Prevalence-adjusted optimisation of fuzzy models for species distribution

Most performance criteria which have been applied to train ecological models focus on the accuracy of the model predictions. However, these criteria depend on the prevalence of the training set and often do not take into account ecological issues such as the distinction between omission and commission errors. Moreover, a previous study indicated that model training based on different performance criteria results in different optimised models. Therefore, model developers should train models based on different performance criteria and select the most appropriate model depending on the modelling objective. This paper presents a new approach to train fuzzy models based on an adjustable performance criterion, called the adjusted average deviation (aAD). This criterion was applied to develop a species distribution model for spawning grayling in the Aare River near Thun, Switzerland. To analyse the strengths and weaknesses of this approach, it was compared to model training based on other performance criteria. The results suggest that model training based on accuracy-based performance criteria may produce unrealistic models at extreme prevalences of the training set, whereas the aAD allows for the identification of more accurate and more reliable models. Moreover, the adjustable parameter in this criterion enables modellers to situate the optimised models in the search space and thus provides an indication of the ecological model relevance. Consequently, it may support modellers and river managers in the decision making process by improving model reliability and insight into the modelling process. Due to the universality and the flexibility of the approach, it could be applied to any other ecosystem or species, and may therefore be valuable to ecological modelling and ecosystem management in general.     

Embedding the human factor in road tunnel risk analysis

The paper is focusing on road tunnel safety and builds upon the Directive 2004/54/EC launched by the European Commission; the latter sets basic requirements and suggests the implementation of risk assessment in several tunnel cases apart from technical measures imposed on the basis of tunnel structural and operational characteristics. Since the EU Directive does not indicate the method for performing risk assessment, a wide range of methods have been proposed, most of them based on quantitative risk assessment (QRA). Although the majority of current road tunnel QRAs assess physical aspects of the tunnel system and consider several hazards concerning the transportation of dangerous goods through a tunnel, they do not take into account, sufficiently, several organizational and human-related factors that can greatly affect the overall safety level of these critical infrastructures. To cope with this limitation this paper proposes a fuzzy logic system based on CREAM method for human reliability analysis (Hollnagel, 1998) in order to provide more sophisticated estimations of the tunnel operator's performance in safety critical situations. It is deduced that a human reliability analysis component to analyze operator performance, like the fuzzy system proposed here, is important for risk analysts. Consideration of organizational and human factors will enhance risk analysts’ studies and highlight the uncertainty related to human performance variability.     

基于支持向量机的采空区稳定性分级

针对采空区稳定性分级的影响因素众多且关系复杂的特点,提出采用支持向量机理论对采空区稳定性进行分级。根据分级评价指标选取原则,选取岩体结构、地质构造、岩石的质量指标、地下可见水、地下水体、周边开采的影响、相邻空区的情况、工程布置、跨度、面积、高度、矿柱的尺寸及布置、埋藏深度和采空区的规格14个影响因子,建立了采空区稳定性评价指标体系,引入支持向量机理论,选择有向无环图方式构造多类分类器,得到采空区稳定性分级的支持向量机模型。将该模型用于山东黄金矿业西山矿区的25个采空区进行分级,并与未确知测度方法的分级情况     

应用支持向量机评价土壤环境质量

基于野外采样和室内分析相结合的方法,采用电感耦合等离子体质谱法(ICP-MS)对羊草沟煤矿研究区表层土壤样品中的Cd、Cr、Zn、Pb和Cu含量进行测定,应用非线性支持向量机模型中的分类支持向量机,选用sigmoid核函数,利用MATLAB编写程序,进行土壤环境质量评价,并利用模糊综合评判法对评价结果进行验证.在此基础上,运用对应分析方法对样品和变量进行了关联分析,进一步了解重金属污染特征.评价结果表明,研究区土壤环境质量多为Ⅰ类,与模糊综合评判法的相同率达到91.67%,将支持向量机用于土壤环境质量评价是可行的.相比于传统的评价方法,支持向量机采用结构风险最小化原则,将复杂的非线性问题转化为线性问题,成功的解决了多分类、高维运算等问题.     

舟山本岛大气污染输送过程的数值模拟分析

利用HYSPLIT-4后向轨迹模式和NCEP(美国国家环境预报中心)的2012年GDAS(全球资料同化系统)气象数据,结合NO₂、PM_2.5、PM₁₀和SO₂等常规大气污染物的质量浓度数据,对舟山本岛2012年4月、7月、10月和12月的大气污染输送过程进行了模拟,并通过聚类分析和潜在源区分析〔包括PSCF(潜在源贡献)和CWT(浓度权重轨迹)计算〕,确定大气污染传输路径及影响源区. 结果表明:舟山本岛气流后向轨迹呈明显的季节变化特征,4月主要受来自黄海海面气流轨迹的影响,其占总轨迹数的36.7%,ρ(PM₁₀)为(53.24±24.33)μg/m3;7月以途经琉球群岛和东海气流轨迹为主,占总轨迹数的48.4%,对ρ(NO₂)、ρ(PM_2.5)、ρ(PM₁₀)和ρ(SO₂)贡献分别为(24.63±6.33)、(28.60±4.83)、(52.89±18.76)和(8.67±3.11)μg/m3;10月气流轨迹主要来自于东海海面,占总轨迹数的49.2%;12月气流则主要来自辽宁南部和黄海,占总轨迹数的66.1%,对ρ(NO₂)、ρ(PM_2.5)、ρ(PM₁₀)和ρ(SO₂)贡献分别为(28.48±15.14)、(58.71±14.10)、(69.83±38.94)和(20.83±13.28)μg/m3. 舟山本岛PM_2.5的潜在源主要为毗邻城市间局地污染,集中于浙江沿海城市及杭州湾、上海等地.      

Meta-analysis of susceptibility of woody plants to loss of genetic diversity through habitat fragmentation

Shrubs and trees are assumed less likely to lose genetic variation in response to habitat fragmentation because they have certain life-history characteristics such as long lifespans and extensive pollen flow. To test this assumption, we conducted a meta-analysis with data on 97 woody plant species derived from 98 studies of habitat fragmentation. We measured the weighted response of four different measures of population-level genetic diversity to habitat fragmentation with Hedge's d and Spearman rank correlation. We tested whether the genetic response to habitat fragmentation was mediated by life-history traits (longevity, pollination mode, and seed dispersal vector) and study characteristics (genetic marker and plant material used). For both tests of effect size habitat fragmentation was associated with a substantial decrease in expected heterozygosity, number of alleles, and percentage of polymorphic loci, whereas the population inbreeding coefficient was not associated with these measures. The largest proportion of variation among effect sizes was explained by pollination mechanism and by the age of the tissue (progeny or adult) that was genotyped. Our primary finding was that wind-pollinated trees and shrubs appeared to be as likely to lose genetic variation as insect-pollinated species, indicating that severe habitat fragmentation may lead to pollen limitation and limited gene flow. In comparison with results of previous meta-analyses on mainly herbaceous species, we found trees and shrubs were as likely to have negative genetic responses to habitat fragmentation as herbaceous species. We also found that the genetic variation in offspring was generally less than that of adult trees, which is evidence of a genetic extinction debt and probably reflects the genetic diversity of the historical, less-fragmented landscape.     

Varying responses of northeastern North American amphibians to the chytrid pathogen Batrachochytrium dendrobatidis

Chytridiomycosis, caused by Batrachochytrium dendrobatidis (Bd), is widespread among amphibians in northeastern North America. It is unknown, however, whether Bd has the potential to cause extensive amphibian mortalities in northeastern North America as have occurred elsewhere. In the laboratory, we exposed seven common northeastern North American amphibian species to Bd to assess the likelihood of population-level effects from the disease. We exposed larval wood frogs (Lithobates sylvaticus) and postmetamorphic frogs of six other species to two different strains of Bd, a northeastern strain (JEL404) and a strain that caused die-offs of amphibians in Panama (JEL423), under ideal in vitro growth conditions for Bd. Exposed American toads (Anaxyrus americanus) all died; thus, this species may be the most likely to die from Bd-caused disease in the wild. Both Bd strains were associated with mortalities of wood frogs, although half the metamorphs survived. The Bd strain from Panama killed metamorphic green frogs (L. clamitans), whereas the northeastern strain did not, which means novel strains of Bd may lead to death even when local strains may not. No mortality was observed in four species (bullfrogs [L. catesbeianus], northern leopard frogs [L. pipiens], spring peepers [Pseudacris crucifer], and blue-spotted salamanders [Ambystoma laterale]) and in some individuals of green frogs and wood frogs that we exposed. This finding suggests these six species may be Bd vectors. Our results show that systematic exposures of amphibian species to Bd in the laboratory may be a good first step in the identification of species susceptible to Bd-caused declines and in directing regional conservation efforts aimed at susceptible species.     

基于高光谱技术的土壤镉污染分级评价研究

提出一种利用高光谱技术进行土壤镉污染分级评价的方法。以FieldSpec 3地物光谱仪采集厂矿区土壤光谱反射率175份,随机分成校正集（135份）和检验集（40份）。光谱经小波去噪和标准归一化（SNV）处理后,以主成分分析法（PCA）降维。将降维所得的前5个主成分数据为输入变量,分别采用Fisher线性判别、Byes逐步判别、模糊模式识别、BP-ANN判别以及SVM 5种方法建立了土壤镉污染分级评价模型,并利用40个未知样对模型进行检验。结果表明：Fisher线性判别准确率为77.5%,Byes逐步判别与模糊模式识别预测为80.0%,BP-ANN模型预测精度为82.5%,SVM模型预测精度最高,达85.0%。说明采用高光谱技术进行土壤镉污染分级评价是可行,其中SVM是建模的优选算法。     

基于多类支持向量机的化学物质生态危害分类研究

应用多类支持向量机(M-SVMs)方法研究了化学物质生态危害程度的分类,以提高分类的准确性和效率.对采集到的61种环境优先污染物的环境行为和生物毒性方面的7项指标进行相关性分析,建立了M-SVMs分类模型并对数据集进行10折交叉验证以评价模型的分类能力,运用所建模型对7种化学物质的生态危害进行预测.结果表明,去除与鱼毒有信息重叠的溞毒指标,选取鱼毒、藻毒、降解性、蓄积性、分配系数和吸附系数6项指标用于构建M-SVMs分类模型;M-SVMs模型识别率较高,交叉验证平均分类正确率达86.89％;对7种化学物质生态危害的预测结果与实际情况基本相符.