Bagging-based machine learning algorithms for landslide susceptibility modeling

Natural Hazards - Landslide hazards have attracted increasing public attention over the past decades due to a series of catastrophic consequences of landslide occurrence. Thus, the mitigation and...     

生物土壤结皮固沙理论与实践

生物土壤结皮是由土壤微生物、藻类、地衣和苔藓等孢子植物类群与土壤颗粒形成的有机复合体,在全球干旱区地表广泛分布,是干旱地表生物覆被层的主要构建者.生物土壤结皮是荒漠植物群落演替的先锋类群,能够提高荒漠地表的稳定性,固定碳和氮等营养元素,增加土壤肥力,并在保持土壤水分方面发挥重要作用,因此在干旱区受损地表的生态修复方面具...     

海南东北部地区地壳地震各向异性特征初步研究

基于海南省地震台网2000~2013年的区域地震波形数据,用剪切波分裂系统分析方法(SAM)获得了海南琼东北部地区"九五"数字台网中2个台站的剪切波分裂参数。结果表明,快剪切波偏振优势方向代表了原地最大主压应力方向。七星岭台NE方向的快剪切波偏振优势方向与区域水平主压应力场方向不一致,与NE走向的断裂一致,体现了局部构造和局部应力场的复杂性;青山岭NNE向的快剪切波偏振优势方向揭示了NNE走向断裂的构造意义。同时,本研究证实,位于活动断裂上或几条活动断裂交汇部位的台站的快剪切波偏振优势方向与对所选用的小地震起控制作用的活动断裂走向一致,而快剪切波偏振优势方向较为离散则反映了该区域复杂的断裂构造和应力分布特征。     

北祁连山西段早元古代变质火山岩的地球化学特征及其构造背景

产于北大河群的早元古代火山岩是北祁连山西段最主要的三期火山岩之一,其地球化学特征表明,该期火山岩的原岩为具有双峰式组合的拉斑玄武岩和中酸笥凝灰岩。两者具有完全不同的稀土分布型式,前者以极低的富集轻稀土为特征,而后者以强烈富集轻稀土的特征;基性火山岩的微量元素分布曲线指示了其形成了裂陷槽环境。结合Ｎｄ同位素特征以及区域２得出本区早元古代变质火山岩形成于裂陷槽环境。     

南大洋氟的生物地球化学研究——Ⅰ.南极磷虾富氟异常的研究

本文通过对南极磷虾不同部位中氟含量的测定,结合氟在测区环境中的分布特征及南极磷虾体中其它元素的分析资料,讨论了氟在磷虾体内的分布和赋存形式,并对其可能的富集机理及与环境的关系作了初步的探讨。     

Characteristics and cause analysis of heavy haze in Changchun City in Northeast China

Northeast China has been reported as having serious air pollution in China with increasing occurrences of severe haze episodes. Changchun City, as the center of Northeast China, has longstanding industry and is an important agricultural base. Additionally, Changchun City has a long winter requiring heating of buildings emitting pollution into the air. These factors contribute to the complexity of haze pollution in this area. In order to analyze the causes of heavy haze, surface air quality has been monitored from 2013 to 2015. By using satellite and meteorological data, atmospheric pollution status, spatio-temporal variations and formation have been analyzed. Results indicated that the air quality in 88.9% of days exceeding air quality index(AQI) level-1 standard(AQI 50) according to the National Ambient Air Quality Standard(NAAQS) of China. Conversely, 33.7% of the days showed a higher level with AQI 100. Extreme haze events(AQI 300) occurred frequently during agricultural harvesting period(from October 10 to November 10), intensive winter heating period(from Late-December to February) and period of spring windblown dust(April and May). Most daily concentrations of gaseous pollutants, i.e., NO_2(43.8 μg/m~3), CO(0.9 mg/m~3), SO_2(37.9 μg/m~3), and O_3(74.9 μg/m~3) were evaluated within level-1 concentration limits of NAAQS standards. However, particulate matter(PM_(2.5) and PM10) concentrations(67.3 μg/m~3 and 115.2 μg/m~3, respectively) were significantly higher than their level-1 limits. Severe haze in spring was caused by offsite transported dust and windblown surface soil. Heavy haze periods during fall and winter were mainly formed by intensive emissions of atmospheric pollutants and steady weather conditions(i.e., low wind speed and inversion layer). The overlay emissions of widespread straw burning and coal combustion for heating were the dominant factors contributing to haze in autumn, while intensive coal burning during the coldest time was the primary component of total emissions. In addition, general emissions including automobile exhaust, road and construction dust, residential and industrial activities, have significantly increased in recent years, making heavy haze a more frequent occurrence. Therefore, both improved technological strategies and optimized pollution management on a regional scale are necessary to minimize emissions in specified seasons in Changchun City, as well as comprehensive control measures in Northeast China.     

卫星面降水估计人工神经网络方法

介绍了人工神经网络原理和卫星云图估计降水的原理.从GMS红外卫星云图资料中抽取12个降水云图特征量,构造了网络结构为12-98-7的降水估计人工神经网络模型,并用1993年的小时地面雨量资料和GMS数字云图资料对神经网络模型进行训练,用1992和1994年资料对该神经网络模型分别进行测试.在日面降水估计试验中,地面雨量计值和卫星估计降水之间的相关系数分别为0.94和0.97,相对误差分别为41%和32%.     

Spatiotemporal characteristics of precipitation and extreme events on the Loess Plateau of China between 1957 and 2009

Spatiotemporal trends in precipitation may influence vegetation restoration, and extreme precipitation events profoundly affect soil erosion processes on the Loess Plateau. Daily data collected at 89 meteorological stations in the area between 1957 and 2009 were used to analyze the spatiotemporal trends of precipitation on the Loess Plateau and the return periods of different types of precipitation events classified in the study. Nonparametric methods were employed for temporal analysis, and the Kriging interpolation method was employed for spatial analysis. The results indicate a small decrease in precipitation over the Loess Plateau in last 53 years (although a Mann–Kendall test did not show this decrease to be significant), a southward shift in precipitation isohyets, a slightly delayed rainy season, and prolonged return periods, especially for rainstorm and heavy rainstorm events. Regional responses to global climate change have varied greatly. A slightly increasing trend in precipitation in annual and sub‐annual series, with no obvious shift of isohyets, and an evident decreasing trend in extreme precipitation events were detected in the northwest. In the southeast, correspondingly, a more seriously decreasing trend occurred, with clear shifts of isohyets and a slightly decreasing trend in extreme precipitation events. The result suggests that a negative trend in annual precipitation may have led to decreased soil erosion but an increase in sediment yield during several extreme events. These changes in the precipitation over the Loess Plateau should be noted, and countermeasures should be taken to reduce their adverse impacts on the sustainable development of the region. Copyright © 2013 John Wiley & Sons, Ltd.     

地下水测年方法及其在地震监测中的应用展望

对地下水动力学特征的观测与研究,是地震前兆监测的重要方法之一。地下水动力学研究中,利用地下水测年方法,可以对地下水的补给、径流特征及更新能力作出定性和定量的分析。本文介绍了近年来国际上常用的地下水测年方法,总结了前人在该领域的主要研究进展,重点分析了地下水更新能力与地震前兆信息的关系、地下水运动规律与地震构造活动的关系以及地下水浅层补给与异常干扰排除的方法等。已有的研究成果表明,地下水年龄的测定与分析方法对于了解监测点映震能力、评价构造活动与地震活动程度,以及在观测资料异常变化的现场核实等方面,可以发挥重要作用。