【作者】 钱鹏；

【导师】 郑祥民；

【作者基本信息】 华东师范大学 ， 第四纪地质学， 2012， 博士

【摘要】 长江三角洲位于我国东部沿海,属亚热带海洋性季风气候,是当今海陆交互作用最为敏感和复杂地带。在大气环流作用下,古冬季风、西风携带的粉尘在黄土高原地区、中国东部平原(包括长江三角洲地区)和海域陆续沉积下来,形成了风尘堆积体,记录下了地质历史时期的古气候、古环境演变过程。而在现代沙尘暴期间,冬季风在蒙古、西伯利亚高压作用下,从西北、北方沙漠地区向中国东部及东部海域输送粉尘物质,使中国东部大气颗粒物中的化学和矿物成分等理化特性改变,记录下了粉尘源区及季风强度的信息。分析现代大气颗粒物的理化特征,是论证风尘黄土物质来源的一个新视角,并可为解析长江三角洲乃至全球古气候变化信息提供重要的线索和参考资料。本文以长江三角洲、沙尘输沙沿途地区的大气颗粒物和风尘黄土为研究对象,分别从颗粒物浓度、地球化学、色度、磁学、矿物组成等几方面研究长江三角洲、沙尘输沙沿途地区沙尘暴、非沙尘暴期间大气颗粒物与黄土堆积物理化特征,分析沙尘暴与非沙尘暴期间大气颗粒物与黄土理化特征的异同点,探索长江三角洲大气颗粒物中具有远源指示作用的指标和方法。对大量实测数据的综合分析,本文主要得到以下几点结论：(1)长江三角洲各采样点大气颗粒物质量浓度较低,呈冬高夏低的趋势,受沙尘暴事件、气象要素和污染控制投入的影响。上海市近地表大气颗粒物中,所测三个采样点的大气颗粒物(TSP、PM10、PM2.5)质量浓度顺序分别是：普陀>青浦>闵行、青浦>普陀>闵行、青浦>普陀>闵行；大气颗粒物质量浓度四季变化规律是秋、冬季最高,春季次之,夏季最低。空气质量评价发现,2009-2010年上海市大气颗粒物环境质量总体无污染,只有2009年12月下旬普陀、青浦、闵行三个采样点颗粒物污染程度为轻微污染。大气颗粒物中PM10占总悬浮颗粒物65-73%,PM2.5占40-50%,其中细颗粒污染物浓度较高,对城市居民的身体健康存在危害。质量浓度受气象要素、沙尘暴事件及污染排控等因素的影响。气温较低、湿度较小时,颗粒物浓度较大,反之则较小；沙尘暴使上海市大气颗粒物质量浓度明显增加,且在沙尘暴期间颗粒物中含有较多的粗颗粒物质。(2)上海市闵行、普陀、青浦区三个区同一时间的大气颗粒物元素体积浓度大小排列顺序及各元素随时间变化曲线均基本一致,表明三个区的大气颗粒物的化学成分非常相似,表明物源相似规律。其中,元素Zn、Cu、Mn、Ni、Cr、Ti、 Fe、Na富集在粗颗粒中；Ca, Mg, Si, Al富集在细颗粒中。Pb、Sr在粗细颗粒中富集程度相当。(3)富Na、Mg,贫Si, Al的特征以及Mg/Al比值可作为北方远源物质示踪的特征指标。上海市三个采样点大气颗粒物元素富集因子分布特征一致,研究显示Ni、Zn、Pb属于人为源污染,Sr、Cu、Mn、Cr、Ti、Ca、Al、Mg、Fe、Na等元素为地壳源物质。沙尘输沙沿途样品Pb、Zn、Ni元素来自人为源排放；Sr、Cu、Cr、Ti、Ca、Na元素的EF值较上海市大气颗粒物各元素EF值均有增加,属部分人为源物质,部分壳源物质；Mg、Al、Mn、Fe、K元素为壳源物质。沙尘暴、非沙尘暴和黄土的主量元素UCC标准化分布图显示,沙尘暴样品主量元素含量均比非沙尘暴样品更接近于黄土的主量元素含量,表明沙尘暴样品主要物质来源与黄土相似,可能与黄土物质具有同源性,即主要来源于西部沙漠,部分物质为局地源。因此,富Na、Mg而贫Si、Al的特征以及Mg/Al比值可作为北方远源物质示踪的特征指标。(4)稀土元素地球化学特征显示沙尘暴样品物质来源与黄土接近以壳源物质为主,非沙尘暴样品与沙尘暴样品、黄土物质来源不同,有非壳源物质的混合。稀土元素配分模式中,不同地点的沙尘暴颗粒物配分模式相似,且与北方黄土相似；不同地点的非沙尘暴颗粒物稀土元素配分模式间差异较大,与沙尘暴及黄土的配分模式差异也较大。稀土元素(La/Yb)CH vs.LaCH散点图显示,大部分黄土样品与沙尘暴样品落于同一区域。稀土元素三角图显示,沙尘暴样品与黄土Ce-La-Sm组成接近于UCC,说明沙尘暴样品物质来源与黄土接近,以壳源物质为主,但非沙尘暴样品与沙尘暴样品、黄土、UCC的Ce-La-Sm组成差别较大,说明非沙尘暴样品与沙尘暴样品、黄土物质来源不同,有非壳源物质的混合。因此推断,对比稀土元素配分模式及稀土元素(La/Yb)CH vs. LaCH散点图也可作为沙尘暴颗粒物物源示踪方法。(5)沙尘暴大气颗粒物与黄土的黄度特征非常接近,黄度可作为指示沙尘暴颗粒物源的有效指标。2009年10月至2010年10月间,上海市大气颗粒物明度、红度值总体呈增加趋势,黄度值仅在沙尘暴期间明显增加,其余则相对较小。北方城市春季大气颗粒物色度变化规律与上海市大气颗粒物黄度变化特征一致,说明在各采样点沙尘暴期间大气颗粒物的物质来源相似。沙尘暴大气颗粒物的色度特征介于非沙尘暴大气颗粒物和黄土之间,指示沙尘暴期间大气颗粒物与黄土具有相同的物源联系,部分物质来源于人为源污染。沙尘暴大气颗粒物与黄土的黄度特征非常接近,黄度可作为指示沙尘暴颗粒物源的有效指标。(6)环境磁学特征显示沙尘暴沙尘与黄土具有相似的物源联系,沙尘暴沙尘在输沙沿途吸附污染物,并带入长江三角洲地区。上海市沙尘暴期间TSP、PM10的χ1f及SIRM值升高,说明沙尘暴带来了污染物含量较高的粉尘,PM25的χ1f、SIRM值降低,说明污染元素富集于TSP、PM10颗粒上。北方城市沙尘暴大气颗粒物中污染物含量较小,沙尘暴粉尘比较清洁,但经过长距离输送,中途吸附了人为活动排放的污染产物。χfd值在冬季相对较高,超顺磁颗粒的贡献较大,表明质量浓度较高的非沙尘暴期间,颗粒物中所含颗粒粒径较细。χARM值与颗粒物质量浓度、元素质量浓度均在冬季升高,说明冬季颗粒物以单筹铁磁性矿物为主导,污染物质较易富集于细颗粒的单畴颗粒上。χIf-SIRM散点图显示,黄土样品点在沙尘暴样品点的延长线上,且黄土磁化率值较沙尘暴低,黄土、沙尘暴样品以亚铁磁性矿物及不完整反铁磁性矿物为主,且物源相近；非沙尘暴期间颗粒物中除了亚铁磁性矿物及不完整反铁磁性矿物外,还有其他磁性矿物存在。黄土、沙尘暴沙尘的χIf、SIRM与χARM均具有明显的相关性,揭示了黄土、沙尘暴沙尘物源的相似性。沙尘暴样品的χIf-SIRM明显高于黄土,说明在沙尘暴沙尘中不仅含有壳源物质,还有来自人为源的污染物。非沙尘暴样品与前二者不同,在xIf-χARM散点图与SIRM-χARM散点图中线性关系不明显,指示大气颗粒物除受非滞后剩磁贡献外,主要受其它颗粒影响,人为源污染物是大气颗粒物的主要物质组成。(7)上海市及北方输沙沿途城市大气颗粒物中主要矿物类型主要有：石英(Q)、石膏(G)、方解石(Cal)、白云石(Dol)、赤铁矿(Hem)、钠长石(Alb)、斜绿泥石(Clc)、蒙脱石(S)、高岭石(K)、伊利石(Ⅰ)、石盐(Hl)、菱铁矿(Sd)、黑云母(Bio)、白云母(Ms)等,沙尘暴期间大气颗粒物中矿物组成与黄土中矿物组成相似,代表二者具有物源联系。大气颗粒物中矿物组成在一年中变化明显,即秋、冬季和春季的沙尘暴时期矿物种类较多,春、夏季节的非沙尘暴期间矿物种类较少,这一四季变化规律与颗粒物浓度、颗粒物中元素浓度四季变化特征基本一致。沙尘暴时期大气颗粒物中石英、白云母、黑云母、蒙脱石、斜绿泥石的平均百分含量均大于非沙尘暴期间样品的矿物平均百分含量,石膏、方解石、白云石、高岭石、菱铁矿、伊利石平均百分含量均小于非沙尘暴期间样品中该矿物的平均百分含量。沙尘暴期间大气颗粒矿物组成较接近于黄土,说明二者物源相似。扫描电子显微镜观察发现,上海市大气颗粒物中不规则颗粒主要为矿物集合体,规则气溶胶小颗粒为球状燃煤颗粒、细小椭球形颗粒组成的汽车尾气絮状集合体等。能谱分析显示,沙尘暴样品中所含元素有O、C、Si、Al、K、Ca、 Fe、Mg、S、Na、Cl等元素；非沙尘暴期间大气颗粒物中所含元素有O、C、Si、 Al、Mg、Ni、Na、Cl、Ca、S、Fe、K等,除少量人为源污染元素外,推测矿物成分与XRD研究结果相似。南通、郑州、北京、西安沙尘暴期间的总悬浮颗粒物能谱分析显示,各采样点样品中所含元素相似,推测主要矿物与上海市沙尘暴期间矿物组成相似,指示它们之间存在物源联系。(8)色度、地化特征、磁学参数的二元Logistic回归模型显示,黄度b的系数最高,高达0.96,表明黄度b是指示远源物质的优良指标,磁学参数χ1f次之,是指示远源物质的良好指标,元素Al和Na体积浓度则可以作为指示远源物质的参考指标。用二元Logistic回归方法建立沙尘暴、非沙尘暴期间上海市大气颗粒物色度及磁学参数之间的数学模型；对沙尘暴、非沙尘暴期间上海市大气颗粒物元素的体积浓度运用灰熵关联度法,筛选出与因变量y关联更为密切的元素：Al、Si、Na、Mg作为自变量,并运用Logistic回归方法建立回归模型。从所建的三个回归模型来看,黄度b、磁学参数χ1f、元素Al和Na体积浓度的回归系数依次为0.96>0.003>3.3*104,其中,黄度b的系数最高,高达0.96,表明黄度b是指示远源物质的优良指标,磁学参数χ1f次之,是指示远源物质的良好指标,元素Al和Na体积浓度则可以作为指示远源物质的参考指标。在今后的研究中,应加强和改进黄土和大气颗粒物的采集和测试方法,获得足够多的大气颗粒物样品和更准确的实验数据。从黄土角度,详细分析同一时期的长江三角洲地区黄土与北方黄土高原黄土之间的区别与联系,结合大气颗粒物理化特征研究,更好地运用“将今论古”方法探讨远源示踪指标。更多还原

【Abstract】 The Changjiang river delta is located at the east coasts of China, where belong to Subtropical maritime monsoon climate, and it is the most sensitive and complex area of east China.Under the action of atmospheric circulation, particulate matters were trasported by paleo-winter monsoon and westlies to the Loess plateaus, East China plain respectively, including Changjiang river delta and Sea area, forming aeolian deposit, and recorded the process of paleo-environmental evolutions. Biside it, during the time of dust storm, winter monsoon transported dust materials from desert of Northwest and North area of China to East China and east areas by the action of Mongolia and Siberian high pressure, changing the Chemical and mineral compositons of aeolian dust of East China, and recorded the information of the dust sources and the intensity of monsoon. Comparative analysising of the physicochemical characteristics of atmospheric particle matters and aeolian loess is the new angle of view to study the material origin, and it can provide important reference data for the Changjiang river delta and even global paleo-climat changing research.In this paper, the atmospheric particle matter and aeolian loess of Changjiang river delta and the areas on the way of dust transporting were used as study objects, and studied the Physicochemical characteristics of atmospheric particle matter of dust storm and non-dust storm periods and aeolian loess in Changjiang river delta and the areas on the way of dust transporting on the aspects of concentration, cheminstry, color index, magnetic characters, mineral compositon, etc., and analyzed the similarities and differences of the physicochemical characteristics and the distan-origin indicating indexes and methods of atmospheric particle matters of dust storm and non-dust storm periods and aeolian loess. Through analysis by synthesis of abundance data, several main conclusions come into being: (1) The order of mass concentrations of different grain sizes-atmospheric particulate matters in Shanghai city is TSP> PM10>PM2.5.The order of mass concentrations of TSP, PM10and PM2.5from three different sampling points are Putuo>Qingpu>Minhang, Qingpu>Putuo> Minhang, Qingpu>Putuo>Minhang. The seasonal order of mass concentration of atmospheric particulate matters is that the autom and winter are higher than spring, the spring is higher than summer. The assessment of air quality shows that the environments of atmospheric particulate matters in shanghai city has no pollution from2009to2010, except for the slight pollution of three sampling points of Putuo, Qingpu and Minhang districts in Dec.2009. The percent of PM10is65-73%in TSP, and PM2.5is40-50%, therefore, the fine particle pollution is high and harmful to health of city residents. The main effect factors of mass concentration of air quality are climate, dust storm and orgnation’s pollution controll. Their relationships are that lower temperature and humidity is match along with higher concentration, on the contrary, higher temperature and humidity is match along with the lower concentration. The dust storm event will make the mass concentration to increase obviously, and there will be more coarse particles in dust storm.(2) The order of element volume concentrations and their seasonal changes of atmospheric particulate matters are probably identical in Putuo, Qingpu and Minhang districts of Shanghai, which show that the element compositions of atmospheric particulate matters in these three districts are exactly similar. This indicates the same sources of the atmospheric particulate matters. The elements of Zn, Cu, Mn, Ni, Cr, Ti, Fe, Na are easy to be absorbed in coarse particles, and Ca, Mg, Si, Al in fine particles, Pb, Sr in both grain sizes particles.(3) The distributions of enrichment factors (EF) values of elements in atmospheric particulate matters from the three sampling points in Shanghai are identical, which shows that elements of Ni, Zn, Pb belong to anthropogenic pollutions, Sr, Cu, Mn, Cr, Ti, Ca, Al, Mg, Fe, Na belong to crust original materials. In the samples from the sampling points along aeolian dust transporting ways, the elements of Pb, Zn, Ni come from anthropogenic source pollutions, and the EF values of Sr, Cu, Cr, Ti, Ca, Na in them are higher than the one in atmospheric particulate matters of Shanghai, which indicates that part of these elements come from anthropogenic materials, part of them from crustal materials. The elements of Mg, Al, Mn, Fe, K are from crustal materials.The standarded major elements by UCC in dust storm, non-dust storm and loess show that, the advantage concentrations of major elements in dust storm dusts are more closed to them in loess than in non-dust storm dusts, which indicates that the material origin of the dust storm dust is indentical to the source of loess, which mainly come from the deserts in western China,only part of them come from located places. The chemical characteristic of rich Na, Mg but poor Si, Al and the Mg/Al ratio can be used as indexs to indicat far source materials.(4) The rare element patterns of dust storm dust from all sampling points are similar to loess’s of northern China, however, the rare element patterns of non-dust storm dust from different places are obviously different with each other and are different with dust storm dust and loess. Most of the loess and dust storm samples fall in the same field of scatter diagram of (La/Yb)CH vs. LaCH. The Ce-La-Sm compositions of loess and dust storm in triangular plot are close to UCC that indicates the origin of dust storm dust is identical to loess, which is mainly crust original material. However, the Ce-La-Sm compositions of non-dust storm dusts are different with the dust storm dust, loess, UCC, which indicates the material origin of non-dust storm dust is different with the dust storm dust and loess, and the dusts mixed with non-crust original material. Therefore, the dust storm dust has the same origin with the northern loess. (5) From Oct.2009to Oct.2010, the seasonal variations of color indexes of atmospheric particulate matters from three sampling points are consilient with each other, the rightness and redness of atmospheric particulate matters increased with time lapes, yellowness increased obviously during the time of dust storm, and decreased during the time of non-dust storm. The color indexes of atmospheric particulate matters in spring from the Northern cities changes with time lapes, and its variation of yellowness has good correlation with atmospheric particulate matters from Shanghai, which indicate that the source region of atmospheric particulate matters in spring from different places are similar to each others.The characteristics of color indexes of atmospheric particulate matters during the time of dust storm fall in between the characteristics of non-dust storm and loess, which indicates that part of the materials in atmospheric particulate matters during the time of dust storm have the same source relation with the loess, part of the materials comes from artifical sources. Therefore, color indexes especially yellowness are the effective indexes of source region indicating for atmospheric particulate matters during the time of dust storm.(6) High χlf and SIRM values of TSP, PM10in period of dust storm infor that high concentrations of pollution were carried from Northern by dust storm dust. Low χlf and SIRM values of PM2.5indicates that the contamination elements are enriched in TSP and PMio-The pollution concentration of dust storm particulate matters from Northern cities is low, so the dust storm dust is clean. But during the long way of transportation, many anthropogenic pollutions were absorbed on the dust, χlf value is relatively higher in winter than other seasons, and the superparamagnetic particles do contribution to the values, which indicate that high mass concentration of atmospheric particulate matters in non-dust storm period are mianly fine grain particles.χARM and particle’s mass concentration, elements’mass concentrations all maintain at high value in winter, which indicates the SD ferromagnetizm dominants magnetic minerals in atmospheric particulate matters, and the pollution material is easy to be absorbed in SD particles. The dust storm samples fall in the extended line of loess in χ1f-SIRM scatter diagram, and it’s magnetic susceptibility is higher than loess’s, which suggests that the loess and dust storm samples are dominated by ferrimagnetism minerals and incompleted anti-ferromagnetism minerals, the non-dust storm dust contents not only ferrimagnetism minerals and incompleted anti-ferromagnetism minerals, but also other magnetic minerals. The obvious correlation of χ1f、SIRM,and χARM between loess and dust storm dust reveals the similar sources of loess and dust storm dust. The χ1f and SIRM value of dust storm dust are higher than loess, which indicates that there are not only crust original materials,but also anthropogenic pollutions in dust storm dusts. Non-dust storm dusts are different with loess and dust storm dusts, and there is no clear correlation in χ1fχARM and SIRM-χARM scatter diagrams, which suggests that the atmospheric particulate matters are effected by other particles except for ARM’S contribution, and the main compositions of the atmospheric particulate matters are anthropogenic pollutions.(7) The main minerals in atmospheric particulate matters of shanghai and northern cities are Q,G,Cal, Dol, Hem, Alb, Clc, S, K, I, H1, Sd, Bio, Ms, etc. Mineral composition in atmospheric particulate matters changes in different periods. More kinds of minerals are found in autumn, winter samples and dust storm dusts, and less in non-dust storm of spring and summer samples.This seasonal changing discipline is similar with the particle concentration and elements’concentrations.The average percentage concentrations of Q, Ms, Bio, S, Clc in atmospheric particulate matters of dust storm period are greater than that of non-dust storm period, G, Cal, Dol, K, Sd, I in in atmospheric particulate matters of dust storm period are less than that of non-dust storm period. The mineral compositions of atmospheric particulate matters in dust storm period are closed to loess, which suggests the resemble material sources of them.Watching by scanning electron microscope (SEM), we find that the irregular atmospheric particulate matters from Shanghai are mainly mineral combinations, the regular particles in them are globular fire coal,the cotton shaped combinations are wee ellipsoidal vehicle exhuast. Researched by EDS, the dust storm samples contain elements of O, C, Si, Al, K, Ca, Fe, Mg, S, Na, Cl,etc., the non-dust storm samples contain elements of O,C,Si,Al, Mg, Ni, Na, Cl, Ca, S, Fe, K, ect. Excepting for a little anthropogenic pollutions, the result of mineral compostions of samples studied by EDS are similar with the result studied by XRD. The study results by EDS of TSP sampled from Nantong, Zhengzhou, Beijing, Xi’an in dust storm period show that the element compositions of samples from every sampling point are similar with each other. Therefore, the mineral compostions of these sampling points are limilar with the samples from Shanghai in dust storm period, which infor the source relation of these tow kinds of samples.(8) Using the method of Binary Logistic Regression to build chroma and magnetic mathematical models of atmospheric particulate matters of shanghai during dust storm and non-dust storm period. Using the method of Ash grey relation entropy to screen out the elements whose volume concentrations of atmospheric particulate matters during dust storm and non-dust storm period have closer relations with dependent variable y. These elements are Al, Si, Na, Mg, which will be used as independent variable, and to build Logistic Regression model. Seen from these three regressions, the regression coefficient of Yellowness (0.96) is the highest one among them, and the regression coefficient of the magnetic parameter χlf (0.003) take the second place, the last one is the regression coefficient of Al, Si (3.3*10-4). Therefore, if to indicate far source materials, Yellowness b is the excellent index,χlf is the good index, and volume concentrations of Al, Si are the reference indexs for us.These conclusions are mainly identical with the above paragraphs. In the future research, we should enhance sampling frequency, improve sampling methods, and get enough samples of atmospheric particulate matters and more accurate sampling data. Combining with the physicochemical characteristics of the atmospheric particulate matters, analyze the differences and relations between loesses from Changjiang river delta and the Northern China of the same period, and discuss the tracing index of far source materials using the method of "present being a key to past".更多还原