基于机器学习的滑坡崩塌地质灾害气象风险预警研究

在划分气象风险等级时,传统地质灾害气象风险预警方法忽略了承灾体脆弱性因素,且气象风险预报等级整体偏高,导致高等级风险区空报率较高。基于此,提出基于机器学习的滑坡、崩塌灾害气象风险预警方法。利用信息量法,分析气象因素影响程度。选取坐标点、降雨量、易发生等级,将其作为机器学习人工神经网络的输入节点,判断是否发生崩塌、滑坡灾害;针对地质灾害区域,根据影响程度计算气象引发因子指数,结合滑坡、崩塌灾害潜势度G和承灾体脆弱性M,确定气象风险预警指数R,划分预警级别,完成滑坡、崩塌灾害气象风险预警。实验结果表明,设计方法有效降低了三级预报和四级预警空报率,提升了预警精细化程度。     

地下水机器学习方法研究

随着我国地下水监测工作的高速发展,高频率高密度水位监测数据的出现催生了对其进行深入信息挖掘的需求。在传统地下水模型研究中,地下水水位监测值常位于模型构建过程的下游,当水位监测的时空密度逐渐增大时,新增信息无法有效传导至模型的规划阶段并指导概念模型的修订。文章提出了一种地下水系统补排边界的识别方法,在不建立地下水数值模型的前提下,以监测井空间位置为节点,按照德劳内原则建立三角网格。在此网格系统中,首先定义一个水力梯度变换函数gradF,以求取网格中任意位置的水力梯度;借鉴机器学习领域的优化算法,使用水力梯度场驱动含水层中随机分布质点的运行轨迹,并以此推断和识别区域内地下水补给和排泄边界。在环境地学计算平台EnviFusion-CGS中实现,并构建了详细工作流程。以山东省青岛市大沽河中下游含水层为示范区,对含水系统的补给区和排泄区的空间分布及其动态变化进行了分析,取得了良好效果。本研究为构建和修订已有含水层概念模型提供了新思路。     

地下水污染风险预警等级及阈值确定方法研究综述

预警等级及阈值的科学确定对地下水污染风险预警系统的建立起着至关重要的作用。当前我国地下水污染风险预警相关研究尚处在探索阶段,未形成科学完备的理论与方法体系。文章基于国内外已有研究报道,较系统地梳理了地下水污染风险预警相关研究,归纳总结了预警等级及阈值确定方法的研究现状,并对进一步研究做了展望。目前阈值主要的划定方法有相关标准法、临界值法、综合评判法等。相关标准法适用于大尺度宏观区域,临界值法在基础资料较完善的研究区域具有较强的针对性和准确性,而综合评判法能够处理多个变量因素,能对定性因素做出量化评价,适用于变量、定性因素较多的研究区域,为不同条件研究区地下水污染风险预警等级划分及阈值划定方法的选择提供了理论依据。最后指出在完善我国地下水污染风险预警体系理论方法的工作中,应当进一步关注：（1）多学科研究成果的交叉渗透,借助多方法交互应用;（2）加强与地下水污染相关的水文地质调查与监测工作,不断提高地下水污染调查研究程度;（3）综合分析地质、污染源、地下水价值及动态等地下水污染影响因素,以提高预警结果科学性。     

基于实时地质灾害监测数据的预警预报动态阈值分析方法

目前在地质灾害监测领域,监测预警信息的发布主要基于各类监测设备的阈值设定。阈值是根据经验或专家估计设定,不仅对地质灾害不同类型、不同环境缺乏针对性,而且设定后较长时间不变,或根据经验略微浮动,缺少数据样本分析的科学性。另外监测设备容易受到卫星信号等环境因素影响,因此在实际运行中可能会出现误报、漏报的情况。为了解决上述问题,文中提出了一种预警阈值自学习自修正从而进行动态调整的方法,引入了两种可变阈值,并提出了一种基于优先级和门以及半马尔可夫过程 VTAS的性能指标优化新方法。半马尔可夫过程的应用使该方法能够考虑具有非高斯分布的工业测量。此外,文中还提出了一种基于遗传算法的优化设计过程,用于优化参数设置,提高性能指标。通过数值仿真以及与以往研究的比较,说明了该方法的有效性。将该方法在实测点位上进行应用,根据结果可知,相比于使用固定阈值,该方法能有效地减少系统误报、漏报,提高地质灾害预警的准确性,从而更好地保护人民生命财产安全。     

Big data and machine learning in geoscience and geoengineering: Introduction

In recent years,we have entered the so-called Fourth Paradigm with the regular production of huge amount of observational data.Big data is often characterized by the three‘V's:Volume of data,Variety and Velocity.The concept of big data can potentially address some existing issues in areas of geoscience and geoengineering.Large-scale,comprehensive,multidirectional and multifield geotechnical monitoring is becoming a reality in the very near future.     

基于WEBGIS与三库一体模式的土壤盐渍化监测预警系统平台研发

随着计算机应用操作技术不断进步,软件技术的不断完善以及可视化技术的蓬勃发展,使WEBGIS系统的可操作性能越来越好。本文以新疆渭库绿洲盐渍地为背景,提出了基于GIS、网络技术与安全监测相结合的WEBGIS与"三库一体模式"技术的平台。首先根据大量野外实测数据、文献资料、人文经济数据、基础地理地质等数据建立SQL Server数据库,之后以集成适用于盐渍地特点的21种监测模型的模型库为主导,调用盐渍化分析方法,提取数据进行相关计算,对模型求解并辅助专家进行决策。本系统采用B/S模式,HTML(超文本标记语言)设计并建立盐渍化监测预警模型,分析评价盐渍化程度、影响及其发展趋势,从而开发出盐渍化监测与预警智能分析相结合的系统工程应用网络。系统实现了数据库的储存、分析和管理;盐渍化监测模型分析;信息结果发布等功能。提高了地理空间信息的表现力,促进了干旱区信息化应用水平和信息共享平台扩展,为区域可持续发展提供了强有力的技术支撑。     

Comparing methods for determining landslide early warning thresholds: potential use of non-triggering rainfall for locations with scarce landslide data availability

Landslides - Rainfall intensity-duration landslide-triggering thresholds have become widespread for the development of landslide early warning systems. Thresholds can be in principle determined...     

Flood risk mapping for the lower Narmada basin in India: a machine learning and IoT-based framework

Natural Hazards - Floods have a significant economic, social, and environmental impact in developing countries like India. Settlements in flood hazard zones increase flood risk due to a lack of...     

Estimation and scaling of hydrostratigraphic units: application of unsupervised machine learning and multivariate statistical techniques to hydrogeophysical data

Numerical models provide a way to evaluate groundwater systems, but determining the hydrostratigraphic units (HSUs) used in constructing these models remains subjective, nonunique, and uncertain. A three-step machine-learning approach is proposed in which fusion, estimation, and clustering operations are performed on different data sets to arrive at HSUs at different scales. In step one, data fusion is performed by training a self-organizing map (SOM) with sparse borehole hydrogeologic (lithology, hydraulic conductivity, aqueous field parameters, dissolved constituents) and geophysical (gamma, spontaneous potential, and resistivity) measurements. Estimation is handled by iterative least-squares minimization of the SOM quantization and topographical errors. Application of the Davies-Bouldin criteria to k-means clustering of SOM nodes is used to determine the number and location of discontinuous borehole HSUs with low lateral density (based on borehole spacing at 100 s m) and high vertical density (based on cm-scale logging). In step two, a scaling network is trained using the estimated borehole HSUs, airborne electromagnetic measurements, and numerically inverted resistivity profiles. In step three, independent airborne electromagnetic measurements are applied to the scaling network, and the estimation performed to arrive at a set of continuous HSUs with high lateral density (based on sounding locations at meter (m) spacing) and medium vertical density (based on m-layer modeled structure). Performance metrics are used to evaluate each step of the approach. Efficacy of the proposed approach is demonstrated to map local-to-regional scale HSUs using hydrogeophysical data collected at a heterogeneous surficial aquifer in northwestern Nebraska, USA.     

A spatiotemporal Bayesian maximum entropy-based methodology for dealing with sparse data in revising groundwater quality monitoring networks: the Tehran region experience

Data inadequacy is a common problem in designing or updating groundwater monitoring systems. The developed methodologies for the optimal design of groundwater monitoring systems usually assume that there is a complete set of data obtained from existing monitoring wells and provide a revised configuration for the system by analyzing the current data. These methodologies are not usually applicable when the current groundwater quantity and quality data are highly sparse. In this paper, a new simulation–optimization approach based on Bayesian maximum entropy theory (BME) is proposed for revising spatial and temporal monitoring frequencies in a sparsely monitored aquifer. The BME is used to simulate the spatial and spatiotemporal variations of groundwater indicators, incorporating the space/time uncertainties due to insufficient data. Comparing the obtained estimations with observations, the best BME model was selected to be linked with an optimization model. The main goal of optimization was to find out the spatial and temporal sampling characteristics of the monitoring stations using the concepts of Entropy theory and a groundwater vulnerability index. The results show the BME estimations are less biased and more accurate than Ordinary Kriging in both spatial and spatiotemporal analysis. The improvements in the BME estimates are mostly related to incorporating hard (accurate) and soft (uncertain) data in the estimation process. The applicability and efficiency of the proposed methodology have been evaluated by applying it to the Tehran aquifer in Iran which is suffering from high groundwater table fluctuations and nitrate pollution. Based on the results, in addition to the existing monitoring wells, seven new monitoring stations have been proposed. Few stations which potentially can be removed or combined with other stations have been identified and a monthly sampling frequency has been suggested.     

滑坡对降雨响应的多指标监测及综合预警探析:以赣南罗坳滑坡为例

降雨量和位移是当前降雨型滑坡监测预警最常用的指标。然而,降雨量和位移监测结果只能反映降雨作用下滑坡的变形情况,不能揭示滑坡内在物理力学性状对降雨的响应。因此,除降雨量和位移监测之外,建立包括体积含水率、基质吸力等反映滑坡动态演化过程的关键指标监测体系必将成为今后更真实地把握滑坡内在演化趋势、更准确地建立滑坡综合预警判据的最有效手段。笔者对赣南地区典型降雨型滑坡进行了多指标监测及综合预警示范研究。结果表明:(1)在降雨条件下滑坡土体内部体积含水率、基质吸力和温度等多指标均产生有规律的动态响应;(2)随着降雨的持续,滑体体积含水率与基质吸力的变化均具有显著的滞后现象;(3)体积含水率和基质吸力变化速率与滑体位移具有显著的正相关性;(4)滑体温度分布变化规律受大气温度和体积含水率的共同影响。以实测数据的滑坡稳定性分析为基准,在考虑实际降雨入渗深度与滑坡稳定性的关联度上,建立了包括日降雨量、体积含水率增加速率、基质吸力减小速率以及位移速度多元指标预警方法体系,提出了基于关键指标综合预警体系及确定方法,旨在为降雨滑坡准确预警提供新模式。     

Hybrid machine learning model with random field and limited CPT data to quantify horizontal scale of fluctuation of soil spatial variability

Acta Geotechnica - The scale of fluctuation (SOF) is the critical parameter to describe the soil spatial variability, which significantly influences the embedded geostructures. Due to the limited...     

基于普适型仪器的滑坡监测预警初探——以甘肃兰州岷县三处滑坡为例

在甘肃省滑坡灾害较为严重的兰州市和岷县,选择了三处滑坡,建设了基于普适型仪器的专群结合监测预警系统。按照《地质灾害专群结合监测预警技术指南(试行)》的要求,选用裂缝计、GNSS、土体含水量仪、雨量计、声光报警器等,对地质灾害体变形破坏、相关因素、宏观前兆等指标开展专业化立体综合监测。监测设备采用蓄电池加太阳能的方案来供电,保证24小时不间断工作;通讯系统采用现场LoRa组网配合2/3/4 G移动通信的方案,能够保证数据传输的效率,同时降低通信成本;监测数据同时发送到国家和省级地质灾害监测数据平台,能够高效支撑地灾预警工作;监测平台能够对实时采集的监测数据自动进行分析,支持用多种预警模型进行判别;监测数据发生变化触发预设条件时,能够自动发送预警信息。通过近四个月的系统试运行,捕捉到了滑坡mm级的蠕变变形,数据可靠,能够满足监测预警的需求。通过三处滑坡监测预警工程的实施,一方面对滑坡变形和环境因素实现了实时监控,同时也为类似问题提供了一个可供参考的解决方案。     

Application of the monitoring and early warning system for internal solitary waves:Take the second natural gas hydrates production test in the South China Sea as an example

Internal solitary waves(ISWs) contain great energy and have the characteristics of emergency and concealment. To avoid their damage to offshore engineering, a new generation of monitoring and early warning system for ISWs was developed using technologies of double buoys monitoring, intelligent realtime data transmission, and automatic software identification. The system was applied to the second natural gas hydrates(NGHs) production test in the Shenhu Area, South China Sea(SCS) and successfully ...     

A numerical method for obtaining multiple linear regression parameters with physically realistic signs and magnitudes: Applications to the determination of equilibrium constants from solubility data

A common problem in experimental geochemistry is the derivation of equilibrium constants from solubility experiments. A simple method of deriving these equilibrium constants, multiple linear regression, often results in the appearance of negative values. This has been a significant obstacle to continuing research in this field. The problem occurs for the most part because of significant correlations among the “independent” concentration variables. These correlations are an inescapable result of the nature of the experiments and the physical model being fitted. Ridge regression is an appropriate modification to simple linear regression which overcomes this difficulty. Ridge regression results in a simple procedure to obtain physically plausible, yet statistically rigorous stability constants. Of course, other problems may further degrade the quality of derived equilibrium constants, e.g. uncertainty in activity coefficients and no purely statistical method can overcome these types of problems. However, ridge regression is an effective procedure to overcome the multicolinearity which is the main cause of negative equilibrium constants. We demonstrate the use of ridge regression with a general mathematical model and then illustrate its use in the determination of iron-chloro complex equilibrium constants from solubility studies of pyrite-pyrrhotite-magnetite in NaCl solutions at 250°C. Ridge regression may also be of use in other geochemical problems where one must estimate parameters with a physical interpretation and where the independent variables are significantly intercorrelated.     

Comparison of GA-BP and PSO-BP neural network models with initial BP model for rainfall-induced landslides risk assessment in regional scale: a case study in Sichuan,China

Communities everywhere are being subjected to a variety of natural hazard events that can result in significant disruption to critical functions. As a result, community resilience assessment in these locations is gaining popularity as a means to help better prepare for, respond to, and recover from potentially disruptive events. The objective of this study was to identify key vulnerabilities relevant to addressing rural community resilience through conducting an initial flood impact analysis, with a specific focus on emergency response and transportation network accessibility. It included a use case involving the flooding of a rural community along the US inland waterway system. Special consideration was given to impacts experienced by at-risk populations (e.g., low economic status, youth, and elderly), given their unique vulnerabilities. An important backdrop to this work is recognition that Federal Emergency Management Agency’s Hazus, a free, publicly available tool, is commonly recommended by the agency for counties, particularly those with limited resources (i.e., rural areas), to use in developing their hazard mitigation plans. The case study results, however, demonstrate that Hazus, as currently utilized, has some serious deficiencies in that it: (1) likely underestimates the flood extent boundaries for study regions in a Level 1 analysis (which solely relies upon filling digital elevation models with precipitation), (2) may be incorrectly predicting the number and location of damaged buildings due to its reliance on out-of-date census data and the assumption that buildings are evenly distributed within a census block, and (3) is incomplete in its reporting of the accessibility of socially vulnerable populations and response capabilities of essential facilities. Therefore, if counties base their flood emergency response plans solely on Hazus results, they are likely to be underprepared for future flood events of significant magnitude. An approach in which Hazus results can be augmented with additional data and analyses is proposed to provide a more risk-informed assessment of community-level flood resilience.