基于深度学习的大数据空气污染预报

为了更好地反映环境污染变化趋势,为环境管理决策提供及时、全面的环境质量信息,预防严重污染事件发生,开展城市空气质量预报研究是十分必要的.本文针对环境大数据时代下的城市空气质量预报,提出了一种基于深度学习的新方法.该方法通过模拟人类大脑的神经连接结构,将数据在原空间的特征表示转换到具有语义特征的新特征空间,自动地学习得到层次化的特征表示,从而提高预报性能.得益于这种方式,新方法与传统方法相比,不仅可以利用空气质量监测、气象监测及预报等环境大数据,充分考虑污染物的时空变化、空间分布,得到语义性的污染物变化规律,还可以基于其他空气污染预测方法的结果(如数值预报模式),自动分析其适用范围、优势劣势.因此,新方法通过模拟人脑思考过程实现更充分的大数据集成,一定程度上克服了现有方法的缺陷,应用上更加具有灵活性和可操作性.最后,通过实验证明新方法可以提高空气污染预报性能.     

Assessment of deep aquifer groundwater quality in a geographically unique climatic region of Southern Western Ghats,India using water quality indices

Water quality index (WQI) models are generally used in hydrochemical studies to simplify complex data into single values to reflect the overall quality. In this study, deep groundwater quality in the Chittur and Palakkad Taluks of the Bharathapuzha river basin of Kerala, India, was assessed by employing the WQI method developed by the Canadian Council of Ministers of the Environment (CCME). The assessment of overall water quality is indispensable due to the specific characteristics of the study area, such as geography, climate, over-drafting, and prevalent agricultural practices. Forty representative samples were collected from the study area for monsoon (MON) and pre-monsoon (PRM) seasons. The results showed a general increase of contents from MON to PRM. The major cations were spread in the order Ca²⁺>Na+>Mg²⁺>K⁺ and the anions HCO₃⁻>Cl⁻>CO₃²⁻ based on their relative abundance. Among various parameters analysed, alkalinity and bicarbonate levels during MON were comparatively high, which is indicative of carbonate weathering, and 90% of the samples failed to meet the World Health Organization (WHO, 2017)/Bureau of Indian Standards (BIS, 2012) drinking water guidelines. The CCME WQI analysis revealed that nearly 50% of the samples during each season represented good and excellent categories. The samples in the poor category comprised 10% in MON and 15% in PRM. The overall WQI exhibited 15% of poor category samples as well. The spatial depiction of CCME WQI classes helped to expose zones of degraded quality in the centre to eastward parts. The spatial and temporal variations of CCME WQI classes and different physicochemical attributes indicated the influence of common factors attributing to the deep groundwater quality. The study also revealed inland salinity at Kolluparamba and Peruvamba stations, where agricultural activities were rampant with poor surface water irrigation.     

Near-term forecasts of stream temperature using deep learning and data assimilation in support of management decisions

Deep learning (DL) models are increasingly used to make accurate hindcasts of management-relevant variables, but they are less commonly used in forecasting applications. Data assimilation (DA) can be used for forecasts to leverage real-time observations, where the difference between model predictions and observations today is used to adjust the model to make better predictions tomorrow. In this use case, we developed a process-guided DL and DA approach to make 7-day probabilistic forecasts of daily maximum water temperature in the Delaware River Basin in support of water management decisions. Our modeling system produced forecasts of daily maximum water temperature with an average root mean squared error (RMSE) from 1.1 to 1.4°C for 1-day-ahead and 1.4 to 1.9°C for 7-day-ahead forecasts across all sites. The DA algorithm marginally improved forecast performance when compared with forecasts produced using the process-guided DL model alone (0%–14% lower RMSE with the DA algorithm). Across all sites and lead times, 65%–82% of observations were within 90% forecast confidence intervals, which allowed managers to anticipate probability of exceedances of ecologically relevant thresholds and aid in decisions about releasing reservoir water downstream. The flexibility of DL models shows promise for forecasting other important environmental variables and aid in decision-making.     

基于Landsat数据的关中盆地腹地AOD时空格局及城市化对其影响

气溶胶光学厚度（aerosol optical depth,AOD）是气溶胶最基本的光学特性之一,表征气溶胶对光的消减作用,对区域大气污染有一定的指示作用.首先,以2000~2019年长时间序列MOD09A1产品数据为基础,利用ASTER波谱数据库,构建了Landsat TM/OLI蓝光波段地表反射率数据集,利用深蓝算法（deep blue algorithm,DB）对关中盆地腹地AOD进行了反演估算.其次,通过测算2000~2019多年平均/各年平均AOD和季节AOD值,研究分析了AOD空间分布特征和时间变化趋势;利用经验正交函数分析方法（empirical orthogonal function,EOF）剖析了AOD主要的空间分布模式.最后,探索分析了城市化对AOD时空分布特征的影响.结果表明：①研究区多年平均AOD空间分布受地形和人类活动影响显著,低海拔且人类活动强度大的盆地内部AOD值和变化幅度均高于周围山系.AOD呈现出明显的季节差异性：春季（0.34） > 夏季（0.33） > 秋季（0.23） > 冬季（0.12）,夏季AOD区域差异性最大.②年均AOD呈先增大后减小再增大的变化趋势,2005年达到最大值;春夏两季AOD数值分布离散,而冬季则表现为低值集聚的状态.③EOF分析结果表明,研究区AOD存在3种主要的空间分布模态：第一,AOD空间分布模式表现为区域一致型,特征向量空间分布与海拔梯度一致,反映了不同海拔下AOD变化程度的差异;第二,以秦岭山脉为界,大致呈现南北反相的分布特征,反映出关中盆地地域发展的独特性和与南部秦岭山区的差异性;第三,"东南-西北"的分布格局,表明了AOD在"城镇-乡村"呈反相变化趋势;④研究区AOD值与常住人口密度（R²=0.707,P<0.05）、不透水面密度（R²=0.377,P<0.05）以及工业POI密度（R²=0.727,P<0.5）呈正相关关系,说明城市化对AOD时空分布有一定影响.本研究对加强关中盆地空气质量监测和城市生态环境建设具有重要的意义.     

泊松比对巷道稳定性影响的数值模拟试验研究

泊松比是分析岩体性质的重要物理力学参数之一。为了分析泊松比对巷道稳定性的影响,从泊松比测定与计算方法着手,采用金尼克假说,以巷道侧压系数为研究对象,运用ANSYS软件,建立典型矿山巷道锚喷支护后的数值模型,通过设置沿拱角径向的安全系数监测点,建立巷道安全稳定范围与泊松比的对应关系。研究表明,随着泊松比μ值的增加,巷道拱角处不稳定区域逐渐减小,并向顶底板延伸,巷道两帮的塑性破坏区无明显变化;泊松比μ值为0.20时,塑性破坏区面积较大,为巷道断面面积的5.10倍;泊松比μ值(x)与安全系数≥1的区域距离拱角径向位置(y)的关系式为:y=61.518x2-54.509x+12.516。研究为泊松比对巷道稳定性影响提供了一定的理论参考依据。     

高瓦斯易自燃工作面高抽巷瓦斯抽采与采空区遗煤自燃相互影响研究

为了实现瓦斯与煤自燃两大灾害的联合防治,首先对布置高抽巷条件下瓦斯与遗煤自燃多因素相互影响关系进行了理论分析和归纳总结。结合淮南潘二煤矿11223高瓦斯易自燃工作面,建立了带有高抽巷的物理模型,利用UDF编译了本煤层与邻近层瓦斯涌出源项、采空区三维孔隙率和低温条件下煤氧化反应氧气消耗速率。在此基础上,分析了高抽巷布置参数和抽采参数以及工作面风量对高抽巷瓦斯抽采效果和采空区自燃带分布相互影响的规律。结果表明,当工作面风量为2 000 m3/min,高抽巷布置在顶板上方40 m时,高抽巷瓦斯抽采浓度和纯量分别达32.3%和29.07 m3/min,占总瓦斯涌出量的69.71%,同时能满足实际防火的要求。研究结果可为类似条件下高抽巷最佳施工与抽采参数提供借鉴。     

深部低渗透强突出煤层群首采下保护层确定和效果考察

软岩保护层在深部低渗透强突出煤层群首采保护层选择中具有潜在的应用前景。在阐述下保护层开采覆岩移动与采场裂隙演化及瓦斯运移关系的基础上,以芦岭矿为工程背景,通过FLAC3D数值试验,研究软岩保护层开采后的卸压效果,并和不同层位的薄煤层开采卸压效果对比分析。结果表明,在采厚相同的情况下,开采10煤保护层相对于软岩保护层,卸压程度弱,保护范围小,保护效果差。软岩保护层开采后,被保护层处于弯曲下沉带的下限范围。受采动影响,有大量的离层裂隙生成。考察期范围内上覆被保护层(8、9煤层)瓦斯抽采率达62.9%,表明软岩保护层开采能够对上覆被保护层(8、9煤层)起到显著的卸压效果。研究结果可为其他矿区保护层开采选择提供参考,丰富国内保护层开采实践。     

软岩大变形巷道支护技术及应用

针对三门峡铝土矿软岩巷道大变形难支护的问题,综合现场工程地质调查、软岩水理作用测试、现场测试等手段和方法,对该地区巷道破坏模式和失稳机理进行详细分析,提出了锚喷-砌碹相结合的互补支护技术。按照互补支护参数建立数值模型并进行稳定性分析,数值计算结果表明在此支护方式下巷道各部位围岩收敛变形控制良好,且支护后围岩安全系数较之前明显提高。将新支护方案进行工程试验,巷道变形监测数据表明,支护后的巷道变形得到了良好的控制,混凝土喷层未出现裂缝,围岩变形速度迅速减小;锚喷10天后围岩变形基本稳定,此时进行二次砌碹支护效果最好,可以保证巷道的长期稳定。     

基于理想点法的邻近基坑高层建筑安全性评价

为更加准确地评价紧邻基坑高层建筑物的安全状况,综合考虑了建筑物稳定性、基坑与建筑之间土体扰动程度和建筑物地基抗扰动能力三方面因素,采用相关系数法选取14个评价指标,构建建筑物安全评价模型。并通过变异系数法和AHP确定各评价指标融合权重,最后采用理想点法对某紧邻基坑的高层建筑物进行安全评价。结果表明,该模型可以有效判断邻近深基坑高层建筑物的安全等级,其评判结果与工程实际相符     

地铁车站深基坑围护桩优化及其数值分析

以北京地铁大屯路公交与地铁换乘车站深基坑工程为研究背景,在对基坑围护结构变形分析的基础上,以围护桩为例,通过理论计算对其配筋进行了优化设计,并使用FLAC程序建立了三维数值模型对优化结果进行了检验分析。结果表明,优化方案和原方案相比不会导致基坑变形增大,还可以节约大量材料（桩间距由1200变为1500,桩径由1000变为800）。优化方案可以保证施工安全,同时节省材料和减小成本,对基坑支护设计优化具有一定参考价值。