数据挖掘在预测组织事故防控效果中的应用*

为进一步探索数据挖据技术在组织事故预防工作中的融入性与适用性,基于24Model构建事故预控基础模型,通过预测准确率数值及接受者操作特性曲线（ROC曲线）对比分析随机森林（RF）、支持向量机(SVM)、决策树（DT）与神经网络（NN）4种方法对组织事故防控效果的预测性能。结果表明:针对事故率控制（Y1）、职业危害预防（Y2）、财产损失3类预测目标（Y3）,RF方法均能达到较高的准确率及稳定性,具有较优的预测性能。根据特征重要度（FI）排序,明确对组织事故水平影响最显著的因素为安全实践活动认知（SC5）及安全管理程序文件（SMS3）,FI值均大于0.150 0。研究结果可为有效预测组织事故防控效果提供方法依据,同时为企业安全工作的规划设计提供思路。     

喀斯特峡谷区工程性缺水原因及解决途径——以贵州省花江峡谷示范区为例

喀斯特峡谷区河深坡陡,地下水深埋,水资源利用难度大,加之经济和水利工程薄弱,是贵州省工程性缺水的典型区域。通过对花江峡谷示范区工程性缺水的原因剖析,提出一系列解决工程性缺水问题的措施,如雨水收集、表层喀斯特水和基岩裂隙水的开发利用、节水、加强对水利工程设施的管理和维护、加大水利投入以及提高植被覆盖率等,这对解决喀斯特峡谷及类似区域的工程性缺水问题有一定的指导和参考意义。     

Modelling the fate of hydrophobic organic contaminants in a boreal forest catchment: A cross disciplinary approach to assessing diffuse pollution to surface waters

The fate of hydrophobic organic compounds (HOCs) in soils and waters in a northern boreal catchment was explored through the development of a chemical fate model in a well-characterised catchment system dominated by two land types: forest and mire. Input was based solely on atmospheric deposition, dominated by accumulation in the winter snowpack. Release from soils was governed by the HOC concentration in soil, the soil organic carbon fraction and soil-water DOC content. The modelled export of selected HOCs in surface waters ranged between 11 and 250 ng day⁻¹ during the snow covered period, compared to 200 and 9600 ng/d during snow-melt; highlighting the importance of the snow pack as a source of these chemicals. The predicted levels of HOCs in surface water were in reasonable agreement to a limited set of measured values, although the model tended to over predict concentrations of HOCs for the forested sub-catchment, by over an order of magnitude in the case of hexachlorobenzene and PCB 180. This possibly reflects both the heterogeneity of the forest soils and the complicated and changing hydrology experienced between the different seasons.     

Evaluating and reducing a model of radiocaesium soil-plant uptake

An existing model of radiocaesium transfer to grasses was extended to include wheat and barley and parameterised using data from a wide range of soils and contact times. The model structure was revised and evaluated using a subset of the available data which was not used for model parameterisation. The resulting model was then used as a basis for systematic model reduction to test the utility of the model components. This analysis suggested that the use of 4 model variables (relating to radiocaesium adsorption on organic matter and the pH sensitivity of soil solution potassium concentration) and 1 model input (pH) are not required. The results of this analysis were used to develop a reduced model which was further evaluated in terms of comparisons to observations. The reduced model had an improved empirical performance and fewer adjustable parameters and soil characteristic inputs.     

Phospholipid fatty acid biomarkers in a freshwater periphyton community exposed to uranium: discovery by non-linear statistical learning

Phospholipid fatty acids (PLFA) have been widely used to characterize environmental microbial communities, generating community profiles that can distinguish phylogenetic or functional groups within the community. The poor specificity of organism groups with fatty acid biomarkers in the classic PLFA-microorganism associations is a confounding factor in many of the statistical classification/clustering approaches traditionally used to interpret PLFA profiles. In this paper we demonstrate that non-linear statistical learning methods, such as a support vector machine (SVM), can more accurately find patterns related to uranyl nitrate exposure in a freshwater periphyton community than linear methods, such as partial least squares discriminant analysis. In addition, probabilistic models of exposure can be derived from the identified lipid biomarkers to demonstrate the potential model-based approach that could be used in remediation. The SVM probability model separates dose groups at accuracies of ∼87.0%, ∼71.4%, ∼87.5%, and 100% for the four groups; Control (non-amended system), low dose (amended at 10 μg U L⁻¹), medium dose (amended at 100 μg U L⁻¹), and high dose (500 μg U L⁻¹). The SVM model achieved an overall cross-validated classification accuracy of ∼87% in contrast to ∼59% for the best linear classifier.     

Projecting Climate Change Effects on Forest Net Primary Productivity in Subtropical Louisiana,USA

This study projected responses of forest net primary productivity (NPP) to three climate change scenarios at a resolution of 5 km × 5 km across the state of Louisiana, USA. In addition, we assessed uncertainties associated with the NPP projection at the grid and state levels. Climate data of the scenarios were derived from Community Climate System Model outputs. Changes in annual NPP between 2000 and 2050 were projected with the forest ecosystem model PnET-II. Results showed that forest productivity would increase under climate change scenarios A1B and A2, but with scenario B1, it would peak during 2011–2020 and then decline. The projected average NPP under B1 over the years from 2000 to 2050 was significantly different from those under A1B and A2. Forest NPP appeared to be primarily a function of temperature, not precipitation. Uncertainties of the NPP projection were due to large spatial resolution of the climate variables. Overall, this study suggested that in order to project effects of climate change on forest ecosystem at regional level, modeling uncertainties could be reduced by increasing the spatial resolution of the climate projections.     

人口长期均衡发展及其评价监测模型的构建与应用

在借鉴已有理论的基础上,界定了人口长期均衡发展的内涵,提出人口长期均衡包含人口内部均衡和人口外部均衡两个方面,人口内部均衡包括人口数量、人口质量和人口结构三个维度,人口外部均衡包括人口与经济、人口与社会、人口与资源、人口与环境四个维度;在此基础上构建了人口长期均衡发展的指标体系和评价模型,指标体系设定为三级,第一级包括人口内部均衡和人口外部均衡两个方面,第二级由人口数量、人口质量、人口结构等七个部分构成;第三级指标包括反映内部均衡的7个指标以及反映外部均衡的13个指标.并使用层次分析法确定了指标的权重;采用2000-2007年31个省级单位的数据对人口长期均衡发展评价模型进行了验证,并对人口总体均衡水平以及影响人口均衡发展的关键要素进行了实证分析.结果表明.北京、上海和浙江分列前三,人口均衡发展程度最高.从影响均衡程度的关键因素看.人口质量和人口结构是制约人口自身均衡的主要因素,人口与社会的均衡状况对人口与外部系统均衡制约最大,其次是人口与经济的均衡状况.     

Quantification of nitrate leaching from German forest ecosystems by use of a process oriented biogeochemical model

Simulations with the process oriented Forest-DNDC model showed reasonable to good agreement with observations of soil water contents of different soil layers, annual amounts of seepage water and approximated rates of nitrate leaching at 79 sites across Germany. Following site evaluation, Forest-DNDC was coupled to a GIS to assess nitrate leaching from German forest ecosystems for the year 2000. At national scale leaching rates varied in a range of 0–>80 kg NO₃–N ha⁻¹ yr⁻¹ (mean 5.5 kg NO₃–N ha⁻¹ yr⁻¹). A comparison of regional simulations with the results of a nitrate inventory study for Bavaria showed that measured and simulated percentages for different nitrate leaching classes (0–5 kg N ha⁻¹ yr⁻¹:66% vs. 74%, 5–15 kg N ha⁻¹ yr⁻¹:20% vs. 20%, >15 kg N ha⁻¹ yr⁻¹:14% vs. 6%) were in good agreement. Mean nitrate concentrations in seepage water ranged between 0 and 23 mg NO₃–N l⁻¹.     

Comparison of the daily global solar radiation from different data sources in Northwest China climate

The solar radiation data are of high importance to the solar energy systems. Conventional methods to obtain the solar radiation data are from weather stations, solar radiation models, commercial software databases, and field measurements. In the present study, a new daily global solar radiation model is proposed, by combining the quadratic function of sunshine fraction and sine function of the day of the year. The solar radiation model calculated data are then compared with China Meteorological Data Sharing System (CMDSS) data, TRNSYS data, and field-measured data in Northwest China climate. It is found that the newly proposed solar radiation model has better performance than the other nine solar radiation models in the literature. The solar radiation model calculated data fit well with the CMDSS annually average data. The TRNSYS data are a bit larger than the CMDSS annually average data in summer half year and a little smaller than those in winter half year. The solar radiation model and the CMDSS annually average data have the best correlation, whereas the TRNSYS data and the field-measured data have the worst correlation. The solar radiation model calculated data have the best correlation with the other three data sources.     

Neural network forecasting of air pollutants hourly concentrations using optimised temporal averages of meteorological variables and pollutant concentrations

The new method for the forecasting hourly concentrations of air pollutants is presented in the paper. The method was developed for a site in urban residential area in city of Zagreb, Croatia, for four air pollutants (NO₂, O₃, CO and PM₁₀). Meteorological variables and concentrations of the respective pollutant were taken as predictors. A novel approach, based on families of univariate regression models, was employed in selecting the averaging intervals for input variables. For each variable and each averaging period between 1 and 97 h, a separate model was built. By inspecting values of the coefficient of correlation between measured and modelled concentrations, optimal averaging periods for each variable were selected. A new dataset for building the forecasting model was then calculated as temporal moving averages (running means) of former variables. A multi-layer perceptron type of neural networks is used as the forecasting model. Index of agreement, calculated for the entire dataset including the data for model building, ranged from 0.91 to 0.97 for the respective pollutants. As suggested by the analysis of the relative importance of the input variables, different agreements for different pollutants are likely due to different sources and production mechanisms of investigated pollutants. A comparison of the new method with more traditional method, which takes hourly averages of the forecast hour as input variables, showed similar or better performance. The model was developed for the purpose of public-health-oriented air quality forecasting, aiming to use a numerical weather forecast model for the prediction of the part of input data yet unknown at the forecasting time. It is to expect that longer term averages used as inputs in the proposed method will contribute to smaller input errors and the greater accuracy of the model.