基于优化的深度极限学习机的柴油车NOx排放预测

用麻雀搜索算法优化的深度极限学习机（SSA-DELM）构建柴油车NOx排放预测模型,对柴油车低速、中速和高速状态下的NOx排放进行预测,并将此模型性能与深度极限学习机（DELM）模型性能进行对比分析。结果表明：SSA-DELM模型的预测效果较好,在低速、中速、高速状态下该模型平均绝对百分比误差MAPE分别为0.061 0、0.044 9、0.039 1;在低速、中速、高速状态下SSA-DELM模型的性能评价指标比DELM模型性能评价指标分别优约23%、44%、11%。     

基于机器学习的短临预报方法及其在空气质量保障中的应用

基于多个数值模式(NAQPMS、CMAQ)预报结果和站点观测资料,采用岭回归机器学习方法构建了一种多污染物短临预报方法,并应用于2017年厦门金砖国家峰会和2019年武汉军运会空气质量保障工作,评估总结了短临预报方法的可行性和预报效果.结果表明:短临预报表现出很好的预报能力,可以有效改进数值模式的预报趋势和量值.从日均...     

Application of SMR to Modeling Watersheds in the Catskill Mountains

Very few hydrological models commonly used in watershed management are appropriate for simulating the saturation excess runoff. The Soil Moisture Routing model (SMR) was developed specifically to predict saturation excess runoff from variable source areas, especially for areas where shallow interflow controls saturation. A recent version of SMR was applied to two rural catchments in the Catskill Mountains to evaluate its ability to simulate the hydrology of these systems. Only readily available meteorological, topographical, and landuse information from published literature and governmental agencies was used. Measured and predicted streamflows showed relatively good agreement; the average Nash–Sutcliffe efficiency for the two watersheds were R ²=72% and R ²=63%. Distributed soil moisture contents and the locations of hydrologically sensitive areas were also predicted well.     

运用方差分析法分析白城市月亮湖水库水质

运用方差分析法中的相关系数法和SNK检验方法,分析了1998～2007年白城月亮湖水库中主要污染物的监测数据.结果显示,溶解氧与氨氮具有线性关系,与高锰酸盐指数正相关,pH值在10年中没有显著变化.     

Application of Ecological Classification and Predictive Vegetation Modeling to Broad-Level Assessments of Ecosystem Health

The Little Missouri National Grasslands (LMNG) of western North Dakota support the largest permitted cattle grazing use within all lands administered by the USDA, Forest Service, as well as critical habitat for many wildlife species. This fact, coupled with the need to revise current planning direction for range allotments of the LMNG, necessitated that a broad-level characterization of ecosystem integrity and resource conditions be conducted across all lands within the study area (approximately 800,000 hectares) in a rapid and cost-effective manner. The approach taken in this study was based on ecological classifications, which effectively utilized existing field plot data collected for a variety of previous inventory objectives, and their continuous spatial projection across the LMNG by maps of both existing and potential vegetation. These two map themes represent current and reference conditions (existing vs. potential vegetation); their intersection allowed us to assign various ecological status ratings (i.e., ecosystem integrity and resource condition) based on the degree of departure between current and reference conditions. In this paper, we present a brief review of methodologies used in the development of ecological classifications, and also illustrate their application to assessments of rangeland health through selected maps of ecological status ratings for the LMNG.     

Statistical Modelling of Extreme Rainfall Indices using Multivariate Extreme Value Distributions

Environmental Modeling & Assessment - Multivariate extreme value models are used to investigate the combined behaviour of several weather variables. To investigate joint dependence of extreme...     

基于机器学习的浮标实时自动化赤潮预警研究

实现赤潮预警对于减轻海洋环境灾害、避免海洋产业特别是海洋渔业重大经济损失具有重要意义。针对当前水文监测数据海量却难以实现实时自动化监测与预警,特别是难以利用传统监测手段实现对危害更大的赤潮的精准实时预测这一显著问题,提出利用浮标数据作为依据,借助机器学习在大数据分析和智能决策方面的优势,建立一种新颖的双重递进式赤潮预警机制的方法。首先,通过相关算法分析历史数据,以确认赤潮初步预警阈值;其次,对叶绿素a、pH、溶解氧等重要监测指标的当前和阶段性变化进行初步分析,判断是否达到预警触发条件;然后,进一步联合分类、回归、聚类、神经网络等机器学习相关方法,对数据进行深度挖掘;最后,通过这种递进式的机制对短期内是否会发生赤潮作出判断,以实现赤潮自动化预警预报。在此基础上,利用宁波梅山湾实际监测数据,证实了该方法在赤潮实时自动化预警中的有效性。     

基于机器学习遥感算法的大型漂浮藻面积提取方法研究

采用基于机器学习的多层感知机算法,利用GOCI(Geostationary Ocean Color Imager)传感器获取的瑞利校正反射率数据,对东中国海大型漂浮藻进行遥感自动识别,采用线性混合像元分解来计算大型漂浮藻的覆盖面积,并利用膨胀和侵蚀法进行大型漂浮藻的分布面积计算。利用L8/OLI(Landsat 8/Operational Land Imager)高空间分辨率资料进行验证,结果表明,基于机器学习遥感算法针对GOCI提取的大型漂浮藻覆盖面积,与L8/OLI结果十分接近,R2达到0. 959,平均绝对误差和平均相对误差分别为39. 32 km2和18. 15%。     

模糊数学在丹江口水库富营养化评价中的应用

丹江口水库是南水北调中线工程水源地,其水质安全将直接影响调水工程的成败。基于2013年6月的水质监测结果,通过对水质指标的聚类,将丹江口水库划分子库区,并尝试使用模糊数学方法进行水质富营养化评价。结果表明,丹江口库区可分为6个子库区,各库区在透明度、溶解氧、浊度、高锰酸盐指数、总磷、硝酸盐氮和叶绿素a等理化指标均有显著差异;富营养化评价显示,库区总体营养状态级别为中营养。其中,坝前区域(库区Ⅰ)和丹库主体部分(库区Ⅵ)营养状态级别为贫营养,汉库原河道区(库区Ⅱ)为中营养,汉江入库口(库区Ⅲ)、汉库最大库湾(库区Ⅳ)、丹江入库口(库区Ⅴ)为轻度富营养。模糊综合评价法较好反映出水质的模糊性、连续性,使评价结果更加准确可靠。     

Support Vector Machine Modeling Using Particle Swarm Optimization Approach for the Retrieval of Atmospheric Ammonia Concentrations

This study was performed in order to improve the estimation accuracy of atmospheric ammonia (NH₃) concentration levels in the Greater Houston area during extended sampling periods. The approach is based on selecting the appropriate penalty coefficient C and kernel parameter σ ². These parameters directly influence the regression accuracy of the support vector machine (SVM) model. In this paper, two artificial intelligence techniques, particle swarm optimization (PSO) and a genetic algorithm (GA), were used to optimize the SVM model parameters. Data regarding meteorological variables (e.g., ambient temperature and wind direction) and the NH₃ concentration levels were employed to develop our two models. The simulation results indicate that both PSO-SVM and GA-SVM methods are effective tools to model the NH₃ concentration levels and can yield good prediction performance based on statistical evaluation criteria. PSO-SVM provides higher retrieval accuracy and faster running speed than GA-SVM. In addition, we used the PSO-SVM technique to estimate 17 drop-off NH₃ concentration values. We obtained forecasting results with good fitting characteristics to a measured curve. This proved that PSO-SVM is an effective method for estimating unavailable NH₃ concentration data at 3, 4, 5, and 6 parts per billion (ppb), respectively. A 4-ppb NH₃ concentration had the optimum prediction performance of the simulation results. These results showed that the selection of the set-point values is a significant factor in compensating for the atmospheric NH₃ dropout data with the PSO-SVM method. This modeling approach will be useful in the continuous assessment of NH₃ sensor discrete data sources.     

基于机器学习的大气PM2.5中金属浓度预测模型的研究

基于2013—2018年哈尔滨市气象数据、大气污染物数据和细颗粒物（PM_2.5）中金属成分数据,采用机器学习方法探索大气PM_2.5中金属浓度预测模型,并选择最优模型进行污染物浓度预测。结果表明,多元线性回归（MLR）、人工神经网络（BP-ANN）、支持向量机（SVM）和随机森林（RF）4种模型中,RF对大气PM_2.5中5种金属［锑（Sb）、砷（As）、铅（Pb）、镉（Cd）、铊（Tl）］的浓度预测效果最佳,在训练集和测试集中表现均较稳定,其中相关系数（r）均＞0.7, 平均绝对误差（MAE）和均方根误差（RMSE）数值较小。RF在大气PM_2.5中金属浓度预测上具有较好的表现,可在缺乏监测和实验数据的情况下,实现对大气颗粒物中金属浓度的快速预测,为全面了解颗粒物中金属污染特征提供数据基础。     

基于轮廓似然函数的PM_(10)极值浓度分析

以R软件为分析工具,选择GEV(generalized extreme value distribution)模型拟合四川省泸州市2003～2007年期间PM10每月最高日平均浓度数据,采用极大似然法估计模型的3个参数即位置参数、尺度参数、形状参数,利用所得的参数估计值计算得出某一标准值(如GB3095—1996)的重现期;进一步利用参数估计值计算轮廓似然函数,估计某一段固定时间间隔的PM10浓度的重现值以及其置信区间。结果表明,GEV模型能很好地拟合泸州市PM10数据,利用轮廓似然函数估计的不同时间间隔的重现值准确度高,统计结果可以为环境主管部门发布污染状况预警信息提供参考。     

Inverse Modeling for Aquatic Source and Transport Parameters Identification and its Application to Fukushima Nuclear Accident

Inverse modeling technique based on nonlinear least square regression method (LSRM) is developed for the identification of aquatic source and transport parameters. Instantaneous line source release model in two-dimensional domain and continuous point source release model in three-dimensional domain are used for the purpose. Case studies have been carried out for both types of releases to illustrate their application. Error analysis has been carried out to identify the maximum error that can be tolerated in the input concentration data used in the inverse model and to specify the minimum number of sampling points to generate such input data. The LSRM is compared with the well-established correlation coefficient optimization method for instantaneous line source release model, and good comparison is observed between them. The LSRM is used to quantitatively estimate the releases of different radionuclides into the Pacific Ocean which has resulted due to the discharge of highly radioactive liquid effluent from the affected Daiichi Nuclear Power Station at Fukushima in Japan. The measured concentrations of these radionuclides in seawater samples collected from two sampling points near Fukushima are used for the estimation. The average release works out to be 1.09?×?10¹⁶ for ¹³¹I, 3.4?×?10¹⁵ Bq for ¹³⁴Cs, and 3.57?×?10¹⁵ Bq for ¹³⁷Cs. Very good agreement is observed between the releases estimated in this study and those estimated by other different agencies.     

基于特征选择和机器学习的阳澄湖叶绿素a遥感反演研究

利用2016-2020年Sentinel-2多光谱遥感影像和同步实测叶绿素a浓度数据,提出了一种基于特征选择和机器学习的叶绿素a遥感反演方法,并应用于阳澄湖.结果表明,特征选择方法在反演模型的自变量选取上具有较好的应用效果,基于此建立的随机森林模型在阳澄湖叶绿素a反演上具有较优的验证精度;2016-2020年阳澄湖叶绿...     

Extreme Nitrogen Oxide and Ozone Concentrations in Athens Atmosphere in Relation to Meteorological Conditions

In the present study, we investigate the variation of NO _x (NO + NO₂) and O₃ concentrations and the relation between the extreme events (episodes) of NO _x and O₃ concentrations and the relevant meteorological conditions in the urban atmosphere of the Athens basin. Hourly data of NO, NO₂ and O₃ concentrations from 10 representative monitoring sites located in the Athens basin were used, covering the 10-year time period from 1994 to 2003. The results of our analysis show that the concentrations of air pollutants differ significantly from one monitoring site to another, due to the location and proximity of each station to the emission sources. For each site, there are also significant differences in NO _x and O₃ concentrations from day to day, as well as from month to month and/or from season to season. The annual and seasonal variations show higher NO values in winter and lower in summer. On the contrary, NO₂ and O₃ values are higher in summer (photochemical production of O₃) and lower in winter. These differences are attributed, to a large extent, to the prevailing synoptic and meteorological conditions, the most important between them being the wind direction and speed as well as the atmospheric pressure. Our analysis of the identified 179 extreme NO _x air pollution events shows that most of them took place under anticyclonic conditions, associated with calm or weak winds (speed <2.5 ms⁻¹) of mostly southern to southwestern directions, as well as with low air temperatures and intense stable surface atmospheric conditions. There exists a significant decreasing tendency in NO _x air pollution episodic events over the 10-year study period, resulting in very few to none events in the period from 2000 to 2003. As far as it concerns the extreme O₃ concentrations, 34 air pollution events were identified, occurring under high air temperatures, variable weak winds and intense solar irradiation. The trends of O₃ concentrations are stronger in suburban sites than in urban ones.     

Stationary and Non-stationary Extreme Value Approaches for Modelling Extreme Temperature: the Case of Riyadh City,Saudi Arabia

Climate change is one of the most fiercely debated scientific issues in recent decades, and the changes in climate extremes are estimated to have greater negative impacts on human society and the natural environment than the changes in mean climate. Extreme value theory is a well-known tool that attempts to best estimate the probability of adversarial risk events. In this paper, the focus is on the statistical behaviour of extreme maximum values of temperature. Under the framework of this theory, the methods of block maxima and threshold exceedances are employed. Due to the non-stationary characteristic of the series of temperature values, the generalized extreme value distribution and the generalized Pareto distribution were extended to the non-stationary processes by including covariates in the parameters of the models. For the purpose of obtaining an approximately independent threshold excesses, a declustering method was performed and then the de-clustered peaks were fitted to the generalized Pareto distribution. The stationary Gumbel distribution was found a reasonable model for the annual block maxima; however, a non-stationary generalized extreme value distribution with quadratic trend in the location was recommended for the half-yearly period. The findings also show that there is an improvement in modelling daily maxima temperature when it is applied to the declustered series and the given model outperforms the non-stationary generalized Pareto distribution models. Furthermore, the retained generalized Pareto distribution model proved better than the generalized extreme value distribution. Estimates of the return levels obtained from both extreme value models show that new records on maximum temperature event could appear within the next 20, 50 and 100 years.     

The Application of Water Quality Monitoring Data in a Reservoir Watershed Using AMOS Confirmatory Factor Analyses

This study investigates six water quality monitoring stations in the watershed of the Feitsui Reservoir. It uses nine parameters of water quality collected in an interval of two and half years for factor analyses, which was first conducted to determine four types of factors, respectively, those for organic pollution, eutrophication, seasonal influence, and sediment pollution. The analysis results effectively help to determine water quality in the watershed of the reservoir. The authors reutilize analysis of moment structures (AMOS) to acquire further results in order to confirm the goodness of fit of the previous factor analysis model. During the confirmation, we examine the hypothesized orthogonal results as well as utilize oblique rotation to explore the goodness of fit of the reflective indicators of the orthogonal rotation. As shown in the algorithm results, as long as the covariance curve is included in the four factors, no related issues are detected in the goodness of fit of reflective indicators and interior and external quality is reported with excellence. The orthogonal model, thus, stands. Additionally, when the analysis of structural equation modeling (SEM) is conducted, sample data mismatches the hypotheses of multivariate normality. Therefore, this study adopts the generalized least square (GLS) for an algorithm. Research results of this study have been submitted to the reservoir management authorities in Taiwan for the improvement of statistical application and strategic evaluation of water quality monitoring data in order to strengthen the managerial effectiveness of water quality in watersheds.     

深度学习在裸地扬尘源监测中的应用研究

裸地是扬尘的重要来源,施工建设过程中形成的裸地极易在大风天气作用下造成扬尘污染。因此,快速、有效地定位裸地位置,并确认其管控措施落实情况,对于开展裸地扬尘源监管具有重要意义。基于高分辨率遥感监测数据,结合人工解译裸地扬尘源数据集,以北京市大兴区为例,利用深度学习方法对裸地和防尘网覆盖裸地进行分类识别。同时,利用颜色匹配法对大兴区防尘网覆盖裸地进行识别,横向评估深度学习方法的识别精度。结果显示：深度学习方法对防尘网覆盖裸地的识别精度达97%,对裸地的识别精度达61%;颜色匹配法对防尘网覆盖裸地的识别精度达85%。防尘网覆盖裸地的颜色特征鲜明,深度学习方法和颜色匹配法对防尘网覆盖裸地的识别精度都在85%以上。深度学习方法对于面积大于2 000 m²的图斑有着较好的识别精度。深度学习方法可以提高裸地遥感解译的效率,实现规范化图像识别,可以作为人工判读的辅助手段。在实际应用中,可通过进一步积累样本来增强模型性能。深度学习方法适用于裸地扬尘源线索快速发现、工地防尘网措施落实情况快速检测等场景。     

Bivariate Extreme Value Analysis of Rainfall and Temperature in Nigeria

Environmental Modeling & Assessment - The rising cases of floods and the onset of drought in different parts of Nigeria require urgent attention particularly because Nigeria accommodates the...