人工神经网络和决策树模型在滑坡易发性分析中的性能对比

机器学习模型广泛应用于区域性滑坡易发性分析。模型的选择关系到评价结果的可信度、准确率和稳定性。现有滑坡易发性分析模型对比研究侧重模型的预测精度。模型的稳定性和数据量敏感性对机器学习模型的性能评估同样非常重要。本文以福建省南平市蔡源流域为研究区,以四川省绵阳市北川县为验证区,从预测精度、稳定性和数据量敏感性3个方面深入对比BP（Back Propagation）人工神经网络模型和CART（Classification and Regression Tree）决策树模型在滑坡易发性分析中的效果,主要结论如下:① 在逐渐增加一定数量训练样本的过程中,BP人工神经网络模型预测精度的增长率更高。在蔡源流域内,当训练样本数量增加10 000时,BP人工神经网络模型的预测精度上升5.22%,CART决策树模型的预测精度上升2.11%。② BP人工神经网络的预测精度高于CART决策树模型,且较为稳定。在100组数据集上,BP人工神经网络模型验证集预测精度的均值和验证集滑坡样本预测精度的均值分别为81.60%和84.86%,高于CART决策树模型的72.97%和76.59%。与此同时,BP人工神经网络模型对应预测精度的标准差分别是0.32%和0.37%,小于CART决策树模型的0.35%和0.67%。③ BP人工神经网络模型分析的滑坡易发区相比CART决策树模型,更接近实际滑坡的空间分布。最后,北川县的验证实验也出现了相同的现象。     

Glacial forest refugium in Howard Valley,South Island,New Zealand

Fossils of forest habitat beetles and leaves of Nothofagus menziesii provide evidence of a forest refugium at times between ca. 34 000 and ca. 18 500 cal. a BP at an upland site in Howard Valley, located adjacent to glaciated valleys in South Island, New Zealand. The stratigraphy of the glacial‐aged terrace sequence of organic‐rich silts and fluvial sand/gravels indicates that soil development occurred episodically for around 15 000 a. Fifty‐four beetle taxa represent seven habitat types: forest, forest or scrub, riparian and aquatic, litter, grass/tussock, marshland and moss habitats. Leaf and beetle fossils indicate that forest dominated by N. menziesii persisted at the site for most of the time period represented, and tree line taxa such as Taenarthrus sp. 1 (Carabidae) and Podocarpus sp. (Podocarpaceae) indicate that the site may represent the upper tree limit for full‐glacial time. The finding of forest at this elevated site adds to the growing fossil evidence for multiple forest refugia in New Zealand during the last glaciation and is consistent with the pollen records, which have consistently indicated the presence of forest species during the last glaciations. Copyright © 2009 John Wiley & Sons, Ltd.     

Central England temperature since AD 1850: the potential of stable carbon isotopes in British oak trees to reconstruct past summer temperatures

In regions with seasonal temperate climatic regimes, tree growth is rarely controlled by any single environmental factor. As a consequence, the development of robust palaeoclimate reconstructions has proved challenging. Tree‐ring stable carbon isotope ratios (δ¹³C), however, are controlled primarily by photosynthetic rate, not by net growth. Therefore, at sites where climatic controls on tree‐ring growth are not strongly expressed, a robust (isotopic) palaeoclimate signal may still potentially be preserved. This hypothesis was tested using a 160‐year record of δ¹³C measured from the pooled latewood cellulose of six Quercus petraea L. (sessile oak) trees from Allt Lan‐las in West Wales, UK. Raw δ¹³C values were corrected for changes in the isotopic ratio of atmospheric carbon dioxide and for changes in the behaviour of trees due to the increasing availability of atmospheric CO₂ since AD 1850. Strong correlations with local summer temperature and sunshine are reported, and also with the Central England Temperature record over the full length of the isotopic chronology (AD 1850–2010) (r = 0.69, P < 0.001). We conclude that tree‐ring stable isotopes can be used to extract strong palaeoclimate signals even from oak trees growing in a temperate maritime climate. This demonstrates the potential for extracting robust palaeoclimatic information from the very long and well‐replicated oak chronologies which have been developed in western and central Europe primarily for dating rather than palaeoclimatic research purposes. Copyright © 2012 John Wiley & Sons, Ltd.     

基于GIS与SDM技术的可视化空间数据分类研究

提出将GIS与可视化空间数据挖掘技术之集成的基本框架。在此基础上,基于VisualC++6.0和Ma-pObject2.0组件技术设计和开发了一个可视化交互空间数据挖掘分类系统,系统采用决策树方法和贝叶斯网络作为数据挖掘方法的基本算法,采用训练与学习相结合实现空间数据的分类。文中用实例数据对系统性能、算法和规则有效性进行了验证。结果表明,该系统是一个适用的、可扩展的可视化交互空间数据挖掘工具,系统能够实现数据挖掘实时动态的交互控制,实现了数据挖掘过程的可视化、挖掘模型的可视化和结果的可视化显示、可视化思考、可视化分析与评价。     

应用描述词汇约简的OGC地理信息服务演绎推理

针对OGC地理信息服务在地理空间知识的有效组织和表达方面的能力比较弱,缺乏对服务信息的描述,数据丰富而知识缺乏,致使已有的数据在知识的表示和检索上存在缺陷,通过引入地理本体,按照基础地理信息要素分类,对地理信息服务的关键描述词汇进行提取,构建了地理信息服务本体库和实例数据库,应用粗糙集理论建立关键描述词汇约简模型,基于该模型实现了对地理信息服务实例数据库进行知识约简,形成最优实例数据库。最后应用地理信息服务本体库和实例数据库,基于演绎推理模型开发实例原型系统,实现了对地理信息服务对象的语义检索和推理,并通过试验从查全率和查准率两个指标验证了该方法的可行性、有效性和准确性。     

基于决策树的洪泽湖湿地信息提取

湿地是地球的重要组成部分,快速而准确地提取湿地信息,是湿地动态监测和可持续发展的一项基础而重要的工作.以洪泽湖淡水湿地为研究对象,采用2006年8月19日的Landsat5 TM遥感影像为数据源,采用经K-T变换光谱增强后的数据及利用灰度共生矩阵分析影像第一主成分的纹理统计量作辅助分类变量,基于地物光谱特征、纹理特征和形状特征,运用决策树分类法提取洪泽湖湿地植被信息,将其分类结果与最大似然法的分类结果进行比较.结果表明:1)洪泽湖地区的湿地植被比较丰富,面积约占全湖的10.74％,其中以浮水植物为主.2)基于决策树的分类结果的精度有了明显的提高,总精度由77.33％提高到86.33％,Kappa系数由0.7292提高到0.8354,证明基于决策树分类方法是提取淡水湿地植被信息的有效手段.     

TREE INVASION IN MEADOWS OF LASSEN VOLCANIC NATIONAL PARK,CALIFORNIA*

The causes of tree invasion into meadows in Lassen Volcanic National Park were investigated by dating invasive trees and associating age structure patterns with regional variations in fire history, grazing, and climate change. Massive invasion was associated primarily with cessation of livestock grazing and burning between 1905 and 1933, but climatic change may also have contributed. Maintenance of meadows may require continued use of unnatural processes due to uncertainties about pre-European fire regimes and other resource considerations.     

Palaeoecological evidence of changes in vegetation and climate during the Holocene in the pre‐Polar Urals,northeast European Russia

This study investigated Holocene tree‐line history and climatic change in the pre‐Polar Urals, northeast European Russia. A sediment core from Mezhgornoe Lake situated at the present‐day alpine tree‐line was studied for pollen, plant macrofossils, Cladocera and diatoms. A peat section from Vangyr Mire in the nearby mixed mountain taiga zone was analysed for pollen. The results suggest that the study area experienced a climatic optimum in the early Holocene and that summer temperatures were at least 2°C warmer than today. Tree birch immigrated to the Mezhgornoe Lake area at the onset of the Holocene. Mixed spruce forests followed at ca. 9500–9000 ¹⁴C yr BP. Climate was moist and the water level of Mezhgornoe Lake rose rapidly. The hypsithermal phase lasted until ca. 5500–4500 ¹⁴C yr BP, after which the mixed forest withdrew from the Mezhgornoe catchment as a result of the climate cooling. The gradual altitudinal downward shift of vegetation zones resulted in the present situation, with larch forming the tree‐line. Copyright © 2003 John Wiley & Sons, Ltd.     

On minimum spanning trees and the intergradation of clusters

Cluster analysis groups samples, but does not generally show gradations between clusters. To illustrate these relationships, principal-coordinate analysis and multidimensional scaling can be employed, but these methods may not be appropriate due to structural distortion. A minimum spanning tree (MST) computes a point-to-point path through the original matrix, and clusters are mapped on the MST diagram. The MST linkages, therefore, provide a natural continuum between the clusters, without distortion. Forty-six Silurian Tonoloway Limestone samples were analyzed for 16 sedimentary features indicative of subaerial exposure. Cluster analysis identified nine clusters which were reduced to five final, ordered groups by MST.     

A study on the bio-hydro mechanical effects of the tree roots on the Agaclı coal reclaimed – forested land

The research area that is in the north of Istanbul (Turkey) and it is a land of open pit coal mining residuals reclaimed and turned to forest in 1988–1989. The materials that are open pit mine spoils are formed by sandy loam, sandy clay loam, heavy loam and clay (noncalcareous Pliocene I sediments). Pseudoacacia (Robinia pseudoacacia), Maritima Pine (Pinus pinaster) and Stone Pine (Pinus pinea) are planted on these materials. The aims of this research were to determine the strength of the material provided by tree roots and evaluate the performance of the forestation practices at lands especially where open pit coal mining spoil (residual) materials piled up. The research has been performed at two stages. The first one was hydro-mechanical effect concerned about material moisture depletion as a result of transpiration and the second one was bio-mechanical effects due to strength parameters of the roots and their growth pattern within the material. All results were evaluated in relation to tensile strength and displacement curves due to pre and post failure behaviors of materials and root systems. The results showed the increase of the displacement of the material was the significant indicator of the bio-mechanics tensile strength of the material provided by the tree roots. The fast growing type of tree roots have stabilized the material up to a depth for 80–100 cm by covering the coal residual materials within 12 years.