Dust storms and Chinese loess sources over the last 22 million years

沙尘循环是气候的一个基本组成部分，但仍然是地球系统中最不为人所知的方面之一。大气中的尘埃影响海洋生产力和大气中的二氧化碳水平、云的形成以及大气吸收或反射的辐射量，所有这些都推动了气候变化。然而，人们对过去环境变化如何控制粉尘的产生和排放知之甚少，因此无法量化其对气候的影响，这一事实在最近的IPCC 2007年报告中得到了承认。这个项目通过试图确定世界上最大和最长的活动沙尘库——中国黄土高原的精确来源来解决这一差距。中国黄土高原包含2200万年的黄土（沉积尘埃）沉积档案，涵盖了了解地球历史最重要的时期之一。中国黄土对了解亚洲季风和沙尘暴的历史变化作出了重要贡献，是目前最有价值的气候档案之一。然而，它的解释依赖于这种尘埃的精确来源的详细限制，以及它到达黄土高原的途径。不幸的是，尽管利用大量样品的地球化学特性进行了调查，但在以下方面存在多种相互竞争的假设：1)精确的黄土源区，2)对粉尘产生的环境控制，3)来源是否随时间变化，4)大气机制是否用于粉尘运输，以及5)记录中最古老的部分（8-22 Ma）是否确实是风吹的。这些分歧严重限制了我们对这些沉积物起源的理解，使我们无法限制过去的大气、构造和海洋条件，这些条件负责灰尘的排放和运输，并破坏了黄土中某些气候代用物的使用。反过来，这限制了我们对过去在中国运行的主要环境过程的理解，以及当前大气环流系统的起源。克服这一差距的关键是合理约束黄土的来源。我们的试点工作表明，广泛应用的黄土和源区沉积物的块状沉积物地球化学分析将掩盖黄土粉尘多源的细节。只有某些重矿物的单个颗粒（> 2.8比重）可以作为来源诊断，因为每个颗粒将有一个来源，并且其某些地球化学和地质年代学特征可以对此进行诊断。因此，将使用多代理单粒地球化学方法对从黄土和邻近沙漠沉积物中分离出来的锆石和其他重矿物进行多粒单粒分析，以最大限度地提高成功的可能性。在我们的试点工作之前，这种方法还没有在中国黄土上进行过。这些分析的样本将取自中国多个典型的黄土序列，从而可以确定时间和空间上的源方差。采样将集中在关键的时间间隔，如西藏的隆起、冰期冰川的开始和最近时期冰川强度的增强。这些结果将使我们能够在相互矛盾的黄土沙尘来源和输运假设之间进行检验，从而约束沙尘排放的基本控制及其输运所涉及的大气机制。反过来，通过与独立记录的比较，这将使我们能够评估全球和区域气候、构造和海洋变化对尘埃的影响。这是目前知之甚少的地球系统的关键组成部分。最后，研究结果将检验质量堆积率和粒度等广泛使用的沉积指标在黄土中重建过去环境变化的有效性。

The dust cycle is a fundamental component of climate, but remains one of the least understood aspects of the Earth-system. Dust in the atmosphere affects oceanic productivity and atmospheric carbon dioxide levels, cloud formation and the amount of radiation absorbed or reflected by the atmosphere, all driving climatic change. However, how dust production and emission is controlled by environmental change in the past is poorly understood, preventing quantification of its effects on climate, fact acknowledged in the recent IPCC 2007 report. This project addresses this gap by attempting to pinpoint the precise sources of the world's largest and longest active dust sink, the Chinese Loess Plateau. The Chinese Loess Plateau contains a 22 million year archive of loess (deposited dust) deposition, covering one of the most significant time periods for understanding the Earth's history. Chinese loess has made crucial contributions to understanding past changes in the Asian monsoon and dust storms and is one of the most valuable climate archives available. However, its interpretation relies on detailed constraint of the precise sources of this dust, and the pathways by which it reaches the Loess Plateau. Unfortunately, despite investigation using the geochemical properties of bulk samples there are multiple competing hypotheses over 1) the precise loess source regions, 2) the environmental controls on dust production, 3) whether sources shift through time, 4) the atmospheric mechanisms are for dust transport and indeed 5) whether the oldest part of the record (8-22 Ma) is indeed wind-blown at all. These disagreements severely limit our understanding of the very origin of these deposits, preventing us from constraining the past atmospheric, tectonic and oceanographic conditions responsible for the emission and transport of dust and undermining the use of certain climate proxies in loess. In turn this restricts our understanding of the dominant environmental processes operating in China in the past, and the origin of the current atmospheric circulation systems. Key to overcoming this gap is to properly constrain the sources of loess. Our pilot work has demonstrated that widely applied bulk sediment geochemical analysis of sediment from loess and source regions will mask the detail of the multiple sources of loess dust. Only individual grains of certain heavy minerals (> 2.8 specific gravity) can be source diagnostic as each grain will have one source and certain of its geochemical and geochronological characteristics may be diagnostic of this. Thus analysis of multiple single grains of zircons and other heavy minerals isolated from loess and adjacent desert deposits will be undertaken using a multi-proxy single-grain geochemical approach to maximise the likelihood of success. Until our pilot work this has not been conducted on Chinese loess previously. Samples for these analyses will be taken from multiple, typical loess sequences in China, allowing determination of source variance through time and space. Sampling will concentrate on key intervals, such as the uplift of Tibet, the onset of Ice Age glaciation and the enhanced intensity of glaciation in more recent times. The results will allow us to test between conflicting hypotheses of loess dust source and transport, enabling constraint of the fundamental controls on dust emission and the atmospheric mechanisms involved in their transport. In turn and through comparison with independent records, this will allow us to assess the effect of global and regional climatic, tectonic and oceanographic changes on dust. This is currently a poorly understood yet critical component of the Earth-system. Finally, the results will test the validity of widely used sedimentary proxies such as mass accumulation rates and grain size in reconstructing past environmental changes from loess.