Earthquake-induced landslide prediction using back-propagation type artificial neural network: case study in northern Iran

Natural Hazards - Landslides can cause extensive damage, particularly those triggered by earthquakes. The current study used back propagation of an artificial neural network (ANN) to conduct risk...     

Historical evidence of climate change impact on drought outlook in river basins: analysis of annual maximum drought severities through daily SPI definitions

In the context of major outcomes of a steadily changing climate, extreme climatic conditions and the associated events in various forms of weather-related natural disasters, e.g. droughts, floods, and heat waves, are more frequently experienced on the global scale in recent years. In support of this argument, there are adequate numbers of explicit signals over such a persistent outlook, which is greatly illustrated by historical data and observations. This study, which is mainly oriented to investigating the drought behaviour in Thracian, Aegean and Mediterranean transects of Turkey's major river basins, is actually inspired by the foreseen potential of using annual maximum drought severity series (based on drought definition through the standardized precipitation index (SPI)) within a framework that resembles the use of flood discharge directly from flow measurements in a river basin. To this end, a series of spatial analyses were employed to identify different aspects of flood appearance in the study extent, including trend views on annual average drought severity series, shifts in the starting time of the annually most severe flood periods, and changes in spatial coverage views of average drought conditions under different drought severity categories. The framework of the analytical approaches depends greatly on validated international datasets and open-source computational algorithms. The results from the analyses that were conducted in two consecutive periods of 1958–1980 and 1981–2004 revealed that Turkey's western and southern river basin systems seemed to have experienced quite different behaviours between the two periods in terms of drought severity magnitudes, drought durations and annual occurrence times.

     

Flood vulnerability and buildings’ flood exposure assessment in a densely urbanised city: comparative analysis of three scenarios using a neural network approach

Natural Hazards - Advances in the availability of multi-sensor, remote sensing-derived datasets, and machine learning algorithms can now provide an unprecedented possibility to predict flood events...     

Prediction of Rock Abrasivity Index (RAI) and Uniaxial Compressive Strength (UCS) of Granite Building Stones Using Nondestructive Tests

Geotechnical and Geological Engineering - Rock abrasivity index (RAI) and uniaxial compressive strength (UCS) are two key parameters for assessing abrasivity and durability of building stones,...     

基于机器学习的表层土壤成矿元素空间预测: 以稀有金属铷元素为例

近些年随着土地质量地球化学调查工作的开展,获取了大量表层土壤样品数据。然而,这些数据也存在一个明显的缺陷,即1∶50 000大比例尺表层土壤数据往往缺少成矿元素。鉴于土壤成矿元素含量对于矿产资源勘查的重要指示作用,尝试基于现有数据对大比例尺表层土壤成矿元素含量提供一个补全方案。以稀有金属铷元素为例,采用随机森林算法把同一区域2 548组1∶250 000小比例尺表层土壤数据按照8∶2的比例随机分为两组,用80%的数据进行训练建模,20%的数据对模型进行验证。采用变量重要性度量排序和构建学习曲线的组合方法优选了8种元素(K、B、Ni、V、Zn、As、Co、Cu)作为预测变量,模型对训练数据和测试数据的拟合优度R²分别达到0.983 2和0.895 6,说明预测变量的优选方法是有效的。随后将1∶50 000表层土壤的上述预测变量数据作为输入变量导入模型中,得到预测的Rb元素含量,预测结果比较符合实际特征。本研究表明将大数据机器学习随机森林算法引入表层土壤地球化学元素含量空间定量预测具有可行性,可进一步拓展土地质量地球化学数据的服务应用维度。     

河南省农业专业村发展与地理环境关系研究

综合运用GIS技术、逐步回归法、Logit模型,分析了2009—2012年河南省农业专业村数量结构、空间分布的变化特征,并探究了地理环境对农业专业村发展的影响。研究结果表明:农业专业村空间分布不均衡,局部地区集聚程度发生变化;农业专业化率随劳动力资源的增加而降低,丘陵、山区农业专业化率比平原区低,村庄区位显著影响农业专业化率;伴随农区发展,农业专业化生产对劳动力资源的依赖程度有所降低,丘陵、山地地形对农业专业化率的负面影响减弱,市场因素对农业专业化率的影响凸显;劳动力资源对花卉、蔬菜等劳动密集型的专业村形成具有正面影响,邻近批发市场能提高蔬菜、水果、花卉等市场依赖性强的专业村形成概率,靠近地级市、县城的区位条件有利于形成蔬菜、水果等专业村。该结果对农区专业村发展方向和策略制定具有一定借鉴意义。