTs established by Wang(1999b)for the tropical western Pacif c west of 140°E were 2-4 K lower than at present.To what extent these reconstructed SSTs potentially affected the East Asian climate remains of interest.

4.2.2.LGM climate in China from global climate models

4.2.2.1.LGM climate in China

In the early 1990s,a two-level AGCM,constructed at the Institute of Atmospheric Physics(IAP)under the Chinese Academy of Sciences,was used to investigate the LGM July climate with CLIMAP boundary conditions and atmospheric CO2concentrations of 200 ppmv(Wang and Zeng, 1992b).The later version of that model,hereafter referred to as IAP-AGCM,was also used to simulate the LGM climate within PMIP1(Jiang et al.,2003).It reproduced colder and drier climates,with a global annual temperature of 5.3 K less,and a terrestrial precipitation level of 29%less,than at present.In East Asia,both annual temperature and precipitation reduced during that period,which were basically consistent with proxy data(Shi et al.,1997;Farrera et al.,1999; Liu et al.,1999).Another set of AGCM experiments indicated that the LGM featured drier conditions in the east and wetter conditions in the west,while the East Asian monsoon weakened signif cantly(Chen et al.,2000,2001).Using the same AGCM,Liu et al.(2002a)furthershowed that the LGM annual temperature decreased by 2-13 K in China.Summer and annual precipitation in eastern China were only～50% of current levels,whereas precipitation differing little or not at all was found for western China.

Based on the Community Climate Model version 3 (CCM3)AGCM experiments with the SSTs reconstructed by Wang(1999b)for the tropical western Pacif c and reconstructed vegetation,it was shown that winter monsoon strengthened notably in northern China,while summer monsoon weakened notably in the South China Sea and southern China during the LGM(Zhao et al.,2003).For that period, annual temperature and precipitation reduced in East Asia, with the greatest decrease in precipitation in eastern Tibet, on the Chinese Loess Plateau,and in northern China,causing surface soil to lose water and become dry.On the central QTP,surface soil lost less water and became wetter.This process may explain why the LGM water levels in the lakes of the central QTP were higher than at present(Yu et al., 2003a).Owing to an increase in snowfall during the LGM, the depth of snow coverincreased remarkablyin the southern QTP,which provided favorable conditions for the expansion of local glaciers.

A comparison of tropical atmospheric heating and circulationpatternsbetweenthe LGMandthepresentdaywas carried out by Zhao et al.(2004).The largest decrease in atmospheric heat occurred in the tropical regions stretching from the Bay of Bengal to the central Pacif c Ocean,while there werenosignif cantchangesathigherlatitudes.Thisindicated that atmospheric heat in the tropics showed a stronger response to the boundary condition changes between the LGM and the present day than it did at higher latitudes,since the tropical convection amplif ed the response of the atmosphere in the former case.Because of this tropical heat change,the LGM Walker circulation,the transverse monsoon,and the EASM were weaker than at present.

Besides experiments from individual climate models,the results of 25 PMIP1/2 models have also been used to examine the LGM regional climate of China(Jiang et al.,2011). Compared to the baseline climate,annual temperature was decreased by 2-7 K in China,with an average of 4.5 K,for the 25-modelensemble mean.The LGM annualprecipitation and evaporation were 5%-40%less than the baseline levels based on the results of 15 climate models that were selected for their ability to simulate the modern precipitation climatology.Both the geographical distribution and magnitude of changes in temperature,precipitation,evaporation,and effective precipitation varied with the seasons and with the models,particularly at the sub-regional scale.In contrast to the conclusions drawn from sparse proxy data,the intensity of the East Asian winter monsoon during the LGM,as measured by regionally averaged meridional wind speed at 850 hPa,was found to vary both in sign and magnitude,with reference to the baseline period,across the PMIP simulations (Jiang and Lang,2010).It weakened by 4%for the 21-model ensemble mean and by 15%for the ensemble mean of 14 coupled models.At the sub-regional scale,the LGM winter monsoon strengthened north of～30°N but weakened south of this region in East Asia.During LGM summers,all of the 14 models chosen for analysis consistently simulated a weaker-than-baseline East Asian monsoon,with an average weakening of 25%.Changes in zonal and meridional landsea thermal contrast across the regions of concern were re-sponsibleforthosechangesinthe LGMEast Asianmonsoon.

4.2.2.2.Effects of western Pacif c SSTs and an interactive ocean

The effect of LGM SSTs in the tropical western Pacif c, as reconstructedrespectivelybyWang(1999b)andCLIMAP, was evaluated using the CCM3 AGCM by Zhao et al.(2004). The latter led to larger changes in atmospheric circulation in the tropics and high northern latitudes,reproducinga weaker transverse monsoon and a larger seasonal variation of the Walker circulation,with lower temperature in the high northernlatitudesandtheArcticandlowertemperatureinthecoldest monthin Europe.Winter temperaturewas reducedby 4 K in Europe,which was closer to pollen-based reconstructions. Additionally,there was a smaller effect of the western Pacif c SSTs on temperature in the SH during the LGM.

The Asian summer monsoonwas also sensitive to the differences in LGM SSTs(Sui and Zhao,2005).In response to warmer summer SSTs in the tropical western Pacif c(Wang, 1999b),theSouthAfricahighandHadleycirculationoverthe South Indian Ocean intensif ed,with a stronger zonal monsoon circulation in the Indian monsoon areas.Accordingly, the Indian summer monsoon intensif ed,with a strengthened level of water vapor transportation and more precipitation in the Indian monsoon areas.In the East Asian monsoon region,the Wang(1999b)SSTs caused a weaker Australian high,with a weakened cross-equatorial current in East Asia and zonal monsoon circulation north of 20°N.These features demonstrated a weaker East Asian subtropical continental monsoon and a stronger South China Sea monsoon.Moreover,Wang et al.(2009b)examined the effect of LGM SSTs in the tropical Atlantic and eastern equatorial Pacif c,reconstructed respectively by Mix et al.(1999)and CLIMAP,on the Pacif c convergence zone.It was found that the different constructions of the tropical SSTs in the Pacif c and Atlantic oceans caused large uncertainties in simulating the LGM climate.Thus,resolving the apparent disagreements among the different SST reconstructions is necessary.

Within the PMIP1/2,annual surface cooling over China was stronger in coupled models than in atmospheric models (Jiang et al.,2011).Less surface cooling in the latter was,at least partly,attributed to the small reductions in the CLIMAP reconstructed SSTs in the oceans adjacent to the East Asian continent.In the western North Pacif c,for example,the LGM changes in regionally averaged annual SSTs from the experiments of six PMIP2 AOGCMs were colder than the CLIMAP SSTs used as boundary conditions of the PMIP1 AGCMs.Accordingly,the LGM annual surface cooling over China was larger in the AOGCM experiments,because the corresponding colder SSTs in the western North Pacif c gave rise to larger losses of surface heat in the East Asian region during warm months and smaller gains of surface heat during cold months.Since the simulations of coupled models are in better agreement with proxy estimates than those of atmospheric models,interactive ocean appears to be an important component of the LGM climate system in the East Asian monsoon region.

4.2.2.3.Effect of vegetation

A series of sensitivity experiments revealed that reconstructed vegetationcouldexertstrongeffects on the LGM climate in China(Chen et al.,2000,2001;Yu et al.,2001;Liu et al.,2002a);in particular,on the QTP,changes in vegetation increased the differences in temperature,precipitation,and effective precipitation between winter and summer.Based on global paleovegetation compiled from pollen records,the simulationsundertakenbyYuet al.(2003a)reproducedlower temperature and precipitation in eastern China,with positive precipitation in western China more extensively.In general, the effect of vegetation contributed to an increase in both the aridity in eastern China and the humidity in western China during the LGM.Compared to proxy data,additional climate change due to vegetation generally reduced model-data discrepancies in East Asia.

The inf uence of reconstructed vegetation(Yu et al., 2000)and associated soil characteristics on the LGM East Asian climate was further examined by Jiang et al.(2003). Sparser-than-present paleovegetation enlarged regional surface albedo,and hence temperature decreased.As a result, model-data discrepancies in temperature were partly reconciled.Using different vegetation reconstructions in China was found to cause large uncertainties in simulating the East Asian summer climate(Han et al.,2009).Relative to the present day,the degradation of vegetation during the LGM over China increased local surface temperature during summer,strengthening the thermal contrast between the East Asian continent and the adjacent oceans.Accordingly, the summer southwest monsoon strengthened during the LGM.Interms ofthe climatology,whensouthwesterlywinds weaken(strengthen),the rain belt in front of the maximum southwesterly wind center often stays in a more southward (northward)position,leading to more(less)rainfall in southeastern(northern)China(Zhaoet al.,2007,2010).Thus,corresponding to the strengthened southwesterly winds caused by thevegetationchange,there is an increasein summerrainfall overnorthernChina and a decrease in rainfall over southeastern China.In this context,reconstructingmorebelievable vegetation over East Asia is important to reduce the uncertainties due to vegetation.

IAP-AGCM and its asynchronously coupled system with an equilibrium terrestrial biosphere model were used to further investigate vegetation and soil feedbacks during the LGM(Jiang,2008).Thesimulatedvegetationdifferedlargely from the present day,and global vegetation cover tended to be reduced overall in adaptation to colder and drier climates. Vegetation feedback induced an annual temperature decrease of 0.31 K,mainly through changes in surface albedo,on the LGM’s ice-free continental areas.Additional soil feedback reinforced vegetation-induced cooling through surface albedo in the high latitudes of Eurasia and from the eastern Middle East eastward to the Indian Peninsula.In the tropics, a terrestrial annual cooling of 0.45 K was derived from vegetation and soil feedbacks.They partly reduced model-data discrepancies in Central Africa,the Indian Peninsula,South China,North Australia,etc.Meanwhile,inter-model com-parisons have also shown that there were large uncertainties with respect to the LGM vegetation feedback,particularly at theregionalscale(Jiang,2008).Intheexperimentsofthe UK Meteorological Unif ed Model run at Bristol University,vegetation feedback was found to give rise to an annual cooling of 2 K,with a seasonal and regional variability,over China during that period(Jiang et al.,2011).In this regard,the extent to which vegetation feedback behaves in fully coupled atmosphere-ocean-vegetationmodels needs to be further explored.

4.2.3.LGM East Asian climate from regional climate models

Using a regional climate model nested within an AGCM, Qian et al.(1998)simulated the LGM July climate of East Asia.Later,the regional climate model RegCM2,with a 120 km horizontal resolution,was used to reproduce the LGM East Asian climate(Zheng et al.,2003,2004).Those experiments provided more details of the East Asian climate, with respect to global models,particularly in understanding the processes behind the LGM East Asian monsoon changes. Based on the improved version of RegCM2 with a 60 km horizontal resolution,Ju et al.(2007)further examined the LGM East Asian climate,in which the 12-hourly updated lateral boundary conditions were provided by IAP-AGCM. Their simulations indicated that LGM annual temperatures were 2-4 K colder overall than at present over the East Asian continent,with the largest decrease of～8 K in the vicinityof the current coastal areas,where land was exposed due to sea levellowering.Comparedtotheresults ofIAP-AGCMalone, RegCM2 simulations were more consistent with proxy data in East Asia,especially in central-easternand southernChina where RegCM2 simulated a colder LGM climate.Annual precipitation decreased by～60%in the drier-than-present areas of China in RegCM2,which was also closer to proxy records than in IAP-AGCM.

Meanwhile,a 90 km horizontal resolution mesoscale model was nested within an AGCM to evaluate the effect of the LGM boundaryconditionson the East Asian climate(Liu et al.,2007a,2007b).Changes in atmospheric CO2concentrations were found to have a signif cant inf uence in winter but less of an impact in summer during the LGM.Compared to the present,the LGM changes in sea-land distribution in East Asia resulted in a decrease in temperature in winter but an increase in summer,with a decrease in annual precipitation by 25%-50%in the coastal areas of East Asia.

4.2.4.LGM environment on the QTP

There is debate on the existence of a unif ed ice sheet on the QTP during the ice age.Using proxy data,Liu et al. (1999,2002b)examined the glacial environment on the QTP and proposed a relatively large area of snow and glaciers on the plateau.Based on a relationship between summer temperature and annual total precipitation at the equilibrium line altitude of glaciers in western China,Zhao et al.(2003)used CCM3 AGCM results to analyze the glacial environment.It was shown that,as a result of balance between precipitation and temperature,the LGM equilibrium line altitude of glaciers on the QTP was 300-900 m lower than at present, which indicated the existence of a large-scale continental ice sheet on the plateau during that period.Later,Jiang et al. (2004)utilized an equilibrium terrestrial biosphere model to simulate the LGM environment in China.It was found that ice covered almost half of the QTP.

As a further step,Jiang et al.(2003)evaluated the potential inf uence of the QTP ice sheet on the LGM climate of East Asia.When the ice sheet was assumed,the strongest temperature changes of around-7 K appearedin the centralwestern part of China during the LGM.The existence of the QTP ice sheet induced an additional regional cooling in East Asia,leadingtoastronganomalousanticycloneneartheQTP in summer and a weak EASM.Meanwhile,the subtropical highoverthewesternNorthPacif cmovedsoutheastwardand precipitation decreased in the central-eastern part of China. Therefore,if a suitable area of ice sheet appearedon the QTP, a lower temperature would be presented in East Asia,which would improvethe LGM temperaturesimulation in East Asia with respect to proxy data.

4.3.LGM Model-data comparisons in China

As previously mentioned,various experiments have been performed to examine the LGM climate in China.Generally, they can reasonably reproduce colder-and drier-than-present climates over the country during that period.Quantitatively, however,model-data differences are still remarkable at the regionalscale.Intheresultsof25PMIP1/2models,theLGM regionallyaveragedannual temperaturewas reducedon average by 3.4 K in South China,5.1 K on the QTP,4.7 K in the Hexi Corridor,and 4.9 K in North and Northeast China.All of them were smaller than the corresponding reconstructed levels of 7.0±3.5 K,6-9 K,13-15 K,and at least 8-10 K as derived from a variety of proxy records(Jiang et al.,2011). On the other hand,LGM annual precipitationminus evaporation changes from the results of 15 PMIP1/2 models agreed qualitatively with lacustrine records,including drier conditions in eastern China and wetter conditions in the region of(35°-42°N,74°-97°E).By contrast,model-datadisagreements occurred on the QTP and in most parts of northern Xinjiang,where simulated drier conditions were opposite to lacustrine records.On the eastern QTP,drier climates agreed with pollen records but disagreed with lacustrine records.In short,the PMIP models successfully reproduced the LGM surface cooling trend over China but failed to reproduce its magnitude,while their humidity results qualitatively agreed with lacustrine records over China excluding the QTP and northern Xinjiang.These model-data discrepancies possibly arose from uncertainties in the boundary conditions(including tropical SSTs for atmospheric model experiments,vegetation conditions,the environment on the QTP,etc.)used in simulations(Jiang et al.,2003;Zhao et al.,2003).For instance,the ability to simulate temperature change could be improved by replacing local vegetation with continental ice on the QTP.In addition,discrepancies in physical processes between climate models could be important contributors tothe differences among simulations.

4.4.Perspective

LGM climate simulations and comparisons with proxy data have improved our knowledge of the features and dynamic mechanisms relating to ice age climates.Based on the PMIP protocol,Chinese scientists have further designed unique experiments to reveal that feedback mechanisms concerned with vegetation,the effect of Pacif c and Atlantic SSTs,and the inf uence of the QTP environment are all important factors and should be evaluated comprehensively in LGM climate simulations.At this point it is important to reiterate that,due to different boundaryconditions used in simulations,the features of the LGM climate have appeared,to some degree,to be divergent.In fact,distinct differences in the response of the LGM climate have been found in various models,albeit when the same or similar boundary conditions were employed(Jiang et al.,2011).To narrow those uncertainties,many improvements are required in the future,not only in terms of proxy data,to verify simulated results,but also in the properties of climate models themselves,including the accuracy of boundary conditions.Issues relating to the importance of unique boundaryconditions should also be examined in depth.For example,it is necessary to systematically consider vegetation-climate feedback,to analyze the impact of different SST reconstructions in some key regions, and to evaluate glacial environments on the QTP and their climatic consequences.

5. Late MIS3 climate modeling

5.1.Reconstructed late MIS3 climate in China

The late MIS3 refers to the periodca.30-40 ka BP,at which time global temperature was lower than in the last interglacial period and the Holocene,but slightly higher than in the early and late last glacial(Yang et al.,2004,and references therein).At the regional scale,however,climate was warmer and wetter in western China,with an estimated annual temperature that was 2-4 K higher on the QTP at that time thanat present(Yaoet al.,1997;Shi et al.,2001;Shi and Yu,2003;Yang et al.,2006;Yu et al.,2007).Such a situation raises an interesting question:why was regional climate change in China so different from that at the global scale? Examining the response of climate models to the boundary conditions at 35 ka BP and its consistency with proxy data are helpful for understanding East Asian climate change on the orbital scale.

5.2.Late MIS3 boundary conditions

Changes in the Earth’s orbital parameters are believed to be the most important forcing for climate change during the late MIS3(Shi et al.,2001;Shi and Yu,2003).Ice sheet extent at 35 ka BP is prescribed as 50%of the area of ice sheets at 21 ka BP according to glacial sediments(Lambeck and Chappell,2001),with the height of the remaining ice sheets kept the same as at 21 ka BP(Peltier,1994).Vegetation in Chinaat 35kaBP hasbeenreconstructedbasedonpollenand macrofossil data,showing that forests in southeastern China extended northwards into the present northwestern steppe region,while thenorthernboundaryoftropicalforests insouthern China shifted north of 24°N(Shi et al.,2001).Atmospheric CO2concentrations were set to 210 ppmv at 35 ka BP from the present value of 345 ppmv.

5.3.Climate modeling for 35 ka BP

A set of AGCM experiments showed that the climate at 35 ka BP featured warm-wet conditions in northern China and warm-dry conditions in southern China compared to the present day(Yu et al.,2003b,2005,2007).Annual temperature was higher in most mid-and lower latitudes of East Asia,which was mainly the result of increased winter temperature.During winter,although insolation in the mid and lower northern latitudes decreased,changes in vegetation in East Asia led to more heat storage and less ref ection of radiation,while they also suppressed cold airf ow and hence reduced the likelihood of temperature falling.In this sense,reconstructed vegetation modif ed the climate model response to orbital forcing through the coupling of atmospheric circulation with land surface conditions.In addition,the ice sheets of the NH during the Quaternary were also found to play an important role in the temperature drop at the mid-high northern latitudes and also to enhancethe south-northtemperature gradients,which in turn increased moisture transport from low to high northern latitudes and increased monsoonal precipitation on the QTP.Vegetation changes in East Asia were inferred to result from increased temperature in the low latitudes,an extended rain belt northwards into China,and an enlarged area of increased precipitation inland.

During summer,precipitation was signif cantly increased in the East Asian monsoon region,which was directly related to an enhanced monsoon due to increased heating contrasts between high and lower latitudes under larger latitudinal gradientsofinsolationandstrongericesheet impactsat35kaBP than at present.Meanwhile,at 35 ka BP the summer sea level pressuredifferencebetweenthe Pacif cOceanandAsiancontinent was 3-6 hPa above the present day.This pattern also increased the strength of the Asian summer monsoon and therefore increased regional precipitation.During winter,the high pressure system in East Asia weakened,while the Aleutian low pressure system in the northern North Pacif c Ocean strengthened at 35 ka BP with respect to the present day. The decreased difference in the land-sea pressure gradient reduced the winter vapor exchange between land and sea. This caused a winter precipitation decrease in eastern China. Meanwhile,the Siberian high pressure center was relatively weak,weakening the East Asian winter monsoon and the incursion of cold air.

5.4.Model-data comparisons in China

It should f rst be noted that qualitative temperature and/or precipitation estimates for the late MIS3 are based on the views of the original authors who reconstructed the climate from various climate proxies.Since the authors did not al-ways specify if the estimates were for the annual or seasonal mean,we assumed they represented a mean status for a relatively long time interval.At 35 ka BP,geological records haveindicatedameantemperatureof～2Khigherinwestern Tibet and northwestern and southern China,but little change in eastern China and a 1-4 K decrease in southwestern China and on the eastern QTP.The simulated temperaturewas compatible with proxy data,showing a 1-3 K increase in northeastern,northern,and western China and a drop in southwestern China.Reconstructed annual precipitation at 35 ka BP was increased over much of China,with values exceeding 300-500 mm in the western QTP,the Tsaidam Basin, and the Yunnan Plateau,and with values of～0.5-1.0 mm d-1higher than at present in western,central,and northeastern China.The simulated precipitation increase in central and western Tibet and Inner Mongolia was consistent with major patterns as seen in geological records.The simulated precipitation decreases in southeastern China were diff cult to validate,since proxy data coverage was insuff cient there.

In terms of regional comparisons at 35 ka BP,the simulated signif cant warm-wet pattern on the QTP agreed with lacustrine sediment records(Shi et al.,2001;Yang et al., 2004),whilethewarm-wetpatternontheLoessPlateaucompared well to paleosol records(Guo et al.,1994;Chen et al., 1997).By contrast,the warm-dry conditions simulated for the eastern coastal plains of China disagreed with vegetation and climate reconstructions(Zheng and Zhou,1995).Additionally,geological records showed signif cantly warmer and wetter conditions in northwestern China(Yang et al.,2004), whereas the simulations only reproduced a temperature increase,while precipitation was slightly changed.

5.5.Perspective

In response to the changes in solar radiation,glaciation in the NH,and East Asian vegetation at 35 ka BP,preliminary experiments showed that both temperature and precipitation increased in northern China,while temperature rose and precipitation decreased in southern China.The QTP underwent a temperature drop and a precipitation increase.Mechanistically,the temperature gradient between inland Asia and low-latitude oceans enlarged,and the transportation of water vapor from the ocean to the continent strengthened,thus increasing precipitation for inland China. At that time,vegetation change had an amplifying effect for orbital forcing through surface albedo.It caused temperature increases in the lower northern latitudes,a weakened temperature gradient in the high northern latitudes,an enlarged area of increased precipitation in inland Asia,and an extended rain belt northwards in China.In general,the inf uence of the Quaternary ice sheets in northern Europe and North America on the East Asian climate was weak.In future simulations,more climate models should be used to reduce model-dependentuncertainty,and current Earth system models should be used to investigate the effect of ocean and vegetation feedbacks on the East Asian climate during the late MIS3.

6. Pre-Quaternary climate modeling

6.1.Mid-Pliocene climate modeling

6.1.1.Reconstructed mid-Pliocene climate

The mid-Pliocene wasca.3.29-2.97 Ma BP.This most recent warm interval in geological time provides a unique opportunity to improve our understanding of a warmer-thanpresentclimate,whichis expectedto besimilar in manyways to the 21st century climate being predicted by climate models,as a result of anthropogenicactivity(Jansen et al.,2007). Various proxies suggest that the mid-Pliocene climate,compared to the present day,was characterized by a greatly reduced continental ice volume,greatly reduced sea ice,with the Arctic being seasonally ice free,a sea-level rise of 25 m, increased SSTs in the high latitudes and little or unchanged SSTs in the lower latitudes,and the presence of warmth and/or moisture-loving vegetation in the middle to high latitudes and a reduction of desert area in equatorial Africa (Dowsett et al.,1999).In addition,at that time,the landsea distribution and geographical conf guration were similar tothepresentday,atmosphericCO2concentrationswereestimated to be about 35%higher than the pre-industrialvalue of 280 ppmv(Raymo et al.,1996),and continental aridity was lower(Guo et al.,2004).

6.1.2.Mid-Pliocene climate modeling

Under the mid-Pliocene boundary conditions,IAPAGCM was used to simulate the mid-Pliocene climate,with particular attention paid to the East Asian climate(Jiang et al.,2005).The mid-Pliocene surface conditions were provided by the U.S.Geological Survey’s Pliocene Research, Interpretations,and Synoptic Mapping(PRISM)group,i.e. the PRISM2 2°×2°digital dataset(Dowsett et al.,1999). They were composed of monthly SSTs and sea ice extent, continental topography,vegetation,and continental ice sheet coverage.

A set of simulations indicated that warmer-and wetterthan-present climates prevailed during the mid-Pliocene (Jiang et al.,2005).Global annual temperature rose by 2.6 K,with a stronger warming in the high latitudes.Changes in SSTs and sea ice extent were mainly responsible for the simulated warming.The effects of vegetation and continental ice sheet changes played an important role in part of the middle to high latitudes,although their global inf uence was quite limited.Global annual precipitation increased by 4.0%, with a larger increase in the high latitudes.On the contrary, drier conditions occurred in most parts of 0°-30°N,consistent with the weakening of tropical Walker circulation and the poleward expansion of Hadley cells(Sun et al.,2013). In addition,both summer and winter monsoons weakened signif cantly in East Asia,which resulted from a weakened thermal contrast and,in turn,a decrease in sea level pressure gradient between the East Asian continent and the adjacent oceans.Based on Chinese red clay sequences,An et al. (2001)reconstructed a weaker-than-present East Asian monsoon during 3.6-2.6 Ma BP and related it to the extent and height of the Himalaya-Tibetan Plateau.The present simu-lations revealed that the changes in SSTs and sea ice extent can also lead to such changes.Therefore,particular attention should be given to oceanic behaviors when exploringthe PlioceneclimateofEast Asia.Inaddition,the simulatedmid-Pliocene vegetation differed from the present day over 62% of the global ice-free land surface(Jiang,2013).Vegetation feedbackhad little overallimpact on the global climate of the mid-Pliocene.Attheregionalscale,however,interactivevegetation led to statistically signif cant increases in annual temperatureoverGreenland,thehighlatitudesofNorthAmerica, the mid-high latitudes of eastern Eurasia,and western Tibet, and reductions in most of the land areas at low latitudes,owing to vegetation-inducedchanges in surface albedo.

Inter-model comparisons indicated that IAP-AGCM results were overall compatible with earlier simulations.The simulated mid-Pliocenewarming varied from 1.4 K using the GISS(Goddard Institute for Space Studies)AGCM(Chandleret al.,1994),to 1.9K usingtheUKMO(UK Meteorological Off ce Unif ed Model)AGCM(Haywoodet al.,2000),to 2.6 K using IAP-AGCM,and to 3.6 K using the GENESIS (Global Environmental and Ecological Simulation of Interactive Systems)AGCM(Sloan et al.,1996).The geographical distribution of the warming was also broadly consistent among the models.Meanwhile,all the models simulated a slightly wetter climate for the mid-Pliocene.In East Asia,the UKMO AGCM produced signif cantly reduced summer precipitation and an increase over part of the QTP,which agreed with IAP-AGCM results.Additionally,the mid-Pliocene annual precipitation decrease over much of East Asia in the GENESISAGCMwasalsoconsistentwiththeresultsofIAPAGCM.

6.1.3.Model-data comparisons in China

IAP-AGCM simulations indicated that,except for on the QTP where annual temperature declined due to changes in topography,annual temperatures rose by 4-8 K in eastern Chinaand1-4K in westernChina,with respecttothe present day.Annual precipitation reduced largely in eastern China, with anaverageofabove0.5mm d-1,particularlyin the middle reaches of the Yangtze River valleys.Meanwhile,annual precipitation increased slightly in northern Xinjiang,Qinghai,and most parts of Tibet,whereas it reduced in central and southern Xinjiang.

Available proxy data consistently suggest that the mid-Pliocene was warmer(e.g.,Han et al.,1997;Ding et al., 1998),and that the East Asian winter monsoon was weaker than at present(e.g.,Ding et al.,1998;An et al.,2001;WehausenandBrumsack,2002;Tianetal.,2004;SunandWang, 2005;Wan et al.,2007;Sun et al.,2008).Therefore,both the warming and weakening of the East Asian winter monsoon as simulated by IAP-AGCM agreed with proxy data.In contrast,both drier(Han et al.,1997;Wu et al.,2006)and wetter(Ding et al.,1998;Li et al.,2004)climates were reconstructed for the mid-Pliocene by loess-paleosol and red clay sequences from the Chinese Loess Plateau,although the locations of the sample sections were close.Additionally, most proxy data suggest a weaker-than-present EASM(Tian et al.,2004;Wang and Deng,2005;Wu et al.,2006;Wan et al.,2007).However,a similar(Ding et al.,1999),similar or slightly stronger(Wehausen and Brumsack,2002), and stronger(Ding et al.,2005;Sun et al.,2008)EASM has also been reported,and the mid-Pliocene EASM was shown to strengthen in recent multiple models(Zhang et al.,2013b).This means there are large uncertainties in the changes of mid-Pliocene precipitation and EASM,and the extent to which IAP-AGCM-simulated aridity and EASM weakening are compatible with proxy data remains an open question,even in a qualitative sense.

6.1.4.Perspective

Global mean temperatures rose by 0.74±0.18 K during the period 1906-2005(Trenberth et al.,2007),and global emissions of greenhouse gases are expected to lead to a continued and strengthened warming in the future.Based on a variety of emissions scenarios for atmospheric greenhouse gases and aerosols,globally averaged temperature is projected to increase by 1.1-6.4K by 2090-99,relative to 1980-99(Meehl et al.,2007).However,the spatiotemporal pattern of the warming and the dynamic mechanism responsible for the current and forthcoming warming are highly uncertain. At this point it is of interest to examine the mid-Pliocene climate because it is thought to be similar to the model-based climate projections for the 21st century.Moreover,a variety of proxy data are available and well constrained for that period.In general,investigation of the mid-Pliocene climate can provide insights into a warmer-than-present climate regime.They can also help to assess the ability of climate models to reproduce a warm climate and examine the sensitivities of climate models to different boundary conditions and the associated feedbacks that may be operating in a warm climate regime,particularly when investigating climate model sensitivity to an elevated atmospheric CO2scenario.And f nally,theycan help to interpretthe mid-Pliocene geological records.

6.2.Inf uence of continental changes on climate

The present tropical Pacif c Ocean is largely characterized by a strong temperature gradient between the warm pool in the west and the cold tongue in the east.In particular, the western Pacif c warm pool(WPWP)is a region with the highest SSTs of all the oceans,and is also a region with the strongest mass and energy exchanges between the atmosphere and ocean.This region therefore has an important impact upon global climate on the seasonal and longer scales. However,proxy data indicate that the WPWP did not come into existence untilca.10 Ma BP,and the present WPWP was formed only atca.3 Ma BP(e.g.,Wang,1994;Chaisson and Ravelo,2000).Some researchers suggest that there was a close relationship between the formationof the WPWP and continentaldrift because the Australian and South American continentsmovedaway fromthe Antarctic continentand drifted slowly northward since the late Tertiary,which f nally resulted in the closure of the Indonesian seaway linking the Pacif c and Indian oceans,and the closure of the Isthmus ofPanama linking the Pacif c and Atlantic oceans(Zhou et al., 2004b).

6.2.1.Impact of the closure of the Indonesian seaway on climate

UsinganAOGCM andits oceaniccomponentOGCM,Yu et al.(2003c)performed numerical experiments to address the climatic consequences of the closures of the Indonesian seaway and the Isthmus of Panama.According to geological records of plate tectonics(Zhou et al.,2004b),topographical conditions at 6 Ma BP and 14 Ma BP were f rst compiled.Afterthat,sensitivity experimentsusingtheabovecontinental conf gurations were respectively carried out by use of the OGCM.Additionally,Yu et al.(2003c)also used the AOGCM to perform the same four experiments with topography set to that of the present day,at 6 Ma BP,at 14 Ma BP, and at 14 Ma BP,except with the opening of the Isthmus of Panama factored in,respectively.Meanwhile,to investigate the effect of the northward shift of the Australian continent on tropical oceanic circulation and the SH climate,Zhou et al.(2004a,2005)employed the GISS AOGCM to perform two experiments with the Australian plate conf gurationat 14 Ma BP and at present,respectively.

In the OGCM experiments,the closure of the Indonesian seaway had signif cant impacts on oceanic circulation in the tropical Indian and Pacif c oceans(Yu et al.,2003c;Jian et al.,2006).At present,the equatorial undercurrent,which is composed of North Pacif c water and South Pacif c water that turns eastward in New Guinea,f ows eastward and then upwells in the eastern equatorial Pacif c at a subsurface layer.When the Australian plate lay south of its present position and the Indonesian passage accordingly became wider at 14 Ma BP,most of the water from the South Pacif c passed through the Indonesian passage and entered directly into the Indian Ocean.Thus,the contribution of southern equatorial water to the equatorial undercurrent became less than at present.SinceSouthPacif cwaterwas generallywarmerthan North Pacif c water,the simulated equatorial undercurrent, which was mainly supplied by water coming from the North Pacif c at 14MaBP butfromtheSouthPacif c at present,was 1-2 K warmer in the equatorial Pacif c but～1 K colder in the equatorial Indian Ocean at present than at 14 Ma BP.In addition,more net heat energy entered into the surface of the equatorial Pacif c.Furthermore,the AOGCM experiments reproduced similar results as obtained by the above OGCM.

The Australian Plate drift at 14 Ma BP also had significant impacts on atmospheric circulation and climate in the middleandhighsouthernlatitudes(Zhouet al.,2004a,2005). Compared to the present,both anticyclonic circulations over subtropical oceans and cyclonic circulations around 60°-70°S wereintensif ed.Thus,subtropicalhighs andcircumpolar lows strengthened,which resulted in a stronger Antarctic Oscillation at 14 Ma BP.Precipitation and temperature also varied correspondingly.Precipitation decreased at around 40°S but increased at around 60°-70°S,while temperatures rose in the high latitudes of the South Pacif c but descended overthe Weddell Sea andits northernside.Inaddition,dueto the changes in temperature and atmospheric circulation,sea ice extent increased in the Ross Sea and its western side,but decreased in the Weddell Sea.

6.2.2.Effect of the closure of the Central American seaway on climate

The closure of the Central American seaway or the Isthmus of Panama could also have played an important role in the formationof the present tropical oceanic and atmospheric circulations.The AOGCM experiments showed that the closure of the Isthmus of Panama was important in forming the presentWPWP(Yuet al.,2003c;Jianet al.,2006).Theemergence of the Isthmus of Panama was found to induce strong upwelling and uplift of the thermocline in the eastern equatorial Pacif c,but a remarkable decrease of the thermocline in the western Pacif c.As a result,the closure of the Isthmus of Panama isolated heat exchange between the Pacif c and Atlantic oceans,led to a cooling of SSTs and an uplift of the thermocline in the eastern Pacif c,and hence increased the contrast of heat content between the western and eastern Pacif c.The amplitudes of thermocline depth changes between the two experiments indicated that the closure of the Panama seaway played an important role in the evolution of the WPWP.On the whole,this set of experiments implied that the closure of the Indonesian seaway could have resulted in the beginning of the WPWP,but the f nal formation of the WPWP was most probably induced by the closure of the Panama passage.

6.2.3.Model-data comparisons

Using marine sediments from the South China Sea and other lines of evidence,Jian et al.(2006)proposed three stages of WPWP evolution.The f rst was the beginning of the WPWP during 11.5-10.6 Ma BP.This period was coincident with the signif cant narrowing or partial closure of the Indonesian seaway between the Pacif c and Indian oceans, although the timing of the closure of the Indonesian seaway is still a point of vigorous discussion.The second was the remarkably weakened or extremely unstable WPWP during 10.6-4.0 Ma BP.And the last was the formation of the present WPWP atca.4.0-3.2 Ma BP based on the signifcant increases in the south-north thermocline gradient of the South China Sea and the west-east thermocline gradient of the equatorial Pacif c,which is often linked with the closure of the Panama passage,as discussed in Jian et al.(2006).The aforementioned simulations support the notion that the formation of the present WPWP depended strongly on the drift of the continents.However,it should be noted that only the reconstructed topography around the Indonesian and Central American seaways at 14 Ma BP was considered in those experiments.Besides the uncertainties in the history of the seaways,other changes in the QTP,glaciation in the NH,greenhouse gases,etc.,also affected global and regional climates during the late Cenozoic.It is necessary to reconstruct a more complete picture of boundary conditions to accurately estimate the effect of continental drift on oceanic and atmospheric circulations using climate models.

6.3.Climatic consequences of the QTP uplift during the late Cenozoic

6.3.1.Introduction

TheupliftoftheQTP was a majoreventinthe naturalhistory of the Earth.A large quantity of geological evidence indicates that the northwarddrift of the Indian-AustralianPlate and its collision with the Eurasian Plate caused a gradual uplift and the formation of the QTP during the late Cenozoic (Molnar,1989;Ruddiman et al.,1989).Chinese scientists have studied the uplift of the QTP and the history of the East Asian monsoonbased on various lines of geological evidence(e.g.,An,2000;Wang et al.,2005).For example,using records of aeolian sediments from Chinese loess and marine sediments,Anet al.(2001)proposedthreestages ofevolution of Asian climate since the late Miocene and their relationship with the phases of the Himalayan-Tibetan Plateau uplift and NH glaciation.Although there is debate regarding the uplift processes of the plateau among the international academic community(e.g.,Harrison et al.,1992;Li and Fang,1999; Harris,2006;Wang et al.,2008a),there is no dispute that the plateau uplift had signif cant inf uences on global and Asian climates throughout the Cenozoic era.The profound impact of the QTP uplift on the evolution of the Asian monsoon is now being widely recognized.

The important role of the QTP in controlling the Asian climate has been known for a long time.Fl¨ohn(1968)was among the f rst to point out the signif cance of the QTP in maintaining the large-scale Asian monsoon circulation.With the developmentofclimate modelsandcomputertechnology, numerical modeling approaches have been used to explore the role of topography in forming the Asian monsoon.Early in the mid-1970s,AGCMs were used to compare climates under conditions with and without global mountains(Manabe and Terpstra,1974;Hahn and Manabe,1975).Their results indicated that the QTP not only controlled the location and intensity of the Siberian High in winter,but also the establishment and development of the Asian summer monsoon.Qian et al.(1988)examined the impact of the QTP on the East Asian monsoon with a limited-area numerical model and found that atmospheric heat source anomalies related to the large-scale topography affected the development of the EASM circulation.Moreover,CCSM3 AOGCM simulations showed that the QTP has a strong effect on oceanatmosphere interactions in the tropics and North Pacif c and on atmospheric circulation and precipitation in North America,Europe,andthe SH(Zhaoet al.,2009;Zhouet al.,2009). Furthersimulationsconf rmedtheirsignif cantimpactsonthe evolution of the Asian monsoon,the gradual increase in aridity in Central Asia,and the cooling of global climate through the Cenozoic(Kutzbach et al.,1989;Manabe and Broccoli, 1990;Kutzbach et al.,1993).Although there is still debate regarding the role of the QTP in the Asian monsoon system (Boos and Kuang,2010;Wu et al.,2012),it is generally acceptedthattheQTPupliftincreasedsummerheatingandwinter cooling on the plateau,which enhanced seasonal changes in winddirections(RuddimanandKutzbach,1990),thusproducing a marked increase in the intensity of the Asian summer and winter monsoons(Kutzbach et al.,1993).In recent years,Chinese scientists have conducted a number of experiments on the stepwise uplift of the QTP to furtherexplorethe role of tectonic uplift on the evolution of the Asian climate, especially on the East Asian monsoon.The key results from those experiments are now brief y described.

6.3.2.The QTP uplift and East Asian monsoon evolution

Due to the lack of precise and comprehensive estimates for the paleoelevation of the QTP,Liu and Yin(2002)used the COLA(Center for Ocean-Land-Atmosphere Studies) AGCM to conducta set ofexperimentswith ideal topography to examine the impact of stepwise uplift of the QTP on the evolution of the monsoonclimate in East Asia.Eleven experiments were performed to represent the varying topography within(10°-60°N,50°-140°E),where the heightat each grid was prescribed to be 100%,90%,...,10%of the present elevation.A no-topography experiment was also performed for reference.Exceptforthe QTP topography,all otherboundary conditions remained the same as the control experiment,i.e., current conditions.

TheQTP upliftcausedsignif cantchangesin atmospheric circulation.In particular,the uplift was closely associated with the establishment and evolution of the Asian monsoon system.The response of the East Asian monsoon was more sensitive to the uplift than that of South Asia.Moreover,the effect of the uplift on winter monsoon was more prominent than that on summer monsoon in East Asia.In northern East Asia,the formation of monsoon climate was marked mainly by the establishment of a winter monsoonand the appearance of an alternationbetween dominant surface winds with opposite directions in winter and summer,which corresponded to the QTP uplift when the height reached approximately half of its current elevation.However,the monsoon was established much earlier and displayed a non-linear response in southern East Asia.Its existence was found well before the QTP reached half of its current height.Additionally,the uplift of the northern QTP mainly caused intensif ed summer monsoon and increased precipitation in northern East Asia, but had little inf uence on the South Asian monsoon(Zhang and Liu,2010).

Given that the main uplift of the QTP occurred in a time period earlier than the Quaternary(Fort,1996),and that pre-Quaternary boundary conditions were greatly different from the present,Jiang et al.(2008)used IAP-AGCM to perform idealized numerical experiments under the PRISM2 boundary conditions(Dowsett et al.,1999).The aim was to examine the sensitivity of the East Asian climate to the progressive uplift and expansion of the QTP under the reconstructed boundaryconditions for the mid-Pliocene.When the QTP was progressively uplifted,global annual temperature declined gradually and statistically signif cant cooling was registered only in the NH,especially on and around the QTP, with a largermagnitudeoverlandthanoverthe ocean.Onthe contrary,annualtemperaturerose notablyin Central Asia and most parts of Africa,as well as in northeasternmost Eurasiawhen the QTP exceeded half of its current elevation.Additionally,the inf uence of the QTP uplift on annual temperature was limited to within East Asia before half of its current elevation and then extended to most parts of the NH when the QTP continually uplifted.The QTP uplift also led to an increase in annual precipitation on the plateau but a decrease in northern Asia,the Indian Peninsula,most parts of Central Asia,part of western Asia,and the southern portion of northeastern Europe.A similar-to-present EASM system initially appeared when the QTP reached 60%of its current elevation and intensif ed gradually with a continued uplift.At 850 hPa,the uplift of the plateau induced an anomalous cyclonic circulation around the QTP in summer and two anomalous westerly currents respectively located to the south and north of the QTP in winter.In the mid-troposphere,a similar-topresent spatial pattern of the summer western North Pacif c subtropical high was apparent only when the QTP exceeded half of its current elevation,and the East Asian trough deepened steadily in response to the progressive uplift and expansion of the QTP.

6.3.3.Role of the QTP uplift on Asian-Africanaridif cation

Aridif cation is one of the most devastating natural disasters.According to the world distribution map of arid regions developed by UNESCO(1979),the most severe and vast arid regions include northern Africa and the Eurasian interior.Identifying the primary factors involved in the formation of arid regions is undoubtedly important,both scientif cally and from a practical standpoint.Geological evidence has revealed that climate developed toward conditions of aridity in inland Asia and America from the late Cenozoic (Ruddiman et al.,1989).Simulations indicate that the QTP uplift played an important role in the formation of an arid climate in the middle northern latitudes(Broccoli and Manabe, 1992;Kutzbach et al.,1993).The suggestion was that the mid-latitude aridity was largely due to orography.

To understand the impact of the QTP uplift on the development of aridity,Liu and Yin(2002)showed that,as a result of water balance linking precipitation,evaporation,runoff, and percolation,soil moisture was superior to precipitation in indicating aridity changes.As such,the most prominent feature in the calculated percentage changes of annual soil moisture between the different stages was the gradual aridity in the extensive area from Central Asia to Northwest China, and even in North Africa.Moreover,the process of desiccation in these regions intensif ed dramatically during the later stages of the QTP uplift.

Although simulations indicate that the QTP uplift led to the aridity of Central Asia and northern Africa,it cannot totally account for the formation of the arid regions from North Africa to Central Asia.In the distribution of annual precipitation during the course of the QTP uplift,there was a large non-mountainous rainy region,with more than 8 mm d-1, fromIndiato East Asia southof 22°N,whichimplies that this tropical monsoon precipitation essentially did not depend on the QTP(Liu et al.,2001).Moreover,the continental interior of northernAfrica throughto Eurasia,around25°-45°N,still remainedarid to a considerabledegreeeven without the QTP. Therefore,the occurrence of arid regions and aridity in Central Asia and northern Africa should not be solely attributed to the QTP uplift.

6.3.4.Modulation of tectonic uplift on orbital-scale EASM variability

Geological records and numerical simulations have shown that the Asian monsoonclimate variability on the geological scale is modulated by both the Earth’s orbital changes and tectonic uplift of the QTP.Moreover,it is suggested that the response of the South Asian monsoon to orbital forcing could depend upon the elevation of the QTP(Prell and Kutzbach,1997).To explore the role of the QTP uplift in modulating the response of the EASM to orbital forcing, Liu et al.(2003a)performed a set of experiments with the CCSM AGCM.Under scenarios of current mountains and non-mountainous conditions,they examined the response of the EASM to changes in precession and obliquity parameters.With the present orography,summer monsoon in northern East Asia responded signif cantly to orbital forcing.In the absence of the QTP,however,the orbital-scale variability of Asian monsoon reduced markedly.By examining the spatial and temporal variations of climate variables and indices,including surface and upper-air winds,air humidity and soil moisture,precipitation,and monsoon intensity,Liu et al.(2003a)suggested that the QTP may serve as an amplif er of orbital-scale variability of the EASM.

6.3.5.Inter-model and model-data comparisons

Results fromrecentstudies inChinafurtherconf rmthose drawn from earlier simulations with respect to the inf uence of the QTP on the Asian monsoon,including the development of the Siberian-Mongolianhigh in winter(Manabe and Terpstra,1974)and the aridity of Central Asia in summer (Broccoli and Manabe,1992;Kutzbach et al.,1993)during the course of the uplift.On the contrary to previous studies (Hahn and Manabe,1975;Kutzbach et al.,1993),Liu and Yin(2002)indicated that the response of the Indian monsoon to the QTP uplift was not as strong as that of the East Asian monsoon.They emphasized that the monsoon phenomenon featured the alteration of the prevailing winds with almost opposite directions between winter and summer.Therefore, wind difference between winter and summer was considered as the prerequisite for monsoon.Their results suggested that the seasonal contrast of winds,precipitation,and temperature between winter and summer did not change much in South Asia during the whole course of the QTP uplift.Even under no-topography conditions,the South Asian monsoon was clearly visible in the simulations.This conformed to the results of numericalexperimentsby DeMenocaland Rind (1993)and Ramstein et al.(1997),suggesting that there existed monsoonal rain areas south of 30°N even without the QTP,although the precipitation amount was lower than current levels.In this sense,the formation of the South Asian monsoonwas lessdependentontheexistenceoftheQTPthan the East Asian monsoon.

Due to limitations in computing resources and available geological boundary conditions,only sensitivity experiments have been performed in this f eld to date.It is therefore diffcult to compare model results directly to geological records, and only some trends of simulated climate change in Asia could be supported by geological evidence.For example, Liu et al.(2003a)reported that the QTP uplift could amplify the orbital-scale variability of the EASM,which was in good agreement with geological records.The magnetic susceptibility of loess and red clay on the Chinese Loess Plateau also showedthattheprecessionandobliquitycomponentsofmonsoon intensity enhancedsignif cantly during the late Pliocene (Liu et al.,2003a).At the same time,geological evidencehas shown that the QTP uplifted rapidly and extended northward and northeastward considerably during 3.6-2.6 Ma BP(Li et al.,1997;An et al.,2001).Combining this geological evidence with simulations,we believe that the signif cant intensif cation of the orbital-scale EASM variability at both precession and obliquity periods is,at least partially,attributable to the strong growth of the QTP.

It is particularly noteworthy that when the QTP rose to about half of its current height,abrupt changes occurred in the evolution of the northern East Asian monsoon,e.g.the establishment of the northerly winter monsoon and the intermittent strengthening of the summer monsoon(Liu and Yin, 2002;Jiang et al.,2008).This seems to imply that there was a critical height of the QTP for its inf uence on the northern East Asian monsoon.On the other hand,geological evidence from both South Asia(Ruddiman et al.,1989;Molnar,1997) and East Asia(An et al.,1999)shows some abrupt changes in the monsoons on the tectonic scale.The abrupt establishment of the northern East Asia monsoon could correspond to the events of loess deposition in the Loess Plateau in North China(Guo et al.,2002).These agreements in the change trend between simulations and geological evidence may ref ect the important role of the QTP in determining regional climate evolution in East Asia.However,to comprehensively understandthe mechanisms behind East Asian monsoon evolution,it is necessary to clarify the geological history of the QTP uplift,collect more geological records,and then to conduct further numerical experiments in the future.

6.3.6.Perspective

At present,the history of the QTP uplift is still an open question.A lack of knowledge regarding the threedimensional paleoelevation of the QTP restricts the design of numerical experiments and comparisons with geological records.In this case,only idealized experiments,where the QTP topography is usually prescribed as some percentage of current elevation,have been designed in order to assess the effect of the QTP uplift.Although such a simplif ed design helps to isolate the impact of topography from those of other forcings,the actual process of the QTP uplift is more intricate,which might give rise to quite different climatic consequences.With realistic geometry of topography,more reasonable schemes should be taken in future simulations to better assess climatic response to tectonic uplift.Recently,for example,Zhang et al.(2012a)conducted numerical experiments to indicate that the uplifts of the different subregions of the QTP had different effects on the Asian climate.

The evolution of the Asian monsoon and arid climate is also associated with other forcings besides the QTP uplift.From the early Pliocene to late Miocene,other signifcant changes occurred in the paleogeographical environment of the Eurasian continent apart from the QTP uplift.Therefore,the reconstruction of various boundary conditions,such as vegetation,ice sheet,and atmospheric CO2concentrations,are also desirable for climate modeling of geological periods.

Physically-basedclimate models describingvariousfeedback mechanisms are needed in the future.The absence of an interactive ocean is a limitation in previous simulations of the effect of the QTP uplift.Previously,AOGCM experiments have been performed to explore the effect of progressive mountain uplift on the Asian climate(Abe et al., 2003;Kitoh,2004).In addition,snow-ice feedback on the QTP(Bush,2000)and uplift-weathering feedback(Raymo and Ruddiman,1992)may further reinforce the thermal and dynamic inf uence of mountain uplift.Collectively,with the continuous improvement in the reconstruction of threedimensional paleoelevation of the QTP,and the development of climate models,we will be able to obtain a more comprehensive understanding of the tectonic-climate link through climate modeling.

6.4.East Asian climate transition through the Cenozoic

6.4.1.Evolution of the East Asian climate pattern in the Cenozoic

At present,China is dominated by the East Asian monsoon climate.Winter northerly winds bring cold and dry air fromthehighlatitudesoftheEurasiancontinent,andsummer southerly winds bring warm and moist air from the tropical oceans.A large area of desert is distributed across Central Asia.However,the monsoon-dominant climate did not form until the Paleogene/Neogene boundary,also known as the Oligocene/Miocene boundary.In the Paleogene,East Asia featured a zonal climate pattern with an arid band extending from the eastern coast of China to Central Asia(e.g.,Sun and Wang,2005;Zhang and Guo,2005;Guo et al.,2008),but in the Neogene a non-zonal climate pattern with inland aridity dominated East Asia(Zhang et al.,2007a,2007b;Guo et al., 2008).Dating of loess-paleosol sequences and more detailed reconstructions of geological indicators have demonstrated that this major reorganization of the climate pattern occurred neartheOligocene/Mioceneboundary,byca.22MaBP(Guo et al.,2002;Zhang and Guo,2005;Guo et al.,2008).

6.4.2.Impact of the QTP uplift and the Tethys retreat on climate

The mechanisms involved in monsoon climate have received broad attention from scholars for a long time.Classical theory emphasized that the land-sea thermal contrast was the key element to monsoon(Halley,1686).A later theory invoked the seasonal movement of planetary wind as themain cause of monsoon(Fl¨ohn,1956).However,those two theories cannot fully explain the formation of the East Asian monsoon because the East Asian monsoon is a complex of tropical and subtropical monsoons.Subsequently,a number of numerical experiments have focused on the impact of the Himalayan-TibetanPlateauupliftontheintensif cationofthe Asian monsoon and Asian inland aridity(Manabe and Terpstra,1974;An et al.,2001;Liu and Yin,2002;Jiang et al., 2008).On the other hand,several studies have emphasized the important role played by the Tethys retreat(Ramstein et al.,1997;Fluteau et al.,1999).

In this area,Zhang et al.(2007a)attempted to distinguish between the effects of those two major factors by using IAPAGCM.Thirty numericalexperiments were performedunder six Tethys Sea and fve Himalayan-Tibetan Plateau conditions.In the fve Himalayan-Tibetan Plateau conditions,the maximum height of the plateau was progressively increased from 1500 m to 4500 m.In the six Tethys Sea conditions,the sea was gradually closed from a large sea with a full connection to the Arctic to a small epicontinental sea.Other boundary conditions were kept constant throughout the experiments.

This set of experiments conf rmed again that the Himalayan-Tibetan Plateau uplift can strengthen the East Asian monsoon.The uplift remarkably increased the seasonal contrast of precipitation in the monsoon areas and Northwest China.Furthermore,the experiments illustrated details about the impact of the Tethys retreat on the East Asian climate.Two kinds of precipitation f elds were obtained in East Asia.One was a zonal pattern,while the other was non-zonal one.For the former,a def cient rain belt with precipitation less than 1.5 mm d-1was distributed from the eastern coast of China to Central Asia.In the non-zonal pattern,a def cient rain region only appeared in Central Asia. Thus,thesimulated transitionofprecipitationpatternsagreed well with the reorganization of climate suggested by geological records.The transition occurred when the Tethys retreated from the southern part of West Siberia to the Turan Plate.The retreat at this stage led to the reorganizationof the pressure system by a changing land-sea thermal contrast.In summer,a low pressure anomaly was centered on the Turan Plate and Northwest China,where the sea was changed into land.The low pressure anomaly caused a cyclonic circulation anomaly in East Asia.Along the southeastern edge of the anomalous cyclonic circulation,the southwesterly winds strengthened greatly.They carried a larger amount of water vapor and increased precipitation in the monsoon areas.In winter,the retreat led to a high pressure anomaly in Northwest China and a low pressure centered on Mongolia.As a result,northwesterlywinds intensif ed,and it reducedprecipitation and increased aridity in inland China.

6.4.3.Impact of the South China Sea expansion on climate

IAP-AGCM was also used to examine the potential effects of other tectonic changes(Zhang et al.,2007b).Additional factors addressed were the Indian Peninsular drift,the South China Sea expansion,and the East China Sea transgression.The experiments revealed a relatively subordinate role of the Indian Peninsular drift and the East China Sea transgression.The South China Sea expansion,however,was another major forcing,in addition to the important roles of the Tethys retreat and the Himalayan-Tibetan Plateau uplift. The expansion provided suff cient water vapor for summer precipitation in the monsoon areas,and then enhanced the humidity contrast between East Asia and Central Asia.

These simulations demonstrated again that the geographical evolution drove the reorganization of the climate pattern over China during the Cenozoic.The Tethys retreat and the Himalayan-Tibetan Plateau uplift strengthened the East Asian monsoon to provide the dynamic conditions,while the South China Sea expansion provided suff cient water conditions for the major reorganization of the climate pattern in East Asia.These three factors acted together to cause the major reorganizationof regional climate in East Asia.

6.4.4.Perspective

The above-discussed studies were still limited by the coarse horizontal resolution of IAP-AGCM and prescribed SSTs and sea ice extent.A new set of simulations of a high resolution AGCM also showed that East Asia was dominated by a zonal desert/steppe climate in the Eocene(Zhang et al., 2012b),supporting the results described above.The coarse resolution of IAP-AGCM hence has less effect on simulations of basic climate patterns in East Asia.In future simulations,the effect of an interactive ocean should be emphasized.Caused by a large drop in atmospheric CO2concentrations and changes in oceanic circulation,the cooling event that happened at the boundary of the Oligocene and Miocene might have also played an important role in the evolution of climate in East Asia(Zhang et al.,2012b).The task remains to explore the possible relationship between the major reorganization of climate and the changes in the oceans at the boundary of the Oligocene and Miocene.

7. Conclusions and perspective

Over the last millennium,changes in solar radiation and volcanic activity were mainly responsible for pre-industrial climate change,while an increase in atmospheric greenhouse gas concentrations played the most important role in the warming of the 20th century over China.Although simulations agree in several respects with reconstructions,modeldata discrepancies are still notable in China(e.g.,Liu et al., 2005;Peng et al.,2009;Man et al.,2012;Man and Zhou, 2014).To reduce the uncertainties surrounding simulations, the potential effect of changes in atmospheric aerosols and land-use and land-cover needs to be further evaluated in future work.Historical climate simulations derived from a hierarchy of climate models should be compared so as to investigate their common and different responses to the same or similar forcings at the regional scale.The physical processes that affect the sensitivity of climate models to the specif ed natural/anthropogenicforcingagents should also be explored.

Changes in the Earth’s orbital parameters were the principal cause for mid-Holoceneclimate change.However,simulations with this forcing disagree in temperature with proxy data over China,particularly for winter.It is worth emphasizing that an interactive ocean appears to amplify the response of climate models to mid-Holocene orbital forcing and reduce model-data disagreements over China,although the underlying mechanism remains unclear(Wei and Wang, 2004;Jiang et al.,2012).Reconstructed vegetation appears to be an important component for mid-Holocene East Asian climate in AGCM experiments(Wang,1999a,2002;Chen et al.,2002),whereas interactive vegetation has little overall effect on the climate of China based on coupled models with a dynamic vegetation model(Jiang et al.,2012;Tian and Jiang,2013).In this regard,both vegetation reconstruction and climate-vegetation interaction require further study. The mid-Holocene annual and winter temperature changes over China from multiple proxy data are opposite to those from 36 PMIP models(Jiang et al.,2012).More reconstructions and simulations are urgently needed to investigate whether such mismatch arises from the models,from the proxy data,or from both sides.In addition,more attention should be given to both transient simulations for the whole Holocene and time-slice simulations for climate events,such as those that occurred at 8.2 and 4 ka BP,so as to enhance our knowledge of the nature and cause of Holocene climate change.

In response to the larger ice sheets of the NH in the LGM,more extensive sea ice,colder SSTs,and lower atmospheric CO2concentrations,a number of experiments have reliably reproduced the main features of the glacial climate overChina.However,thesimulatedchangesareweakeroverall than those suggested by proxy data(Jiang et al.,2011).A series of sensitivity experiments suggested that the effect of reconstructed vegetation,the impact of reconstructed SSTs in the western Pacif c,and the potential inf uence of the QTP environment change were all remarkable(Chen et al.,2000; Yu et al.,2001;Liu et al.,2002a;Jiang et al.,2003;Yu et al., 2003a;Zhao et al.,2003;Jiang et al.,2004;Zhao et al.,2004; Sui and Zhao,2005).Climate change due to those factors can partly reduce model-datadisagreements overChina,particularly on the QTP.In this sense,it is necessary to systematically evaluate the effect of an interactive ocean and vegetation,the impact of different SST reconstructions in key regions,and the environmental conditions on the QTP and their climatic consequences for the East Asian monsoon region in future simulations.Additionally,to what extent the latest PMIP3 simulations are consistent with earlier simulations and proxy data should be evaluated,which will be helpful in understanding the LGM East Asian climate.

Orbitally induced insolation changes are hypothesized to be responsiblefor the climate change in China duringthe late MIS3(Shi et al.,2001;Shi and Yu,2003).Preliminaryexperiments have revealed that the glaciations in the NH and vegetation change can modify climate model response to orbital forcing,and therefore also contribute signif cantly to the late MIS3 climate in China(Yu et al.,2003b,2005,2007).Although part of the simulated late MIS3 climate is supported by reconstructions,model-data disagreements are still not settled over the country.To what extent the obtained results are model-dependent and how ocean feedback behaves need to be specif cally investigated.

Preliminary simulations have revealed basic characteristics of the mid-Pliocene global climate(Chandler et al., 1994;Sloan et al.,1996;Haywood et al.,2000;Jiang et al.,2005).The spatial pattern of the mid-Pliocene climate and the underlying mechanism,however,are highly unclear (Jansen et al.,2007).Over China,model-data disagreements in changes of precipitation and the EASM need to be further addressed(Jiang,2009;Zhang et al.,2013b).Of particular interest is to employ state-of-the-art climate models to simulate the mid-Pliocene climate,which will improve our understanding of a warmer-than-present climate regime.During the Cenozoic,the QTP uplift was one of the most important events on Earth.Although the spatial and temporal evolution of the QTP remains unresolved,a series of idealized numerical experiments consistently revealed that the progressive uplift and expansion of the QTP played important roles in forming the modern Asian climate,particularly fortheEastAsian monsoonclimate(LiuandYin,2002;Jiang et al.,2008;Zhang and Liu,2010;Zhang et al.,2012a).In addition,the Tethys Sea retreat and the South China Sea expansion could have had important effects on the formation of the East Asian monsoon-dominant environment pattern during the late Cenozoic(Zhang et al.,2007a,2007b,2012b).

Finally,we would like to stress that a wealth of historical and geological data have been used to reconstruct past climate change over China,e.g.,historical documents,tree rings,stalagmites,lacustrine and f uvial sediments,marine sediments,ice cores,loess and red clays,paleosols,and pollen data.It is promising to employ climate and Earth system models to simulate past climate change and events in the East Asian monsoonregion,and to comparesimulations with reconstructions.The application of regional climate models should be helpful when model-data comparisons are performed at the regional scale.All of these studies will provide insights into the dynamic mechanisms of climate change on a range of timescales.Interdisciplinary cooperation will undoubtedly advance our knowledge of past,present,and future climate change in China,which comprises～25%of the world’s population.

Acknowledgements.We would like to sincerely thank Prof. WANG Huijun for his helpful comments and suggestions,which helped greatly in improving the manuscript.DJ is supported by the Strategic Priority Research Program of the Chinese Academy of Sciences(Grant No.XDB03020602).GY is supported by the Key Directional Program of the Knowledge-innovation Project of the Chinese Academy of Sciences(Grant No.KZCX2-YW-338-2).PZ is supported by the National Basic Research Program of China(Grant No.2007CB815901).XC is supported by the National Natural Science Foundation of China(Grant No.40875043). JL is supported by the National Basic Research Program of China (Grant No.2010CB950102).

REFERENCES

Abe,M.,A.Kitoh,and T.Yasunari,2003:An evolution of the Asian summer monsoon associated with mountain uplift—Simulation with the MRI atmosphere-ocean coupled GCM.J.Meteor.Soc.Japan,81,909-933.

An,C.-B.,L.Y.Tang,L.Barton,and F.H.Chen,2005:Climate change and cultural response around 4000 cal yr B.P.in the western part of Chinese Loess Plateau.Quaternary Research, 63,347-352.

An,Z.S.,2000:The history and variability of the East Asian paleomonsoon climate.Quaternary Science Reviews,19,171-187.

An,Z.S.,and Coauthors,1999:Eolian evidence from the Chinese Loess Plateau:The onset of the late Cenozoic great glaciation in the Northern Hemisphere and Qinghai-Xizang Plateau uplift forcing.Science in China(D),42,258-271.

An,Z.S.,J.E.Kutzbach,W.L.Prell,and S.C.Porter,2001:Evolution of Asian monsoons and phased uplift of the Himalaya-Tibetan plateau since Late Miocene times.Nature,411,62-66.

Bartlein,P.J.,and Coauthors,2011:Pollen-based continental climate reconstructions at 6 and 21 ka:A global synthesis.Climate Dyn.,37,775-802.

Berger,A.,1978:Long-termvariationsof dailyinsolationand quaternary climatic changes.J.Atmos.Sci.,35,2362-2367.

Boos,W.R.,and Z.Kuang,2010:Dominant control of the South Asian monsoon by orographic insulation versus plateau heating.Nature,463,218-222.

Broccoli,A.J.,and S.Manabe,1992:The effects of orography on midlatitude northern hemisphere dry climates.J.Climate,5, 1181-1201.

Bush,A.B.G.,2000:A positive climatic feedback mechanism for Himalayan glaciations.Quaternary International,65-66, 3-13.

Cane,M.A.,and Coauthors,2006:Progress in paleoclimate modeling.J.Climate,19,5031-5057.

Chaisson,W.P.,and A.C.Ravelo,2000:Pliocene development of east-west hydrographic gradient in the equatorial Pacif c.Paleoceanography,15,497-505.

Chandler,M.,D.Rind,and R.Thompson,1994:Joint investigations of the middle Pliocene climate II:GISS GCM northern hemisphere results.Global and Planetary Change,9,197-219.

Chen,F.H.,J.Bloemendal,J.M.Wang,J.J.Li,and F.Oldf eld, 1997:High-resolution multi-proxy climate records from Chinese loess:Evidence for rapid climatic changes over the last 75 kyr.Palaeogeography,Palaeoclimatology,Palaeoecology, 130,323-335.

Chen,X.,G.Yu,and J.Liu,2000:A preliminary simulation of climate at 21 ka BP in China.Journal of Lake Sciences,12, 154-164.(in Chinese)

Chen,X.,G.Yu,and J.Liu,2001:An AGCM+SSiB model simulation on changes in palaeomonsoon climate at 21 ka BP in China.Acta Meteorologica Sinica,15,333-345.

Chen,X.,G.Yu,and J.Liu,2002:Mid-Holocene climate simulation and discussion on the mechanism of temperature changes in eastern Asia.Science in China(D),32,335-345.(in Chinese)

Chu,K.-C.,1973:A preliminary study on the climatic f uctuations during the last 5,000 years in China.Science in China(A), 16,226-256.

CLIMAP Project members,1981:Seasonal reconstructions of the Earth’s surface at the last glacial maximum.Geological Society of America,Map and Chart Series,Vol.36,Geol.Soc.of Amer.,Boulder,Colorado,18 pp.

Crowley,T.J.,2000:Causes of climate change over the past 1000 years.Science,289,270-277.

DeMenocal,P.B.,and D.Rind,1993:Sensitivity of Asian and African climate to variation in seasonal insolation,glacial ice cover,sea surface temperature,and Asian orography.J.Geophys.Res.,98(D4),7265-7287.

Ding,Z.L.,J.M.Sun,T.S.Liu,R.X.Zhou,S.L.Yang,and B. Guo,1998:Wind-blown origin of the Pliocene red clay formation in the Central Loess Plateau,China.Earth and Planetary Science Letters,161,135-143.

Ding,Z.L.,S.F.Xiong,J.M.Sun,S.L.Yang,Z.Y.Gu,and T.S. Liu,1999:Pedostratigraphy and paleomagnetism of a～7.0 Ma eolian loess-red clay sequence at Lingtai,Loess Plateau, north-central China and the implications for paleomonsoon evolution.Palaeogeography,Palaeoclimatology,Palaeoecology,152,49-66.

Ding,Z.L.,E.Derbyshire,S.L.Yang,J.M.Sun,and T.S. Liu,2005:Stepwise expansion of desert environment across northern China in the past 3.5 Ma and implications for monsoon evolution.EarthandPlanetaryScience Letters,237,45-55.

Dowsett,H.,J.Barron,R.Poore,R.Thompson,T.Cronin,S.Ishman,and D.Willard,1999:Middle Pliocene paleoenvironmental reconstruction:PRISM2.USGS Open f le Report 99-535.[Available online at http://pubs.usgs.gov/openf le/of99-535]

Farrera,I.,and Coauthors,1999:Tropical climates at the last glacial maximum:A new synthesis of terrestrial palaeoclimate data.I.Vegetation,lake-levels and geochemistry.Climate Dyn.,15,823-856.

Fl¨ohn,H.,1956:Der indische sommermonsun als Glied der planetarischen Zirkulation der atmosphere.Berichte des Deutichen Wetterdienstes,22,134-139.

Fl¨ohn,H.,1968:Contributions to a meteorology of the Tibetan highlands.Atmospheric Science Paper,No.130,Department of Atmospheric Science,Colorado State University,122 pp.

Fluteau,F.,G.Ramstein,and J.Besse,1999:Simulating the evolution of the Asian and African monsoons during the past 30 Myr using an atmospheric general circulation model.J.Geophys.Res.,104(D10),11 995-12 018.

Fort,M.,1996:Late Cenozoic environmental changes and uplift on the northern side of the central Himalaya:A reappraisal from f eld data.Palaeogeography,Palaeoclimatology,Palaeoecology,120,123-145.

Ge,Q.S.,J.Zheng,X.Fang,Z.Man,X.Zhang,P.Zhang,and W.-C.Wang,2003:Winter half-year temperature reconstruction for the middle and lower reaches of the Yellow River and Yangtze River,China,during the past 2000 years.The Holocene,13,933-940.

Gou,X.H.,J.Peng,F.Chen,M.Yang,D.F.Levia,and J.Li, 2008:A dendrochronological analysis of maximum summer half-year temperature variations over the past 700 years on the northeastern Tibetan Plateau.Theor.Appl.Climatol.,93, 195-206.

Guiot,J.,H.B.Wu,W.Y.Jiang,and Y.L.Luo,2008:East Asian monsoon and paleoclimatic data analysis:A vegetation point of view.Climate of the Past,4,137-145.

Guo,Z.,and T.J.Zhou,2013:Why does FGOALS-gl reproducea weak medieval warm period but a reasonable little ice age and 20th century warming?Adv.Atmos.Sci.,30,1758-1770, doi:10.1007/s00376-013-2227-8.

Guo,Z.T.,D.S.Liu,and Z.S.An,1994:Paleosol in the loess deposition in Weinan and the environments since the past 150,000 yr B.P..Quaternary Sciences,14,256-269.(in Chinese)

Guo,Z.T.,and Coauthors,2002:Onset of Asian desertif cation by 22 Myr ago inferred from loess deposits in China.Nature, 416,159-163.

Guo,Z.T.,S.Z.Peng,Q.Z.Hao,P.E.Biscaye,Z.An,and T. S.Liu,2004:Late Miocene-Pliocene development of Asian aridif cationasrecorded inthered-earth formationinnorthern China.Global and Planetary Change,41,135-145.

Guo,Z.T.,and Coauthors,2008:A major reorganization of Asian climatebytheearlyMiocene.Climateof thePast,4,153-174.

Hahn,D.G.,and S.Manabe,1975:The role of mountains in the south Asian monsoon circulation.J.Atmos.Sci.,32,1515-1541.

Halley,E.,1686:An historical account of the trade winds and monsoons observable in the seas between and near the tropics with an attempt to assign the physical cause of the said wind.Philosophical Transactions of the Royal Society London,16, 153-168.

Han,J.,W.S.Fyfe,F.J.Longstaffe,H.C.Palmer,F.H. Yan,and X.S.Mai,1997:Pliocene-Pleistocene climatic change recorded in f uviolacustrine sediments in central China.Palaeogeography,Palaeoclimatology,Palaeoecology, 135,27-39.

Han,Y.,P.Zhao,and G.B.Zhou,2009:Modeling of impacts of regional vegetation change in China on East Asian summer rainfall under the background of the last glacial maximum.Quaternary Sciences,29,1071-1077.(in Chinese)

Hansen,J.,M.Sato,A.Lacis,R.Ruedy,I.Tegen,and E. Matthews,1998:Climate forcings in the industrial era.Proceedings of the National Academy of Sciences of the United States of America,95,12 753-12 758.

Harris,N.,2006:The elevation history of the Tibetan Plateau and its implications for the Asian monsoon.Palaeogeography Palaeoclimatology Palaeoecology,241,4-15.

Harrison,T.M.,P.Copeland,W.S.F.Kidd,and A.Yin,1992: Raising Tibet.Science,255,1663-1670.

Haywood,A.M.,P.J.Valdes,and B.W.Sellwood,2000:Global scale palaeoclimate reconstruction of the middle Pliocene climateusing theUKMO GCM:Initial results.Global and Planetary Change,25,239-256.

Huang,R.H.,J.L.Chen,L.T.Zhou,and Q.Y.Zhang,2003: Studies on the relationship between the severe climatic disasters in China and the East Asia climate system.Chinese J. Atmos.Sci.,27,770-787.(in Chinese)

Jansen,E.,and Coauthors,2007:Palaeoclimate.Climate Change 2007:The Physical Science Basis.Contribution of Working Group I to the Fourth Assessment Report of the Intergovernmental Panel on Climate Change,Solomon et al.,Eds.,Cambridge University Press,Cambridge,United Kingdom and New York,NY,USA,434-497.

Jian,Z.,Y.Yu,B.Li,J.Wang,X.Zhang,and Z.Zhou,2006: Phased evolution of the south-north hydrographic gradient in the South China Sea since the middle Miocene.Palaeogeography,Palaeoclimatology,Palaeoecology,230,251-263.

Jiang,D.B.,2008:Vegetation and soil feedbacks at the last glacial maximum.Palaeogeography,Palaeoclimatology,Palaeoe-cology,268,39-46.

Jiang,D.B.,2009:Numerical simulation of the middle Pliocene climate in China.Quaternary Sciences,29,1033-1043.(in Chinese)

Jiang,D.B.,2013:Vegetation feedback at the mid-Pliocene.Atmos.Oceanic Sci.Lett.,6,320-323.

Jiang,D.B.,and X.M.Lang,2010:Last glacial maximum East Asian monsoon:Results of PMIP simulations.J.Climate,23, 5030-5038.

Jiang,D.B.,H.J.Wang,H.Drange,and X.M.Lang,2003:Last glacial maximum over China:Sensitivities of climate to paleovegetation and Tibetan ice sheet.J.Geophys.Res.,108(D3), 4102,doi:10.1029/2002JD002167.

Jiang,D.B.,H.J.Wang,and X.M.Lang,2004:On the possibility of ice sheet over the Tibetan Plateau at the last glacial maximum.Chinese J.Atmos.Sci.,28,1-6.(in Chinese)

Jiang,D.B.,H.J.Wang,Z.L.Ding,X.M.Lang,and H.Drange, 2005:Modeling the middle Pliocene climate with a global atmospheric general circulation model.J.Geophys.Res.,110, D14107,doi:10.1029/2004JD005639.

Jiang,D.B.,Z.L.Ding,H.Drange,and Y.Q.Gao,2008:Sensitivity of East Asian climate to the progressive uplift and expansion of the Tibetan Plateau under the mid-Pliocene boundary conditions.Adv.Atmos.Sci.,25,709-722,doi: 10.1007/s00376-008-0709-x.

Jiang,D.B.,X.M.Lang,Z.P.Tian,and D.L.Guo,2011: Last glacial maximum climateover China from PMIP simulations.Palaeogeography,Palaeoclimatology,Palaeoecology, 309,347-357.

Jiang,D.B.,X.M.Lang,Z.P.Tian,and T.Wang,2012:Considerable model-data mismatch in temperature over China during the mid-Holocene:Results of PMIP simulations.J.Climate, 25,4135-4153.

Jiang,D.B.,X.M.Lang,Z.P.Tian,and L.X.Ju,2013a:Mid-Holocene East Asian summer monsoon strengthening:Insights from Paleoclimate Modeling Intercomparison Project (PMIP)simulations.Palaeogeography,Palaeoclimatology,Palaeoecology,369,422-429.

Jiang,D.B.,Z.P.Tian,and X.M.Lang,2013b:Mid-Holocene net precipitation changes over China:Model-data comparison.Quaternary Science Reviews,82,104-120.

Jin,L.,A.Ganopolski,F.Chen,M.Claussen,and H.J.Wang, 2005:Impacts of snow and glaciers over Tibetan Plateau on Holocene climatechange:Sensitivityexperiments withacoupled model of intermediate complexity.Geophys.Res.Lett., 32,L17709,doi:10.1029/2005GL023202.

Jin,L.,Y.Peng,F.Chen,and A.Ganopolski,2009:Modeling sensitivity study of the possible impact of snow and glaciers developing over Tibetan Plateau on Holocene African-Asian summer monsoon climate.Climate of the Past,5,457-469.

Joos,F.,S.Gerber,I.C.Prentice,B.L.Otto-Bliesner,and P.J. Valdes,2004:Transient simulations of Holocene atmospheric carbon dioxide and terrestrial carbon since the last glacial maximum.Global Biogeochemical Cycles,18,GB2002,doi: 10.1029/2003GB002156.

Joussaume,S.,and K.E.Taylor,1995:Status of the Paleoclimate Modeling Intercomparison Project(PMIP).Proceedings of the First International AMIP Scientif c Conference,WCRP-92,WMO/TD-732,W.L.Gates,Eds.,World Meteorological Organization,Geneva,425-430.

Ju,L.X.,H.J.Wang,and D.B.Jiang,2007:Simulation of the last glacial maximum climate over East Asia with a re-gional climate model nested in a general circulation model.Palaeogeography,Palaeoclimatology,Palaeoecology,248, 376-390.

Kitoh,A.,2004:Effects of mountain uplift on East Asian summer climate investigated by a coupled atmosphere-ocean GCM.J. Climate,17,783-802.

Kutzbach,J.E.,P.J.Guetter,W.F.Ruddiman,and W.L.Prell, 1989:Sensitivity of climate to late Cenozoic uplift in southern Asia and the American West:Numerical experiments.J. Geophys.Res.,94(D15),18 393-18 407.

Kutzbach,J.E.,W.L.Prell,and W.F.Ruddiman,1993:Sensitivityof Eurasian climateto surface uplift of the Tibetan plateau.Journal of Geology,101,177-190.

Lambeck,K.,and J.Chappell,2001:Sea level change through the last glacial cycle.Science,292,679-686.

Li,J.J.,and X.M.Fang,1999:Uplift of the Tibetan Plateau and environmental changes.Chinese Science Bulletin,44,2117-2124.

Li,J.J.,and Coauthors,1997:Late Cenozoic magnetostratigraphy (11-0 Ma)of the Dongshanding and Wangjiashan sections in the Longzhong Basin,western China.Geologie en Mijnbouw, 76,121-134.

Li.,X.Q.,C.S.Li,H.Y.Lu,J.R.Dodson,and Y.F.Wang,2004: Paleovegetation and paleoclimate in middle-late Pliocene, Shanxi,central China.Palaeogeography,Palaeoclimatology,Palaeoecology,210,57-66.

Li,Y.F.,and S.P.Harrison,2008:Simulations of the impact of orbital forcing and ocean on the Asian summer monsoon during the Holocene.Global and Planetary Change,60,505-522.

Li,Y.F.,S.P.Harrison,P.Zhao,and J.Ju,2009:Simulations of the impacts of dynamic vegetation on interannual and interdecadal variability of Asian summer monsoon with modern and mid-Holocene orbital forcings.Global and Planetary Change,66,235-252.

Liu,J.,G.Yu,and X.Chen,2002a:Palaeoclimate simulation of 21 ka for the Tibetan Plateau and eastern Asia.Climate Dyn., 19,575-583.

Liu,J.,X.Chen,S.Wang,and Y.Zheng,2004:Palaeoclimate simulation of Little Ice Age.Progress in Natural Science,14, 716-724.

Liu,J.,H.Storch,X.Chen,E.Zorita,J.Zheng,and S.Wang,2005: Simulatedand reconstructed winter temperature inthe eastern China during the last millennium.Chinese Science Bulletin, 50,2872-2877.

Liu,J.,B.Wang,Q.Ding,X.Kuang,W.Soon,and E.Zorita, 2009a:Centennial variations of global monsoon precipitation in the last millennium:results from the ECHO-G model.J. Climate,22,2356-2371.

Liu,J.,B.Wang,H.Wang,X.Kuang,and R.Ti,2011:Forced response of the East Asian summer rainfall over the past millennium:Results from a coupled model simulation.Climate Dyn.,36,323-336.

Liu,T.S.,X.S.Zhang,S.F.Xiong,and X.G.Qin,1999:Qinghai-Xizang Plateau glacial environment and global cooling.Quaternary Sciences,19,385-396.(in Chinese)

Liu,T.S.,X.S.Zhang,S.F.Xiong,X.G.Qin,and X.P.Yang, 2002b:Glacial environments on the Tibetan Plateau and global cooling.Quaternary International,97-98,133-139.

Liu,X.D.,and Z.Y.Yin,2002:Sensitivity of East Asian monsoon climate to the uplift of the Tibetan Plateau.Palaeogeography,Palaeoclimatology,Palaeoecology,183,223-245.

Liu,X.D.,L.Li,and Z.S.An,2001:Tibetan Plateau uplift and drying in Eurasian interior and northern Africa.Quaternary Sciences,21,114-122.(in Chinese)

Liu,X.D.,J.E.Kutzbach,Z.Y.Liu,Z.S.An,and L.Li,2003a: The Tibetan Plateau as amplif er of orbital-scale variability of the East Asian monsoon.Geophys.Res.Lett.,30,1839,doi: 10.1029/2003GL017510.

Liu,Y.,J.H.He,W.L.Li,and L.X.Chen,2007a:MM5 simulations of the China regional climate of the LGM I:Inf uences of changes of CO2and earth orbit changes.Acta Meteorologica Sinica,65,139-150.(in Chinese)

Liu,Y.,J.H.He,W.L.Li,and L.X.Chen,2007b:MM5 simulations of the China regional climate of the LGM II:Inf uences of changes of land area,vegetation,and large-scale circulationbackground.ActaMeteorologica Sinica,65,151-159.(in Chinese)

Liu,Y.,J.H.He,W.L.Li,L.X.Chen,W.Li,and B.Zhang,2009b: MM5 simulations of the China regional climate during the mid-Holocene.Acta Meteorologica Sinica,24,468-483.

Liu,Z.,J.Kutzbach,and L.Wu,2000:Modeling climate shift of El Nino variability in the Holocene.Geophys.Res.Lett.,27, 2265-2268.

Liu,Z.,B.Otto-Bliesner,J.Kutzbach,L.Li,and C.Shields, 2003b:Coupled climate simulation of the evolution of global monsoons in the Holocene.J.Climate,16,2472-2490.

Liu,Z.,and Coauthors,2009c:Deglaciation with a new mechanism for B?lling-Aller?d warming.Science,325,310-314.

Man,W.M.,and T.J.Zhou,2011:Forced response of atmospheric oscillations during the last millennium simulated by a climate system model.Chinese Science Bulletin,56,3042-3052.

Man,W.M.,and T.J.Zhou,2014:Regional-scale surface air temperature and East Asian summer monsoon changes during the last millennium simulated by the FGOALS-gl climate system model.Adv.Atmos.Sci.,31,765-778,doi:10.1007/s00376-013-3123-y.

Man,W.M.,T.J.Zhou,J.Zhang,C.Q.Wu,and B.Wu,2010:The equilibrium response of LASG/IAP climate system model to prescribed external forcing during the little ice age.Chinese J.Atmos.Sci.,34,914-924.(in Chinese)

Man,W.M.,T.J.Zhou,and J.H.Jungclaus,2012:Simulation of the East Asian summer monsoon during the last millennium with the MPI earth system model,J.Climate,25,7852-7866.

Man,W.M.,T.J.Zhou,J.H.Jungclaus,2014:Effects of large volcanic eruptions on global summer climate and East Asian monsoon changes during the last millennium:Analysis of MPI-ESM simulations,J.Climate,doi:http://dx.doi.org/ 10.1175/JCLI-D-13-00739.1.

Manabe,S.,and T.B.Terpstra,1974:The effects of mountains on the general circulation of the atmosphere as identif ed by numerical experiments.J.Atmos.Sci.,31,3-42.

Manabe,S.,and A.J.Broccoli,1990:Mountains and arid climates of middle latitudes.Science,247,192-194.

Meehl,G.A.,and Coauthors,2007:Global climate projections.Climate Change 2007:The Physical Science Basis.Contribution of Working Group I to the Fourth Assessment Report of the Intergovernmental Panel on Climate Change,Solomon et al.,Eds.,Cambridge University Press,Cambridge,United Kingdom and New York,NY,USA,748-845.

Mix,A.C.,A.E.More,and N.G.Pisias,1999:Foraminiferal faunal estimates of paleotemperature:circumventing the noanalog problem yields cool ice age tropics.Paleoceanography,14,350-359.

Molnar,P.,1989:The geographic evolution of the Tibetan Plateau.American Scientist,77,350-360.

Molnar,P.,1997:The rise of the Tibetan plateau:From mantle dynamics to the Indian monsoon.Astronomy&Geophysics, 38,10-15.

Peltier,W.R.,1994:Ice age paleotopography.Science,265,195-201.

Peltier,W.R.,2004:Global glacial isostasy and the surface of the ice-age earth:The ICE-5G(VM2)model and GRACE.Annual Review of Earth and Planetary Sciences,32,111-149.

Peng,Y.B.,Y.Xu,and L.Y.Jin,2009:Climate changes over eastern China during the last millennium in simulations and reconstructions.Quaternary International,208,11-18.

Prell,W.L.,and J.E.Kutzbach,1997:The impact of Tibet-Himalayan elevation on the sensitivity of the monsoon climate system to changes in solar radiation.Tectonic Uplift and Climate Change,W.F.Ruddiman,Eds.,Plenum Publishing Corporation,New York,171-201.

Qian,Y.F.,H.Yan,Q.Q.Wang,and A.Y.Wang,1988:Numerical Study of the Effect of Topography on Planetary Atmosphere. Science Press,Beijing,119-147.(in Chinese)

Qian,Y.,Y.F.Qian,and Y.C.Zhang,1998:Study on scenarios and mechanism of the regional climate change of East Asia in the last ice-age.Chinese J.Atmos.Sci.,22,283-293.(in Chinese)

Qin,B.Q.,and G.Yu,1998:Implications of lake level f uctuations at 6 ka and 18 ka in mainland Asia.Global and Planetary Change,18,59-72.

Ramstein,G.,F.Fluteau,J.Besse,and S.Joussaume,1997:Effect oforogeny,platemotionandland-sea distributiononEurasian climate change over the past 30 million years.Nature,386, 788-795.

Raymo,M.E.,and W.F.Ruddiman,1992:Tectonic forcing of late Cenozoic climate.Nature,359,117-122.

Raymo,M.E.,B.Grant,M.Horowitz,and G.H.Rau,1996:Mid-Pliocene warmth:Stronger greenhouse and stronger conveyor.Marine Micropaleontology,27,313-326.

Ruddiman,W.F.,and J.E.Kutzbach,1990:Late Cenozoic plateau uplift and climate change.Transactions of the Royal Society of Edinburgh:Earth Sciences,81,301-314.

Ruddiman,W.F.,W.L.Prell,and M.E.Raymo,1989:Late Cenozoic uplift in southern Asia and the American West:Rational for general circulation modeling experiments.J.Geophys. Res.,94(D15),18 379-18 391.

Shao,X.M.,L.Huang,H.Liu,E.Liang,X.Fang,and L.Wang, 2005:Reconstruction of precipitation variation from tree rings in recent 1000 years in Delingha,Qinghai.Science in China(D),48,939-949.

Shen,C.,W.-C.Wang,Y.Peng,Y.Xu,and J.Zheng,2009:Variability of summer precipitation over eastern China during the last millennium.Climate of the Past,5,129-141.

Shi,Y.F.,and G.Yu,2003:Warm-humid climate and transgressions during 40-30 ka B.P.and the potential mechanisms.Quaternary Sciences,23,1-11.(in Chinese)

Shi,Y.F.,and Coauthors,1993:Mid-Holocene climates and environments in China.Global and Planetary Change,7,219-233.

Shi,Y.F.,B.X.Zheng,and T.D.Yao,1997:Glaciers and environments during the last glacial maximum(LGM)on the Tibetan Plateau.Journal of Glaciology and Geocryology,19,97-113. (in Chinese)

Shi,Y.F.,G.Yu,X.D.Liu,B.Y.Li,and T.D.Yao,2001:Reconstruction of the 30-40 ka BP enhanced Indian monsoon climate based on geological records from the Tibetan Plateau.Palaeogeography,Palaeoclimatology,Palaeoecology,169, 69-83.

Shi,Z.G.,X.Yan,C.Yin,and Z.Wang,2007:Effects of historical land cover changes on climate.Chinese Science Bulletin,52, 2575-2583.

Sloan,L.C.,T.J.Crowley,and D.Pollard,1996:Modeling of middle Pliocene climate with the NCAR GENESIS general circulation model.Marine Micropaleontology,27,51-61.

Sui,W.H.,and P.Zhao,2005:Modeling the impact of the western pacif c sea surface temperature difference on the Asian summer monsoon climate at the last glacial maximum.Quaternary Sciences,25,645-654.(in Chinese)

Sun,D.H.,R.Su,J.Bloemendal,and H.Lu,2008:Grain-size and accumulation rate records from LateCenozoic aeolian sequences in northern China:Implications for variations in the East Asian winter monsoon and westerly atmospheric circulation.Palaeogeography,Palaeoclimatology,Palaeoecology, 264,39-53.

Sun,X.J.,and P.X.Wang,2005:How old is the Asian monsoon system?—Palaeobotanical records from China.Palaeogeography,Palaeoclimatology,Palaeoecology,222,181-222.

Sun,Y.,G.Ramstein,C.Contoux,and T.J.Zhou,2013:A comparative study of large-scale atmospheric circulation in the context of a future scenario(RCP4.5)and past warmth(mid-Pliocene).Climate of the Past,9,1613-1627.

Tan,M.,X.Shao,J.Liu,and B.Cai,2009:Comparative analysis between a proxy-based climate reconstruction and GCM-based simulation of temperatures over the last millennium in China.Journal of Quaternary Science,24,547-551.

Tian,J.,P.X.Wang,and X.R.Cheng,2004:Development of the EastAsian monsoon and Northern Hemisphere glaciation: Oxygen isotope records from the South China Sea.Quaternary Science Reviews,23,2007-2016.

Tian,Z.P.,and D.B.Jiang,2013:Mid-Holocene ocean and vegetation feedbacks over East Asia.Climate of the Past,9,2153-2171.

Trenberth,K.E.,and Coauthors,2007:Observations:Surface and Atmospheric Climate Change.Climate Change 2007:The Physical Science Basis.Contribution of Working Group I to the Fourth Assessment Report of the Intergovernmental Panel on Climate Change,Solomon et al.,Eds.,Cambridge University Press,Cambridge,United Kingdom and New York,NY, USA,236-336.

UNESCO(United Nations Educational,Scientif cand Cultural Organization),1979:Map of the World Distribution of Arid Regions.Map at scale 1:25,000,000 with explanatory note. UNESCO,Paris,54 pp.

Wan,S.,A.Li,P.D.Clift,and J.W.Stuut,2007:Development of the East Asian monsoon:Mineralogical and sedimentologic records in the northern South China Sea since 20 Ma.Palaeogeography,Palaeoclimatology,Palaeoecology, 254,561-582.

Wang,B.,and Q.Ding,2008:Global monsoon:The dominant mode of the annual variations of the global tropical precipitation and circulation.Dyn.Atmos.Oceans,44,165-183.

Wang,C.,and Coauthors,2008a:Constraints on the early uplift history of the Tibetan Plateau.Proceedings of the National Academy of Sciences of the United States of America,105, 4987-4992.

Wang,H.J.,1999a:Role of vegetation and soil in the Holocene megathermal climate over China.J.Geophys.Res.,104(D8),9361-9367.

Wang,H.J.,2000:The seasonal climate and low frequency oscillation in the simulated mid-Holocene megathermal climate.Adv.Atmos.Sci.,17,445-457,doi:10.1007/s00376-000-0035-4.

Wang,H.J.,2002:The mid-Holocene climate simulated by a gridpoint AGCM coupled with a biome model.Adv.Atmos.Sci., 19,205-218,doi:10.1007/s00376-002-0017-9.

Wang,H.J.,and Q.C.Zeng,1992a:Numerical simulation of the climate 9000 years ago.Chinese J.Atmos.Sci.,16,313-321. (in Chinese)

Wang,H.J.,and Q.C.Zeng,1992b:The numerical simulation of the ice age climate.Acta Meteorologica Sinica,50,279-289. (in Chinese)

Wang,L.,1994:Seasurface temperature historyof thelow latitude western Pacif c during the last 5.3 million years.Palaeogeography,Palaeoclimatology,Palaeoecology,108,379-436.

Wang,P.X.,1999b:Response of western Pacif c marginal seas to glacial cycles:Paleoceanographic and sedimentological features.Marine Geology,156,5-39.

Wang,P.X.,S.Clemens,L.Beaufort,P.Braconnot,G.Ganssen, Z.M.Jian,P.Kershaw,and M.Sarnthein,2005:Evolution and variability of the Asian monsoon system:State of the art and outstanding issues.Quaternary Science Reviews,24, 595-629.

Wang,S.W.,2009:Holocene cold events in the North Atlantic: Chronology and climatic impact.Quaternary Sciences,29, 1146-1153.(in Chinese)

Wang,S.W.,T.J.Zhou,J.Cai,J.Zhu,Z.Xie,and D.Gong,2004: Abrupt climate change around 4 ka BP:Role of the thermohaline circulation as indicated by a GCM experiment.Adv. Atmos.Sci.,21,291-295,doi:10.1007/BF02915714.

Wang,S.W.,X.Y.Wen,Y.Luo,W.J.Dong,Z.C.Zhao,and B. Yang,2007:Reconstruction of temperature series of China for the last 1000 years.Chinese Science Bulletin,52,3272-3280.

Wang,S.W.,J.B.Huang,X.Y.Wen,and J.H.Zhu,2008b:Evidence and modeling study of droughts in China during 4-2 ka BP.Chinese Science Bulletin,53,2215-2221.

Wang,S.W.,J.B.Huang,X.Y.Wen,B.Yang,and G.Y.Ren, 2009a:Two modes of summer precipitation variability of Holocene in China.Quaternary Sciences,29,1086-1094.(in Chinese)

Wang,T.,H.J.Wang,and D.B.Jiang,2010:Mid-Holocene East Asian summer climate as simulated by the PMIP2 models.Palaeogeography,Palaeoclimatology,Palaeoecology,288, 93-102.

Wang,Y.,and T.Deng,2005:A 25 m.y.isotopic record of paleodiet and environmental change from fossil mammals and paleosols from the NE margin of the Tibetan Plateau.Earth and Planetary Science Letters,236,322-338.

Wang,Y.,Z.M.Jian,and P.Zhao,2009b:Sensitivity of equatorial Pacif c convergence zone to revised tropical sea surface temperature outside the warm pool during the last glacial maximum.Quaternary Sciences,29,221-231.(in Chinese)

Wehausen,R.,and H.Brumsack,2002:Astronomical forcing of the East Asian monsoon mirrored by the composition of Pliocene South China Sea sediments.Earth and Planetary Science Letters,201,621-636.

Wei,J.F.,and H.J.Wang,2004:A possible role of solar radiation and ocean in the mid-Holocene East Asian monsoon climate.Adv.Atmos.Sci.,21,1-12,doi:10.1007/BF02915675.

Wu,G.X.,Y.M.Liu,B.He,Q.Bao,A.M.Duan,and F.-F.Jin, 2012:Thermal controls on the Asian summer monsoon.Scientif c Reports,2,404,doi:10.1038/srep00404.

Wu,N.,Y.Pei,H.Lu,Z.Guo,F.Li,and T.S.Liu,2006:Marked ecological shifts during 6.2-2.4 Ma revealed by a terrestrial molluscan record from the Chinese red clay formation and implication for palaeoclimatic evolution.Palaeogeography,Palaeoclimatology,Palaeoecology,233,287-299.

Yang,B.,A.Braeuning,K.R.Johnson,and Y.F.Shi,2002:General characteristics of temperature variation in China during the last two millennia.Geophys.Res.Lett.,29,381-384,doi: 10.1029/2001GL014485.

Yang,B.,Y.F.Shi,A.Braeuning,and J.Wang,2004:Evidence for a warm-humid climate in arid northwestern China during 40-30 ka BP.Quaternary Science Reviews,23,2537-2548.

Yang,M.X.,T.D.Yao,H.J.Wang,and X.H.Gou,2006:Climatic oscillations over the past 120 kyr recorded in the Guliya ice core,China.Quaternary International,154-155,11-18.

Yao,T.D.,L.G.Thompson,Y.F.Shi,K.Q.Jiao,and X.P.Zhang, 1997:A study on the climate changes from Guliya ice core records since last interglacial period.Science in China(D), 27,447-452.(in Chinese)

Yin,C.H.,X.D.Yan,Z.G.Shi,and Z.M.Wang,2007:Simulation of the climatic effects of natural forcings during the pre-industrial era.Chinese Science Bulletin,52,1545-1558.

Yu,G.,and Coauthors,2000:Palaeovegetation of China:A pollen data-based synthesis for the mid-Holocene and last glacial maximum.Journal of Biogeography,27,635-644.

Yu,G.,X.Chen,J.Liu,and S.Wang,2001:Preliminary study on LGM climate simulation and the diagnosis for East Asia.Chinese Science Bulletin,46,364-368.

Yu,G.,B.Xue,J.Liu,and X.Chen,2003a:LGM lake records from China and an analysis of the climate dynamics using a modelling approach.Global and Planetary Change,38,223-256.

Yu,G.,G.Y.Lai,J.Liu,and Y.F.Shi,2003b:Late MIS 3 climate simulations.Quaternary Sciences,23,12-24.(in Chinese)

Yu,G.,Y.Q.Zheng,and X.K.Ke,2005:35 ka B.P.climate simulations in East Asia and probing the mechanisms of climate changes.Chinese Science Bulletin,50,58-67.

Yu,G.,F.Gui,Y.F.Shi,and Y.Q.Zheng,2007:Late marine isotope stage 3 palaeoclimate for East Asia:A data and modelling comparison.Palaeogeography,Palaeoclimatology,Palaeoecology,250,167-183.

Yu,Y.,Z.Zhou,and X.Zhang,2003c:Impact of the closure of Indonesian seaway on climate:A numerical modeling study.Chinese Science Bulletin,48(SII),88-93.

Zhang,D.E.,1980:Winter temperature changes during the last 500 years in South China.Chinese Science Bulletin,25,497-500.

Zhang,J.,T.J.Zhou,W.M.Man,and Z.X.Li,2009:Thetransient simulation of Little Ice Age by LASG/IAP climate system model.Quaternary Sciences,29,1125-1134.(in Chinese)

Zhang,J.,L.Z.Li,and T.J.Zhou,2013a:Variation of surface temperature during the last millennium in a simulation with the FGOALS-gl climate system model.Adv.Atmos.Sci.,30, 699-712,doi:10.1007/s00376-013-2178-0.

Zhang,P.,and Coauthors,2008:A test of climate,sun,and culture relationships from an 1810-year Chinese cave record.Science,322,940-942.

Zhang,R.,andX.D.Liu,2010:Theeffects oftectonicupliftonthe evolution of Asian summer monsoon climate since Pliocene.Chinese Journal of Geophysics,53,948-960.

Zhang,R.,D.B.Jiang,X.D.Liu,and Z.P.Tian,2012a:Modeling the climate effects of different subregional uplifts within the Himalaya-Tibetan Plateau on Asian summer monsoon evolution.Chinese Science Bulletin,57,4617-4626.

Zhang,R.,and Coauthors,2013b:Mid-Pliocene East Asian monsoon climate simulated in the PlioMIP.Climate of the Past,9, 2085-2099.

Zhang,Z.S.,and Z.T.Guo,2005:Spatial character reconstruction of different periods in Oligocene and Miocene.Quaternary Sciences,25,523-530.(in Chinese)

Zhang,Z.S.,H.J.Wang,Z.T.Guo,and D.B.Jiang,2007a:Impacts of tectonic changes on the reorganization of the Cenozoicpaleoclimaticpatterns inChina.Earthand PlanetaryScience Letters,257,622-634.

Zhang,Z.S.,H.J.Wang,Z.T.Guo,and D.B.Jiang,2007b: What triggers the transition of palaeoenvironmental patterns in China,the Tibetan Plateau uplift or the Paratethys Sea retreat?Palaeogeography,Palaeoclimatology,Palaeoecology, 245,317-331.

Zhang,Z.S.,F.Flat?y,H.J.Wang,I.Bethke,M.Bentsen,and Z.T.Guo,2012b:Early Eocene Asian climate dominated by desert and steppe with limited monsoons.Journal of Asian Earth Sciences,44,24-35.

Zhao,P.,L.X.Chen,X.J.Zhou,Y.F.Gong,and Y.Han,2003: Modeling the East Asian climate during the last glacial maximum.Science in China(D),46,1060-1068.

Zhao,P.,X.J.Zhou,Z.M.Jian,M.Sparrow,and Y.Han,2004: Modeling the tropical climate and the impact of the western Pacif c sea surface temperature at the last glacial maximum.J.Geophys.Res.,109,D08105,doi:10.1029/2003JD004095.

Zhao,P.,R.H.Zhang,J.P.Liu,X.J.Zhou,and J.H.He,2007: Onset of southwesterly wind over eastern China and associated atmospheric circulation and rainfall.Climate Dyn.,28, 797-811.

Zhao,P.,X.Zhang,Y.F.Li,and J.M.Chen,2009:Remotely modulated tropical-north Pacif c ocean-atmosphere interactions by the South Asian high.Atmospheric Research,94,45-60.

Zhao,P.,S.Yang,and R.C.Yu,2010:Long-term changes in rainfall over eastern China and large-scale atmospheric circulation associated with recent global warming.J.Climate,23, 1544-1562.

Zheng,W.P.,and Y.Q.Yu,2009:The Asian monsoon system of the mid-Holocene simulated by a coupled GCM.Quaternary Sciences,29,1135-1145.(in Chinese)

Zheng,W.P.,P.Braconnot,E.Guilyardi,U.Merkel,and Y.Yu, 2008:ENSO at 6ka and 21ka from ocean-atmosphere coupled model simulations.Climate Dyn.,30,745-762.

Zheng,Y.Q.,G.Yu,S.M.Wang,B.Xue,H.Q.Liu,and X.M. Zeng,2003:Simulations of LGM climate of East Asia by regional climate model.Science in China(D),46,753-764.

Zheng,Y.Q.,G.Yu,S.M.Wang,B.Xue,D.Q.Zhuo,X.M.Zeng, and H.Q.Liu,2004:Simulation of paleoclimate over East Asia at 6 ka BP and 21 ka BP by a regional climate model.Climate Dyn.,23,513-529.

Zheng,Y.Q.,Z.C.Qian,H.R.He,H.P.Liu,X.M.Zeng,and G.Yu,2007:Simulations of water resource environmental changes in China during the last 20000 years by a regional climate model.Global and Planetary Change,55,284-300.

Zheng,Z.,and J.Guiot,1999:A 400000-year paleoclimate reconstruction in tropical region of China.Acta Scientiarum Naturalium Universitatis Sunyaseni,38,94-98.(in Chinese)

Zheng,Z.,and K.S.Zhou,1995:Discovery of pollen Sonneratia in Pleistocene strata in the coastal area of Guangdong Province.Acta Scientiarum Naturalium Universitatis Sunyatsen,34,88-92.(in Chinese)

Zhou,B.T.,and P.Zhao,2009:Inverse correlation between ancient winter and summer monsoons in East Asia?Chinese Science Bulletin,54,3760-3767.

Zhou,B.T.,andP.Zhao,2010:Modeling variationsof summer upper tropospheric temperature and associated climate over the Asian Pacif c region during the mid-Holocene.J.Geophys. Res.,115,D20109,doi:10.1029/2010JD014029.

Zhou,B.T.,and P.Zhao,2013:Simulating changes of spring Asian-Pacif c oscillation and associated atmospheric circulation in the mid-Holocene.International Journal of Climatology,33,529-538.

Zhou,B.T.,P.Zhao,J.He,and H.J.Wang,2004a:Modeling the impact of Australian Plate drift on the equatorial Pacif c.Quaternary Sciences,24,716-723.(in Chinese)

Zhou,B.T.,P.Zhao,Z.Jian,and J.He,2005:Modeling theimpact of Australian Plate drift on Southern Hemisphere climate and environment.Chinese Science Bulletin,50,1495-1502.

Zhou,T.J.,and R.Yu,2006:Twentieth century surface air temperature over China and the globe simulated by coupled climate models.J.Climate,19,5843-5858.

Zhou,T.J.,B.Wu,X.Wen,L.Li,and B.Wang,2008:A fast version of LASG/IAP climate system model and its 1000-year control integration.Adv.Atmos.Sci.,25,655-672,doi: 10.1007/s00376-008-0655-7.

Zhou,T.J.,B.Li,W.M.Man,L.X.Zhang,and J.Zhang,2011a: A comparison of the Medieval Warm period,Little Ice Age and 20th century warming simulated by the FGOALS climate system model.Chinese Science Bulletin,56,3028-3041.

Zhou,X.J.,P.Zhao,J.M.Chen,L.X.Chen,and W.L.Li,2009: Impacts of thermodynamic processes over the Tibetan Plateau on the Northern Hemispheric climate.Science in China(D), 52,1679-1693.

Zhou,X.J.,P.Zhao,G.Liu,and T.J.Zhou,2011b:Characteristics of decadal-centennial-scale changes in East Asian summer monsoon circulation and precipitation during the Medieval Warm Period and Little Ice Age and in the present day.Chinese Science Bulletin,56,3003-3011.

Zhou,Z.Y.,X.C.Jin,L.L.Wang,Z.M.Jian,and C.H.Xu, 2004b:Two closures of the Indonesian Seaway and its relationship to the formation and evolution of the western Pacif c warm pool.Marine Geology&Quaternary Geology,24,7-14.(in Chinese)

:Jiang,D.B.,and Coauthors,2015:Paleoclimate modeling in China:A review.Adv.Atmos.Sci.,32(2),250-275,

10.1007/s00376-014-0002-0.

(Received 5 May 2014;revised 10 July 2014;accepted 4 August 2014)

?Corresponding author:JIANG Dabang

Email:jiangdb@mail.iap.ac.cn