QUANTITATIVE CLIMATE RECONSTRUCTION IN THE ASIAN INTERIOR SINCE THE EOCENE
The aridification of the Asian interior is one of the most significant events during the Cenozoic.However,there is debate over its evolution and associated driving mechanisms.Pollen data,which can generate quantitative results and have sufficient spatio-temporal coverage,is of great value for constraining these issues.However,such pollen-based reconstruction is rarely available during the Cenozoic stage.In addition to the scarcity of long-term terrestrial records,the development of appropriate reconstruction methods is challenging.The taxon-based Modern Analogue Technique(MAT)is one of the most widely used approaches for reconstructing climate change during the Late Quaternary period,while it may suffer from'non-analogue'conditions in the pre-Quaternary times.Plant Functional Types(PFT),grouped by taxa with common phenology and similar bioclimatic space,can efficiently avoid this problem.In this study,we converted the taxa percentages to PFT scores based on the taxon-PFT-biome assignment of China.We then evaluated the performance of the PFT-based MAT model based on leave-one-out(LOO)cross-validation using a dataset comprising 1630 modern pollen samples in China.The results show that the model performed well in reconstructing both the annual mean precipitation(ANNP)(R2=0.93,RMSEP=154.34 mm)and annual mean temperature(ANNT)(R2=0.80,RMSEP=3.18 ℃),which demonstrate that the PFT-based MAT method is a potentially powerful tool for quantitative reconstructions of climate.We collected three well-dated pollen records across Northwest China.One of them came from Jingou River section(44° 10.4′N,85°27.4′E)from the Tienshan region,with an age range of 28.0～4.2 Ma and a resolution of 160 ka.Other two pollen records came from Huoshaogou and Laojunmiao sections(39°47′ N,97° 32′E)from the Jiuxi Basin,with an age range of 40.2～33.4 Ma and 13.00～2.21 Ma,respectively,and a resolution of 56 ka and 17 ka,respectively,and all the results were based on magnetostratigraphy.All the taxa percentages were converted to PFT scores.We used the PFT-based MAT method for palaeoclimate reconstructions of ANNP and ANNT since the Eocene.Our results demonstrate that the ANNP decreased from ca.948 mm to ca.84 mm,while the ANNT ranged between 1～10℃ since the Eocene.During 40～23 Ma,the climate was warmer and humid,with ANNP and ANNT ranging between 386～948 mm and 2.00～9.44 ℃,respectively.During 23～8 Ma,the ANNT ranged between-1.11～8.52 ℃,the ANNP decreased slightly and ranged between 187～953 mm.The ANNP increased to over 400 mm during 17～12 Ma and 10～9 Ma.Since ca.8 Ma,the climate was colder and drier,with ANNP and ANNT ranging between 84～434 mm and-2.84～8.85 ℃,respectively.It is important to note that our estimate is much lower than those estimated from the coexistence approach,as the latter lacks a statistical framework.Compared with existing environmental records,we infer that the retreat of the Paratethys Sea has limited effect on the aridification of the Asian interior during the Eocene.Furthermore,we infer that both the Tibet uplift and global cooling promote the ongoing aridity since ca.23 Ma.
Asian interiorquantitative reconstructionaridificationCenozoicpollenplant functional type
万方数据
维普
