Fifty years of Quaternary palynology in the Tibetan Plateau

来源 :中国科学:地球科学(英文版) | 被引量 : 0次 | 上传用户:zzdj1990
下载到本地 , 更方便阅读
声明 : 本文档内容版权归属内容提供方 , 如果您对本文有版权争议 , 可与客服联系进行内容授权或下架
论文部分内容阅读
Quaternary palynology in the Tibetan Plateau (TP) was initiated in the 1960s to meet the needs of economic development in western China.Pollen analysis was conducted for the first time on a 200-m long core of Quaternary lacustrine sediments taken from the main body of the TP in order to study pollen assemblages as well as vegetation and climate changes of glacial (cold)/interglacial (warm) periods.Pollen analysis of alpine snow and ice began at the first scientific expedition to the TP in the 1970s.After the 1980s,a series of international collaborative programs were carried out under Sino-French,Sino-German,Sino-Australian,and Sino-American cooperation,marking the integration of Chinese Quaternary palynology society with the international community.New methods for Quaternary palynology were gradually promoted and applied,changing the vege-tational and climatic interpretation of Quaternary palynology from qualitative to quantitative.Since the 1990s,many palynol-ogists have carried out extensive Quaternary palynological studies on fossilpollen sites of more than 60 lakes/sections and alpine glaciers in the TP to discuss the spatiotemporal vegetation changes and climatic and environmental evolution of the TP since the Pleistocene.Over the past half-century,Quaternary palynology in the TP has contributed to the establishment of the Chinese Quaternary pollen database and the study of vegetation and climate evolution since the Last Glacial Maximum (LGM) in the TP.Currently available pollen records revealed the spatial and temporal distribution of vegetation in the TP since the LGM,exhibiting expansions and shrinkages of forest,meadow,grassland and desert in different periods such as the LGM,the last deglaciation,and Holocene optimum period.The paleomonsoon reflected by paleovegetation since the LGM has undergone the changes of weak-strengthening-strong-weakening but still active-shrinking,which is mainly affected by solar insolation.
其他文献
在连续变量的量子信息理论中,玻色场中量子态的非Gauss性是一种重要的资源.近十年来,人们引进了多种基于距离和相对熵的非Gauss性度量,用来刻画非Gauss性的不同侧面.由于Gauss量子态的集合在辛变换所对应的酉操作下具有不变性,因此,我们希望非Gauss性的度量也具有这种辛变换下的不变性.基于Gauss态正好是一类量子不确定性关系的最小不确定性态,本文提出一个具有辛不变性的非Gauss性度量并揭示其基本性质.对玻色场中的一些重要量子态,本文计算其非Gauss性.
Stein方法是正态逼近的常用方法.本文介绍非线性期望下Stein方法的基本框架,以及如何利用Stein方法给出非线性期望下中心极限定理和大数定律的收敛速度.
本文介绍了近十年来关于随机微分方程Girsanov变换的路径独立性的主要研究成果.该性质因反映了金融数学中市场的有效性而具有明确的应用背景,也因等价于一些非线性偏微分方程而具有数学研究价值.本文分别就经典的随机微分方程、带跳的随机微分方程、随机偏微分方程和分布依赖随机微分方程,介绍Girsanov变换的路径独立性所联系的非线性偏微分方程,并对其他相关研究加以评注.最后,作为一个新结果,本文使用非线性偏微分方程刻画了随机微分方程解的路径无关性.
为了刻画金融时间序列的长记忆性和非平稳性,众多学者采用次分数Brown运动来描述金融资产价格变化的行为模式.然而,次分数Brown运动不是半鞅,能否直接将其应用于金融市场一直是金融数学领域的热点问题.基于Hurst指数H>1/2情形下次分数Brown运动Donsker逼近定理,本文研究几何次分数Brown运动框架下金融市场的套利问题.首先,采用Skorokhod拓扑下的随机游走理论,构建一个弱收敛于次分数Brown运动驱动的Black-Scholes市场模型.其次,利用积分不等式和次分数二元市场理论证明次
本文考虑保险人模糊厌恶的情形,针对一类包含多个常用准则的混合再保险保费准则,在连续时间模型框架下研究保险人的鲁棒最优再保险策略问题.本文通过最小化保险人折现破产概率和最大化保险人终端时刻期望财富效用,在Cramér-Lundberg跳模型和其扩散近似模型下,均得到了优化问题的值函数表达式和相应的鲁棒最优再保险策略.特别地,对于最小化保险人折现破产概率的优化问题,通过给出并证明优化问题的等价形式,从理论上直接证明值函数具有指数形式.此外,对于每个优化问题,本文通过巧妙地构造优化问题的辅助Lagrange函数
自从Ferguson的里程碑式的工作以来,非参数Bayes模型在统计和机器学习等领域中有着广泛的应用,近年来得到了蓬勃的发展.它的一个重要的理论基础是一个特殊的随机概率测度族,即Dirichlet过程.本文介绍Dirichlet过程的构造、性质、推广以及它在非参数Bayes估计问题中的应用.另外,本文也提到双参数Poisson-Dirichlet过程、Beta过程和更一般的断棍(stick-breaking)过程以及相关性质.
针对在Knight不确定性下的决策问题,本文基于非参数化的思想提出了一种新的概率模型之间差异的度量方法,以此规避参数化不确定性模型描述金融数据特征的损失.该度量方法对概率模型的形式没有限制,并且易于计算.在正态假设下,使用该度量方法的非参数化检验方法可以退化为一种更不易犯第一类错误的参数检验方法.本文将该度量方法应用于美国标准普尔500指数,发现了不确定性指标的双波峰特征,结果表明概率模型之间的差异捕捉到了独特的信息并且在金融市场中可能有特别的预警作用.本文得到的不确定性指标也可以被应用于研究不确定性与资
本文介绍一类系数是测度值的椭圆算子及其对应的扩散过程,这里考虑的测度值系数属于Kato类Kd,1.首先,介绍漂移项为测度值μ的Brown运动及反射Brown运动的弱解的存在唯一性以及弱解的转移概率密度函数估计.然后,给出关于无穷小算子∠=1/2△+μ·▽的二阶椭圆方程的Dirichlet边界值问题和Neumann边界值问题的弱解的存在唯一性及概率表示.最后,介绍漂移项为测度值μ的Brown运动的小时间渐近性和大偏差原理.
Let π be a self-dual irreducible cuspidal automorphic representation of GL2 (AQ) with trivial central character.Its Hecke eigenvalue λπ(n) is a real multiplicative function in n.We show that λπ(n) < 0 for some n << Q2/5π,where Qπ denotes (a special value
With the rapid development of information technologies and cloud computing,sensor networks play an increasingly important role in our society.Over the past few decades,distributed observer theory has attracted unprecedented attention due to its wide poten