基于概率神经网络的沙尘天气预报

为了对敦煌地区沙尘天气过程进行预报,构造概率神经网络模型。尝试组合多种预报因子,选取最优预报因子。通过叠套方法进行预报消空,减小冗余的晴好天气样本。使用Matlab神经网络工具箱提供的函数建立沙尘天气预报模型,并采用遗传算法获得模式层最佳平滑参数,来优化概率神经网络模型结构,使概率神经网络的函数逼近能力和网络性能得到提升。以此进行的沙尘天气预报效果较好,准确率达57%以上。     

A Comparison of the Physical and Optical Properties of Anthropogenic Air Pollutants and Mineral Dust over Northwest China

Emissions of mineral dust and its mixing with anthropogenic air pollutants affect both regional and global climates. Our fieldwork in late spring 2007(April 25–June 15) measured the physical and optical properties of dust storms mixed with local air pollutants at a rural site about 48 km southeast of central Lanzhou. Levels of air pollutants and aerosol optical properties were observed during the experiment, with concentrations of NOx(6.8 ± 3.3 ppb, average ± standard deviation), CO(694 ± 486 ppb), SO2(6.2 ±10 ppb), O3(50.7 ± 13.1 ppb), and PM10(172 ± 180 μg m-3), and aerosol scattering coefficient(164 ±89 Mm-1; 1 Mm = 106m) and absorption coefficient(11.7 ± 6.6 Mm-1), all much lower than the values observed during air pollution episodes in urban areas. During a major dust storm, the mass concentration of PM10 reached 4072 μg m-3, approximately 21-fold higher than in non-dust storm periods. The mixing ratios of trace gases declined noticeably after a cold front passed through. The observed CO/SO2 and CO/NOx ratios during air pollution episodes were 4.2–18.3 and 13.7–80.5, respectively, compared with the corresponding ratios of 38.1–255.7 and 18.0–245.9 during non-pollution periods. Our investigations suggest that dust storms have a significant influence on air quality in areas far from their source, and this large-scale transport of dust and air pollutants produces major uncertainties in the quantification of the global effects of emissions over Northwest China.     

Influence of Surface-Based Stable Layer Development on Asian Dust Behaviour Over Tokyo

The relationship between local meteorological fields and the behaviour of airborne Asian dust that arrived in the Tokyo metropolitan area on 1 April 2007 with the passage of a synoptic-scale cold front has been investigated through Doppler lidar observations, experiments using a regional atmospheric numerical model, and analyses of surface and upper-air meteorological observations. Results of the Doppler lidar observations showed that the Asian dust passed above the metropolitan area with strong south-westerly winds with speeds of 15–26 m s⁻¹. Meteorological fields reproduced by the numerical experiments showed the development of a surface-based stable layer in the metropolitan area caused by nocturnal radiational cooling near the ground surface and south-westerly warm air advection at upper levels. The blocking effect of the mountainous region located to the west of the metropolitan area induced an area of stagnant air inside the metropolitan area and promoted the stable layer development. Although strong downdrafts prevailed in the upper air, the airborne Asian dust did not spread to the ground when the stable layer was formed. These results strongly indicate that the developed stable layer prevented strong downdrafts from spreading to the ground, acting as an obstacle to the transport of the Asian dust particles from the upper air towards the ground. This is considered to be one of the main causes of the low appearance frequency of Asian dust phenomena near the ground in the Tokyo metropolitan area and eastern Japan.     

沙尘天气过程对中国北方城市空气质量的影响

2014年4月23-25日中国北方地区经历了一次大范围的强沙尘天气过程.利用13个城市的空气质量和气象数据结合后向轨迹模式和MODIS deep-blue气溶胶光学厚度数据,研究沙尘天气过程对中国北方城市空气质量的影响,特别是对PM_2.5质量浓度的贡献。结果表明:来源于塔克拉玛干沙漠及其周边地区的沙尘途经甘肃西北部和内蒙古中西部的沙漠区进而影响下游城市。MODIS deep-blue气溶胶光学厚度的空间分布能够较好地反映出沙尘的水平运动特征。沙尘天气过程可以使中国北方城市PM_2.5和PM₁₀质量浓度分别增加7.5%～1 141.2%和344.0%～2 562.1%,可以有效地降低SO₂、NO₂和CO气体污染物的浓度,而对O₃的浓度影响较小。银川、大同和库尔勒沙尘天气过程期间平均SO₂、NO₂和CO浓度较非沙尘天气过程期间降低最明显,分别降低约80.0%、78.7%和62.1%。     

宁夏盐池沙尘暴特征分析

对盐池气象站1961-2000年逐月气象数据进行了较为详细地分析，结果表明，干旱多风的气候特征在时间搭配上的一致性，是沙尘暴发生的主要直接原因；沙尘暴频率与降水量和大风日数显著相关，降水量对沙尘暴的影响滞后1-3个月。从较长时间序列看，影响沙尘暴发生的自然因子（以降水和大风为代表）的作用在逐渐下降，而人为因素的权重在不断加强，反映出生态环境的不断恶化；沙尘暴逐年序列的2-8a周期，是受到厄尔尼诺影响的结果。     

祁连山西段冰川积雪中大气粉尘沉积特征

基于2012 年夏季野外考察、微粒粒度测试和扫描电镜(SEM-EDX) 微观形貌观测研究, 对位于我国青藏高原东北缘的祁连山西段典型极大陆型冰川区老虎沟12 号冰川、野牛沟十一冰川积雪中大气粉尘沉积进行了分析研究。两冰川区积雪中微粒的平均质量浓度分别是3461 μg/kg、2876 μg/kg, 年均沉积通量分别是207.6 μg/cm²、143.8 μg/cm²。将本研究区与其他区域冰川积雪中粉尘浓度对比研究表明, 冰川受周边粉尘源区影响较大。雪坑微粒浓度剖面和离子相关性分析表明, 祁连山西段冰川积雪中污化层富含亚洲粉尘的富Ca²⁺、Na⁺矿物;微粒体积-粒径分布众数介于3~22 μm, 两冰川区的粒径众数分别为12.6 μm和12 μm, 粒径分布均显示了单结构模式, 同时反映了祁连山冰川区与毗邻的天山地区雪冰中粉尘粒径分布模式的相似性和粒径众数的差异性。通过SEM-EDX对粉尘颗粒的微观结构研究发现, 颗粒绝大多数为形貌不规则的矿物粉尘颗粒, 和很少数量的飞灰颗粒等。同时, 对粉尘来源结合Sr-Nd同位素测定和气团后向轨迹分析进行验证, 认为位于研究区北边的巴丹吉林沙漠是祁连山12号冰川区粉尘最可能的源区。     

沙尘天气影响因子的对比分析

对比分析了2002年和2003年春季沙尘天气发生的大气环流及地表条件特征, 并利用具有风沙物理过程的沙尘数值模拟结果和动力诊断技术, 探讨了2002年和2003年典型沙尘天气过程和冷空气过程中大气动力条件的作用。结果表明, 沙尘天气的强度、发生范围主要是由大气的动力条件所决定的, 而沙源区地表特征的变化对沙尘天气的发生起着重要的作用, 沙尘天气的强度和影响范围主要决定于沙尘的垂直输送过程。此外, 还总结出了沙尘天气中垂直输送-水平传输-沉降过程的概念模型。     

尘卷风的研究进展

尘卷风是一种发生在大气对流边界层内,能将沙尘或者碎屑等物体扬到高空、具有温度较高的低压核心和较短生命周期的旋风,是自然界中一种最常见的自然现象,也是气象学中最独特的内容。回顾尘卷风的研究历史,主要包括尘卷风的气象要素与环境效应,尘卷风研究的三种方法及取得的成果,粉尘颗粒在尘卷风中的悬移运动及其分层机制,粉尘活动与颗粒荷电,提出了进一步研究的主要内容。     

2007年春季沙尘暴对辽宁中部城市群空气质量的影响

根据大气降尘量和可吸入颗粒物PM10 、PM2.5、 PM1的观测资料,分析了2007年春季发生于我国北方的两次沙尘暴天气过程对辽宁中部城市群空气质量的影响。结果显示,沙尘天气使各城市大气降尘量与可吸入颗粒物PM10和PM2.5的质量浓度明显增高,空气质量明显下降。在3月31日和5月7日的两次沙尘暴过程中,城市群的大气降尘量比当月日平均值分别增加了1.5倍和2.5倍,各城市PM10的日均质量浓度比沙尘出现前一日分别增加了0.2～2.6倍和1.5～3.8倍。3月31日除铁岭和鞍山空气质量状况为良外,均达到轻微污染程度;5月7日鞍山达严重污染程度;沈阳、辽阳、本溪、铁岭和抚顺的空气质量均为中度污染或中度重污染。沙尘暴期间1.0～10.0 μm的粗粒子为影响辽宁中部城市空气质量的主要成分,其小时平均质量浓度最大时是沙尘出现前的10～30倍。     

Factors affecting the nature and rate of dust production from natural dune sands

This paper evaluates the influence of natural sand particle characteristics on the amount and particle-size distributions of dust produced by aeolian abrasion. It contrasts with previous studies of aeolian abrasion by conducting experiments using: (i) whole sand samples, as opposed to selected size fractions; (ii) natural, mature dune sands, rather than artificial or freshly crushed material; and (iii) weathered sands that have acquired a superficial clay coating, instead of grains with clean surfaces. Whilst previous research has found clear, positive relationships between particle size, sorting, roundness and the amount of dust produced by aeolian abrasion, the relationships determined in this study show some variation according to the geomorphological context from which the original samples were obtained. The most important factor affecting the amount and particle-size characteristics of the dust produced was the presence of a clay coating on the grain surface that is removed by the abrasion process. The dust produced by this mechanism had a modal size of 2–5  μ m and material <10  μ m comprised up to 90% of the particles produced.