西藏芒康滇金丝猴国家级自然保护区变化监测

以第一次全国地理国情普查与基础性地理国情监测数据为数据源,开展西藏芒康滇金丝猴国家级自然保护区内2015、2016、2017、2018年自然保护区建设与管理情况监测,分别从人工地物、林草资源、露天采掘场等方面进行统计,分析人类活动对自然保护区内的生态系统胁迫以及区内生态恢复情况等。     

广州海洋观测网规划建设及观测数据应用探讨

阐述了广州海洋观测网建设的迫切性和必要性,对海洋观测网建设在海洋防灾减灾、海岸工程建设及海水养殖等方面的重要意义进行了分析说明。总结广州市海洋观测网建设现状和成果,并利用已有观测数据对2017年台风^“天鸽^”和^“帕卡^”期间广州海域水文气象特征、2015—2017年枯水期咸潮入侵特征进行了分析。研究表明, 现有观测数据对台风和咸潮活动均具有较明显的反应,且对海洋防灾减灾工作具有重要的意义;为更有效地利用已有观测数据,将观测数据价值最大化,对观测网建设及运行过程中观测工作机制及业务化流程、数据质量控制、观测站布设合理性以及数据应用分析这几个方面存在的问题进行了探讨,提出了广州海洋观测网的优化、改进措施及未来发展方向。     

基于人工神经网络的街道峡谷NO_x浓度的数值模型研究

通过对反向传播人工神经网络的算法和网络结构的研究,发现拟牛顿算法训练速度较快,能够较好地接近误差目标值,同时建立了包括输入层、隐含层、输出层的人工神经网络三层拓扑结构。通过对街道峡谷人工神经网络的训练,模拟计算了街道峡谷NOx浓度分布值。结果显示,训练误差和测试误差比为1.11,训练样本的模拟值与实测值的相关系数为0.93,测试样本的模拟值与实测值的相关系数为0.87,模拟值与实测值的相关系数均高于显著水平为α=0.05与α=0.01所对应检验性表的相关系数临界值。该模型能够用于街道峡谷污染物浓度的模拟计算,具有较好的泛化能力。     

Transport and distribution of artificial gamma-emitting radionuclides in the River Yenisei and its sediment

Discharges from the Krasnoyarsk Mining and Chemical Industrial Complex (KMCIC) near Krasnoyarsk resulted in radioactive contamination of sediments of the River Yenisei. Between 1999 and 2006, 16 sediment cores were collected at different positions 15-1500 km downstream from the discharge point. The concentration of artificial gamma-emitting radionuclides (¹³⁷Cs, ⁶⁰Co, ¹⁵²Eu, and ²⁴¹Am) was determined with the objective to analyze the migration processes leading to the transport of these radionuclides along the river and to their vertical distribution within the sediment. In cores taken in the vicinity of the reactors, the average activity concentration of ¹³⁷Cs, ¹⁵²Eu, and ⁶⁰Co was about 1000 Bq kg⁻¹, and the activity concentration of ²⁴¹Am was about 20 Bq kg⁻¹. Contamination levels of artificial radionuclides were decreasing with increasing distance downstream the KMCIC: The fastest decrease of average activity by a factor of 10 over a distance of 300 km was observed for ²⁴¹Am, whereas for ¹³⁷Cs this decrease occurred over a distance of 1100 km. Sequential extraction experiments revealed that in all depths and at all distances the studied radionuclides were tightly bound to the sediment.To investigate the mechanisms of transport of the ¹³⁷Cs and ⁶⁰Co contamination, mathematical models have been used to describe the contamination in the river water and within the sediments.     

基于人工神经网络的街道峡谷NOx浓度的数值模型研究

通过对反向传播人工神经网络的算法和网络结构的研究,发现拟牛顿算法训练速度较快,能够较好地接近误差目标值,同时建立了包括输入层、隐含层、输出层的人工神经网络三层拓扑结构。通过对街道峡谷人工神经网络的训练,模拟计算了街道峡谷NOx浓度分布值。结果显示,训练误差和测试误差比为1．11,训练样本的模拟值与实测值的相关系数为0．93,测试样本的模拟值与实测值的相关系数为0．87,模拟值与实测值的相关系数均高于显著水平为α=0．05与α=0．01所对应检验性表的相关系数临界值。该模型能够用于街道峡谷污染物浓度的模拟计算,具有较好的泛化能力。     

前馈神经网络在环境评价和预测中的应用进展

本文针对水环境中复杂的不确定性及非线性关系,在水环境不确定性分析的基础上,详细阐述了以BP网络和RBF网络为代表的前馈神经网络法的基本原理,分析了两种方法的优点。同时,本文对两种方法在水环境影响评价工作中的应用现状进行总结,分析了两种方法的研究发展趋势。     

基于人工神经网络的突发环境事故应急系统

突发环境事故现场具有形势紧迫、信息获取困难等特点,引入人工神经网络技术(Artificial Neural Networks,ANN)设计开发的突发环境事故应急决策支持系统(ANN-EEDSS),能够完成现场应急指挥调度的定量化决策支持。关键技术包括突发环境事故神经网络构建方法,事故信息与应急措施的数字化方法,决策信息的翻译表示,事故影响因素的权重分析。最后阐述了系统现场运行和工作的流程。     

重庆市公路网络通达性格局研究

以重庆市38个区县公路网络为研究对象,以最短距离为原则,在构建公路网络拓扑结构的基础上,计算加权平均旅行时间指标和通达性系数来衡量城市通达性水平的高低,并对通达性水平的特点和空间格局进行分析。结论如下:(1)38个区县的通达性水平呈现出"放射状同心圆"特征,以重庆市主城区为圆心,随着同心圆的半径增加,通达性水平逐渐降低,形成明显的环状分异态势;(2)路网的拓扑连接水平较低,路网等级体系不够完善,网络的辐射效率不高,网络结构相对脆弱,网络的有效性较低;(3)高等级路网发育不均衡,区县间出现通达性差异很大的现象,影响区域的协调发展。     

供水管网系统地震易损性风险评估

供水管网易损性风险评估就是通过识别和分析系统的易损性,找出管网中相对比较薄弱的部分,为确定安全策略和采取控制措施提供理论依据。借鉴网络脆弱性评估理论和方法,基于震后供水管网的连通性分析和水力分析结果,采用分层的分析方法,以管网的拓扑位置和节点服务特性作为影响管网易损性的因素,分配给各节点一定的权值,计算其易损性指标,采用定性与定量相结合的方法评判其易损性程度。对一实际管网进行地震易损性风险评估,说明了该方法的有效性与合理性。     

Performance of data acceptance criteria over 50 months from an automatic real-time environmental radiation surveillance network

The automatic real-time environmental radiation surveillance network of Catalonia (Spain) comprises two subnetworks; one with 9 aerosol monitors and the other with 8 Geiger monitors together with 2 water monitors located in the Ebre river. Since September 2006, several improvements were implemented in order to get better quality and quantity of data, allowing a more accurate data analysis. However, several causes (natural causes, equipment failure, artificial external causes and incidents in nuclear power plants) may produce radiological measured values mismatched with the own station background, whether spurious without significance or true radiological values. Thus, data analysis for a 50-month period was made and allowed to establish an easily implementable statistical criterion to find those values that require special attention. This criterion proved a very useful tool for creating a properly debugged database and to give a quick response to equipment failures or possible radiological incidents. This paper presents the results obtained from the criterion application, including the figures for the expected, raw and debugged data, percentages of missing data grouped by causes and radiological measurements from the networks. Finally, based on the discussed information, recommendations for the improvement of the network are identified to obtain better radiological information and analysis capabilities.