北京市城市道路交通噪声的灰色预测

以北京市1991~2002年城市道路交通噪声监测数据为基础,运用灰色系统理论,建立了灰色GM(1,1)预测模型,并采用残差、均方差比值、小误差概率检验等3种办法对所建模型的拟合精度进行了检验.结果表明,此模型精度为一级灰色预测模型.预测精度较高,平均拟合偏差为0.40%.对2003~2005年噪声的预测精度平均高达99.9%.利用常规GM(1,1)模型、新陈代谢GM(1,1)模型和多维灰数递补GM(1,1)模型等三种预测模型对本市未来3年(2006~2008年)的城市道路交通噪声进行了预测.     

基于GM（1,1）模型的城镇生活污水排放量预测

根据2000--2012年全国城镇生活污水排放量数据，建立了相应的GM（1，1）模型和预测函数。通过后验差检验等对预测函数进行了评估，并对2013--2017年城镇生活污水排放量进行了预测，结果表明，2013--2017年全国城镇生活污水的排放量逐年增加，呈上升趋势，从2013年的477．2736亿t上升到2017年的624．1022亿t；灰色预测模型和方法简单易用，利用较少的数据即可进行精度较高的预测。     

用灰色系统理论对乌鲁木齐市大气降水pH值进行预测分析

运用灰色系统理论建立灰色模型ＧＭ，对该模型进行了检验和残差修正。并用其模型分析和预测乌鲁木齐市１９９４＝１９９８年大气降水ｐＨ值。     

灰色新陈代谢GM(1,1)模型在城市道路交通噪声预测中的应用

以郑州１９９１～１９９６年城市道路交通噪声监测的数据,运用灰色系统理论,建立了常规ＧＭ（１,１）和新陈代谢ＧＭ（１,１）预测模型。经用四种不同方法对两种模型的精度进行检验,结果表明,新陈代谢ＧＭ（１,１）模型优于常规ＧＭ（１,１）模型,其精度更高,不失为预测城市道路交通噪声的一种好方法。应用该模型,对郑州城市道路交通噪声未来１０年进行预测,其结果符合郑州城市的实际情况     

公路噪声模式存在的问题与处理

采用《公路建设项目环境影响评价规范》（试行）中推荐的环境噪声影响预测模式（规范模式）和以“城市道路交通噪声预测统计模型”为基础的综合模式分别对高速公路交通噪声进行了计算，并利用实测资料对两种模式进行了检验和比较分析。指出规范模式存在着对低车流量噪声预测值偏大（约4－5dB）等问题，而综合模式可以部分弥补这些缺陷，使预测能力明显提高。     

运用灰色系统的GM（1，1）模型对城市噪声的综合预测

应用GM（1，1）模型，对大同市各功能区环境噪声进行综合预测，得出了大同市未来几年城市噪声的变化趋势。     

江南水网地区中小城市空气质量预报研究

利用湖州2002年7月至2004年12月的空气污染浓度资料及常规气象资料,探讨了江南水网地区气象因素对环境空气质量的影响,并结合区域气候特征及污染现状对影响环境空气质量的因素进行了研究分析.将天气类型科学分类,分别建立了区域空气质量预报的总体回归模型和分类回归模型.两种模型均可通过统计检验,分类统计模型较总体统计模型实际业务预报精度有所提高,最后采用DELPHI语言将预报流程制作了软件.     

环境振动的灰色预测模型

利用GIM(1)的非时序直接建模法预测研究建筑施工的环境振动 ,并将GIM(1)模型与GM (1,1)模型进行比较分析 ,结果表明GIM(1)模型的拟合精度优良 ,对原始资料中白化信息的利用更加丰富 ,拓宽了GIM (1)模型在环境科学领域中的应用范围     

灰色系统GM（1,1）模型在城市地下水总硬度预测中的应用

应用灰色系统GM(1,1)模型,对地下水中总硬度的变化作了预测.检验结果表明,该模型精度较高,是一种较好的预测方法.     

自然降尘与常规气象因子的相关分析

笔者整理统计了１９９２年至１９９４年逐月自然降水和与之相应的常规气象因子监测数据，找到了自然降尘的季节性分布规律。在此基础上，求得了自然降尘分别与月平均气温、月平均气压、月平均风速、月平均温度、月降水量、月蒸发量的相关系数，并进行了检验，为污染预报工作提供了科学依据。     

Modeling of non-point source pollution in a Mediterranean drainage basin

SWAT ver. 2000 was used to predict hydrographs, and sediment, nitrate and total phosphorus loadings from a 1349 km² mountainous/agricultural watershed in Northern Greece. The model was calibrated and verified using continuous meteorological data from eight stations within the drainage area, and runoff, sediment and nutrient concentrations measured at nine stations located within the main tributaries of the watershed, for the time period from May 1st, 1998 to January 31st, 2000. Model validation methodology and resulting input parameters appropriate for Mediterranean drainage basins are presented. Predicted by the model hydrographs, sedimentographs and pollutographs are plotted against observed values and show good agreement. Model performance is evaluated using the root mean square error computation and scattergrams of predicted versus observed data. The validated model is also used to test the effectiveness of three alternative cropping scenarios in reducing nutrient loadings from the agricultural part of the watershed. The study showed that this model, if properly validated, can be used effectively in testing management scenarios in Mediterranean drainage basins.     

Quantifying urban heat island intensity in Hong Kong SAR, China

This paper addresses the methodological concerns in quantifying urban heat island (UHI) intensity in Hong Kong SAR, China. Although the urban heat island in Hong Kong has been widely investigated, there is no consensus on the most appropriate fixed point meteorological sites to be used to calculate heat island intensity. This study utilized the Local Climate Zones landscape classification system to classify 17 weather stations from the Hong Kong Observatory’s extensive fixed point meteorological observation network. According to the classification results, the meteorological site located at the Hong Kong Observatory Headquarters is the representative urban weather station in Hong Kong, whereas sites located at Tsak Yue Wu and Ta Kwu Ling are appropriate rural or nonurbanized counterparts. These choices were validated and supported quantitatively through comparison of long-term annual and diurnal UHI intensities with rural stations used in previous studies. Results indicate that the rural stations used in previous studies are not representative, and thus, the past UHI intensities calculated for Hong Kong may have been underestimated.     

融合拓扑信息与气象信息的空气质量预测网络

通过空气质量监测数据对正在形成或即将到来的空气污染进行预测是一项具有重要意义的工作，而空气质量监测站只能检测其周围一定范围内的空气污染情况。为了衡量整个城市的空气污染情况，获取任意时间、任意位置的空气质量信息，结合交叉注意力机制，提出了一种融合拓扑信息与气象信息的空气质量预测网络(CGMIM)。将西安市空气质量监测数据与气象数据转换为图像拼接起来，作为输入信息。在高阶非线性时空动态神经网络(MIM)的基础上引入注意力机制，并增加拓扑图编码器模块，提高模型提取能力以及对空气质量监测数据中的空间特征的利用率。最后，使用时空损失函数替代传统的均方误差损失函数，提高模型对空间关系的关注。结果表明：CGMIM网络模型能够在准确预测的同时，对位置区域合理填充，能够有效提升空气质量监测数据的空间分辨率。     

基于GIS的新疆年降水量空间插值精度比较研究

以新疆具有完整资料的90个气象站点中的75个站点的基本地理信息和1995—2008年平均降水数据资料为基础,以ArcGIS为数据处理平台,分别采用反距离加权法、径向基函数法、普通克里格法、协克里格法对新疆降水空间分布进行插值,并通过没有参与插值的15个气象站点的实际降水量与插值结果进行相对误差分析。研究结果表明:在年降水量插值的4种算法中,普通克里格法(Kriging)的计算精度最高。利用最优的插值方法的栅格降水空间数据库建立的多年降水资源信息系统,可快速计算研究区内任一地域单元中降水的总量及其空间变化,对降水资源模拟、工程规划、发展战略及决策管理均具有重要的意义。     

基于OPAQ的城市空气质量预报系统研究

空气质量预测在国内的关注度日益提高,传统的空气质量预测系统通常运用数值化学传输模型,利用物理方程来计算污染物的扩散、沉降和化学反应。而化学传输模型的预测准确性很大程度上需要依赖详细的污染源排放信息和气象模型的输出结果。基于统计模型的OPAQ空气质量预报业务系统,采用人工神经网络算法,可预测各污染物的日均值或日最大值。并对北京空气质量预报的结果进行了评价,OPAQ空气质量预报业务系统对空气质量预测的准确性较高,能够利用较低的计算资源得到较为准确的预测结果。与数值预报相比,OPAQ空气质量预报业务系统不需要大量的基础数据作为输入,可弥补数值预报的不足,并成为数值预报的有力补充。     

Monitoring of the Endangered Pinna nobilis Linné, 1758 in the Mar Grande of Taranto (Ionian Sea, Italy)

Two methods were used to calculate the meteorologically adjusted ground level ozone trends in southern Taiwan. The first method utilized is a robust linear regression method. The second approach uses a multilayer perceptron (MLP) artificial neural network (ANN) method. The observations obtained from 16 monitoring stations were analyzed and divided into six groups by hierarchical divisive clustering procedure. The daily maximum 1 and 8 h ozone concentrations for each group are then calculated. The meteorologically adjusted trends obtained by linear regression and MLP methods are smaller than the unadjusted trends for all groups and average time. It indicts that the meteorological conditions in Taiwan tend to increase ambient ozone concentrations in recent years.     

On the evaluation of box model results: the case of BOXURB model

In the present paper, the BOXURB model results, as they occurred in the Greater Area of Athens after model application on an hourly basis for the 10-year period 1995-2004, are evaluated both in time and space in the light of observed pollutant concentrations time series from 17 monitoring stations. The evaluation is performed at a total, monthly, daily and hourly scale. The analysis also includes evaluation of the model performance with regard to the meteorological parameters. Finally, the model is evaluated as an air quality forecasting and urban planning tool. Given the simplicity of the model and the complexity of the area topography, the model results are found to be in good agreement with the measured pollutant concentrations, especially in the heavy traffic stations. Therefore, the model can be used for regulatory purposes by authorities for time-efficient, simple and reliable estimation of air pollution levels within city boundaries.     

Analysis and assessment of ground-level ozone measured at two stations in Tehran

In this study, variations of ambient ozone level are thoroughly analysed according to the monitored data in a mixed residential, commercial and industrial city, Tehran, based on considering the meteorological factors. Ozone as a pollutant shows typical annual, weekly and diurnal cycles. This analysis has shown that the ozone level concentrations were below the WHO guidelines in Tehran during 2000–2003. The relation between ozone level at two different stations (Aghdasieh and Fatemi) is found (r?=?0.51). The ozone level response to meteorological parameters is investigated. The results suggest that the ozone level is affected (positively or negatively) by meteorological conditions, e.g. relative humidity, solar radiation, air temperature, wind speed and wind direction.     

Near-surface ozone levels and trends at rural stations in France over the 1995–2003 period

There is a considerable interest in quantifying near-surface ozone concentrations and associated trends, as they serve to define the impacts on ozone of the anthropogenic precursors reductions and to evaluate the effects of emission control strategies. A statistical test has been used to the ozone air concentrations measured in the French rural monitoring network stations, called MERA, in order to bring out spatio-temporal trends in air quality in France over the 1995-2003 period. The non-parametric Mann-Kendall test has been developed for detecting and estimating monotonic trends in the time series and applied in our study at annual values: mean, 98th percentile and median based on hourly averaged ozone concentrations and applied to daily maxima. In France, when averaged overall 9 stations between 1995 and 2003, a slight increasing trend of the O(3) levels (+0.6 +/- 1.3% year( - 1)) is observed, which is strongly influenced by the concentrations of the high altitude stations. In stations below 1000 m a mean rate of -0.48% year( - 1) from annual average concentrations, of -0.45% year( - 1) for medians and of +0.56% year( - 1) for P.98 over the 1995-2003 period were obtained. In stations above 1,000 m a mean rate of +1.75% year( - 1) from annual averages values, of +4.05% year( - 1) for medians and of +2.55% year( - 1) for P.98 were calculated over the 1997-2003 period. This situation is comparable to the one observed in other countries. In Europe and in France a reduction of precursor emissions is observed whereas a slight increasing trend of the O(3) levels is observed over the 1995-2003 period. One reason is the non-linearity of chemical ozone production with respect to precursor emissions. Possible explanations are an increase in near-surface ozone values caused by a reduced ozone titration by reduced NO( x ), the meteorological parameters change, an increase in bio-geogenic compound concentrations, the intercontinental transport from North America and Asia and the influence of stratospheric-tropospheric exchanges. These possible explanations must be interpreted carefully as on the short time scales considered.     

基于复合LSTM模型的PM2.5浓度预测

在对淄博市19个空气质量监测站点监测数据进行分析后，提出了一种基于机器学习的复合模型——灰色关联度分析(GRA)-改进的完备总体经验模态分解(ICEEMD)-长短期记忆网络(LSTM)模型。通过分析淄博市2019年大气污染物和气象数据，选用LSTM模型预测PM_2.5浓度。由于传统单一模块机器学习模型具有训练时间较长和预测精度较低的问题，提出了复合LSTM模型。该模型由3部分组成：GRA,用于PM_2.5浓度影响因素变量筛选；ICEEMD,用于PM_2.5分解、分量筛选和原始大气污染物及气象数据处理；LSTM,用于PM_2.5浓度预测。预测结果表明：淄博市中部丘陵地带PM_2.5浓度高于南部山区和北部平原，东部高于西部；淄博市逐月PM_2.5浓度呈“U”形分布，1月最高，8月最低；淄博市PM_2.5浓度受PM₁₀和CO影响较大，受湿度和温度影响较小。对比单一LSTM模型和GRA-LSTM模型，GRA-ICEEMD-LSTM模型...