WO2023024463A1 - 一种水体有机污染智能化溯源方法及系统 - Google Patents

一种水体有机污染智能化溯源方法及系统 Download PDF

Info

Publication number
WO2023024463A1
WO2023024463A1 PCT/CN2022/077587 CN2022077587W WO2023024463A1 WO 2023024463 A1 WO2023024463 A1 WO 2023024463A1 CN 2022077587 W CN2022077587 W CN 2022077587W WO 2023024463 A1 WO2023024463 A1 WO 2023024463A1
Authority
WO
WIPO (PCT)
Prior art keywords
pollutants
pollution
water body
organic
pollutant
Prior art date
Application number
PCT/CN2022/077587
Other languages
English (en)
French (fr)
Inventor
吴吉春
刘梦雯
韦斯
祝晓彬
于南洋
Original Assignee
南京大学
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by 南京大学 filed Critical 南京大学
Priority to US18/005,565 priority Critical patent/US11965871B2/en
Publication of WO2023024463A1 publication Critical patent/WO2023024463A1/zh

Links

Images

Classifications

    • GPHYSICS
    • G01MEASURING; TESTING
    • G01NINVESTIGATING OR ANALYSING MATERIALS BY DETERMINING THEIR CHEMICAL OR PHYSICAL PROPERTIES
    • G01N33/00Investigating or analysing materials by specific methods not covered by groups G01N1/00 - G01N31/00
    • G01N33/18Water
    • G01N33/1826Organic contamination in water
    • GPHYSICS
    • G01MEASURING; TESTING
    • G01NINVESTIGATING OR ANALYSING MATERIALS BY DETERMINING THEIR CHEMICAL OR PHYSICAL PROPERTIES
    • G01N30/00Investigating or analysing materials by separation into components using adsorption, absorption or similar phenomena or using ion-exchange, e.g. chromatography or field flow fractionation
    • G01N30/02Column chromatography
    • G01N30/86Signal analysis
    • G01N30/8624Detection of slopes or peaks; baseline correction
    • G01N30/8631Peaks
    • G01N30/8637Peak shape
    • GPHYSICS
    • G01MEASURING; TESTING
    • G01NINVESTIGATING OR ANALYSING MATERIALS BY DETERMINING THEIR CHEMICAL OR PHYSICAL PROPERTIES
    • G01N33/00Investigating or analysing materials by specific methods not covered by groups G01N1/00 - G01N31/00
    • G01N33/18Water
    • GPHYSICS
    • G01MEASURING; TESTING
    • G01NINVESTIGATING OR ANALYSING MATERIALS BY DETERMINING THEIR CHEMICAL OR PHYSICAL PROPERTIES
    • G01N30/00Investigating or analysing materials by separation into components using adsorption, absorption or similar phenomena or using ion-exchange, e.g. chromatography or field flow fractionation
    • G01N30/02Column chromatography
    • G01N30/86Signal analysis
    • G01N30/8693Models, e.g. prediction of retention times, method development and validation

Definitions

  • the invention belongs to the technical field of environmental analytical chemistry, and in particular relates to an intelligent traceability method and system for organic pollution in water bodies.
  • the present invention provides an intelligent traceability method for organic pollution in water And the system can realize the intelligent traceability of pollution sources and key pollutants in the case of unknown pollution sources, and provide technical support for the investigation and control of organic pollution in the water environment.
  • the present invention provides an intelligent traceability method for organic pollution in water, including:
  • high-throughput screening of organic matter in the water sample is carried out to confirm the pollutants in the water body
  • the machine learning classification model is used to determine the key pollutants in the pollution source and quantify their pollution contribution.
  • identifying pollution sources through network analysis includes:
  • the location interval of potential pollution sources is determined from the point where the peak area of the pollutants increases sharply from upstream to downstream, and at the same time, the consistency of the peak area of pollutants in this group increases sharply at this point, combined with actual geographical information , to determine the source of pollution.
  • the calculation of the correlation of the peak area of the organic pollutants, and constructing a correlation-based pollutant network according to the correlation, and classifying the pollutant network into groups includes:
  • drawing the change curve of the peak area of the pollutants in each group and the change curve of the average peak area of each group of pollutants includes:
  • the average value of the peak area of pollutants at each point is calculated, and the change curve of the average peak area of each group of pollutants is obtained.
  • the machine learning classification model is a random forest model.
  • the machine learning classification model is used to determine the key pollutants in the pollution source, and the quantification of its pollution contribution includes:
  • For the determined pollution source determine the organic pollutants in the water sample that flows into the receiving water body; select the organic pollutants that exist in both the pollution source and the receiving water body, and use the peak area of each point in the receiving water body as input ;
  • An index representing the importance of variables is output, and key pollutants and their pollution contributions are determined according to the value of the importance index.
  • outputting indicators representing the importance of variables, and determining key pollutants and their pollution contributions according to the value of the importance indicators include:
  • variable whose average accuracy rate reduction value is greater than the set threshold is considered to be a potential pollution contribution factor, and the peak area is judged between the maximum value of the downstream sample closest to the pollution source and the maximum value of all upstream samples;
  • the peak area of the potential pollution contribution factor is larger in the downstream sample closest to the pollution source than in the upstream sample, that is, its relative abundance increases, it is considered to be a key pollutant in the pollution source, based on the average accuracy of these key pollutants
  • the magnitude of the reduction values quantifies their pollution contribution.
  • organic matter analysis and detection data obtained by high performance liquid chromatography-tandem mass spectrometry of several water samples from the upstream to the downstream of the polluted water body include:
  • the samples were analyzed and detected by high performance liquid chromatography-tandem mass spectrometry.
  • the present invention provides an intelligent traceability system for water body organic pollution, which uses any of the intelligent traceability methods for water body organic pollution provided by the present invention to trace the source of water body pollution, including:
  • the data acquisition unit is used to acquire the organic matter analysis and detection data of high performance liquid chromatography-tandem mass spectrometry of several water samples from the upstream to the downstream of the polluted water body;
  • the pollutant determination unit is used to perform high-throughput screening of organic matter in the water sample and confirm the pollutants in the water body according to the analysis and detection data;
  • a pollution source identification unit is used to identify pollution sources through network analysis based on the determined pollutants
  • the pollution source evaluation unit is used to determine the key pollutants in the pollution source and quantify their pollution contribution according to the identified pollution source and the organic pollutants in the receiving water of the pollution source, using a machine learning classification model.
  • the present invention has the following advantages:
  • the present invention uses the peak area of organic pollutants in the receiving water body, that is, the correlation relationship of relative abundance, to construct a correlation network diagram of pollutants, and can visualize and aggregate the pollutants with similar relative abundance change trends in the receiving water body. Efficiently and quickly find clusters with similar distribution trends in a large number of pollutants, and at the same time quickly focus on clusters with a large number of pollutants, and effectively track the geographic location of the pollution source according to the change trend of the average relative abundance of the clusters. Combining geographic information to realize the identification of pollution sources.
  • Classify the upstream and downstream of the receiving water samples relative to the location of the pollution source build a machine learning classification model, discover potential pollution contributing factors, and screen them according to their relative abundance in the upstream and downstream, which can be used in a large number of pollutants Find the key pollution contribution factors to the receiving water body, effectively identify the key pollutants on the basis of identifying the pollution source, and realize the quantification of the contribution to the pollutant pollution based on the value of the importance index.
  • the whole process comprehensively utilizes high-throughput screening, network analysis and machine learning technologies, so that the intelligent traceability of pollution sources and key pollutants can be realized in the case of unknown pollution sources, which provides a basis for the investigation and control of organic pollution in the water environment.
  • Fig. 1 is the flow chart of the intelligent traceability method of water body organic pollution in the embodiment of the present invention
  • Fig. 3 is the change curve diagram of the pollutant peak area of 8 groups that network analysis obtains in the embodiment of the invention
  • Fig. 4 is a value diagram of the importance index of pollutants output by the machine learning model in the embodiment of the invention.
  • the present invention provides a method for intelligent traceability of organic pollution in water bodies.
  • FIG. 1 shows a flow chart of the intelligent traceability method for organic pollution in water bodies in an embodiment of the present invention.
  • the method in the embodiment include:
  • Step S100 Obtain the organic matter analysis and detection data of the high performance liquid chromatography-tandem mass spectrometry of several water samples from the upstream to the downstream of the polluted water body.
  • this step can be carried out according to sub-steps S110-S140:
  • Step S110 collecting water samples. River water samples were collected from 11 points from upstream to downstream.
  • Step S120 Extraction of organic matter in river water samples.
  • a 1 L river water sample was filtered with a 1 ⁇ m glass fiber filter membrane, and then the organic matter in the sample was enriched by solid phase extraction (SPE).
  • the column flow rate was maintained at approximately 3 mL/min.
  • Activation is required before using the cartridge: for MAX cartridges, add 10ml 2% formic acid methanol solution, 10ml methanol, 10ml Fisher water; for MCX cartridges, add 10ml 5% ammonia methanol solution, 10ml methanol, 10ml Fisher water; For HLB cartridges, add 10ml methanol and 10ml Fisher water in sequence.
  • Step S130 concentrating and constant volume. Blow the eluent to nearly dry with nitrogen, dilute to 1ml with methanol, then centrifuge to get the supernatant, and store it in a sample injection vial.
  • Step S140 on-board detection.
  • the organic matter was analyzed and detected by the combination of high performance liquid chromatography-combined quadrupole orbital ion trap mass spectrometer, and the instrument conditions were as follows:
  • Ion source Electrospray
  • Ion mode positive ion mode and negative ion mode
  • MS Level 1 (MS) full scan range: 80-1000m/z;
  • Spray voltage 3500V (positive ion mode); 2500V (negative ion mode);
  • Step S200 According to the analysis and detection data, perform high-throughput screening of organic matter in the water sample to confirm the pollutants in the water body.
  • the data files obtained in step S100 are imported into MS-DIAL software for peak extraction and alignment, and high-throughput screening of organic substances in water samples is performed using the public large-scale mass spectrometry database , Manually check the matching of the secondary spectrum to remove false positives, and confirm the pollutants in it according to the substance classification information provided by PubChem.
  • software such as PeakView and Compound Discover can also be used to carry out, and public large-scale mass spectrum databases such as MS-DIAL, NIST, MassBank, GNPS mass spectrum databases, etc.
  • the parameters are set as follows: peak response: ⁇ 30000; alignment retention time error: ⁇ 0.2min; alignment quality error: ⁇ 0.01Da; screening quality error: primary ⁇ 0.01Da, secondary ⁇ 0.002Da.
  • the standard for manually checking and removing false positives is as follows: if there is only one fragment ion information in the secondary spectrum in the database, remove substances without fragment ion matches; if there are two or more fragment ion information in the database, remove less than The substance matched by two fragment ions; if there is no fragment ion information in the secondary spectrum in the database, remove the substance.
  • the final screening identified 132 organic pollutants in the river.
  • Step S300 According to the determined pollutants, identify pollution sources through network analysis. Specifically, this step includes:
  • Step S310 Calculate the correlation of the peak areas of the organic pollutants, construct a correlation-based pollutant network according to the correlation, and classify the pollutant network into groups.
  • the Pearson correlation of the peak area of organic pollutants in the receiving water body water samples from upstream to downstream is calculated, and the correlation relationship with significance p ⁇ 0.05 and positive value is retained, which is used as the edge, Put the pollutants as nodes into the Gephi software, build a correlation-based pollutant network, and perform modular analysis to obtain the group classification results of the pollutant network.
  • the correlation-based pollutant network is shown in FIG. 2 , and the pollutant network is divided into 8 different groups, as shown in FIG. 3 .
  • Cytoscape software can also be used to construct a correlation-based pollutant network.
  • Step S320 According to the pollutant groups, draw the change curve of the peak area of the pollutants in each group and the change curve of the average peak area of the pollutants in each group; specifically, the pollutants in the water samples from upstream to downstream can be Peak area standardization, according to the classification of pollutant groups, draw the change curve of the peak area of pollutants in each group. On this basis, the average value of the peak area of pollutants at each point is calculated, and the change curve of the average peak area of each group of pollutants is obtained. In the example provided, this is shown in Figure 3.
  • Step S330 According to the change curve of the average peak area of the pollutants in the large group from the point where the peak area of the pollutant increases sharply from the upstream to the downstream, determine the location interval of the potential pollution source, and at the same time consider the consistency of the peak area of the pollutant in the group that increases sharply at this point, combined with Actual geographic information to identify sources of pollution.
  • steps S100-S300 should be repeated until a specific point source of pollution is identified.
  • the black curve is the change curve of the average peak area of pollutants.
  • the gray curve is the change curve of the peak area of the pollutant.
  • the pollution source can be determined by combining with the actual geographical information. In the example provided, the location intervals of the two potential pollution sources are determined according to the point where the curve increases sharply from upstream to downstream in the first three large groups, and the peak areas of the three groups of pollutants increase sharply at these two points. Consistency, combined with the actual geographical information, to determine the nearby sewage treatment plant and a tributary is the source of its pollution.
  • Step S400 According to the identified pollution source and the organic pollutants in the receiving water body of the pollution source, use the machine learning classification model to determine the key pollutants in the pollution source and quantify their pollution contribution.
  • the machine learning classification model adopted is a random forest model
  • other classification models may also be used, such as decision tree, support vector machine and so on.
  • the random forest model is used for specific description, specifically, steps S410-S440 may be followed.
  • Step S410 For the determined pollution source, measure the organic pollutants in the water sample that flows into the receiving water body; select the organic pollutants that exist in both the pollution source and the receiving water body, and calculate their peaks at each point in the receiving water body area as input.
  • the detection can be carried out according to the method of steps S110-S140, and then the organic pollutants can be screened and identified according to the method of step S200.
  • the final screening identified 76 organic pollutants in a wastewater sample from a wastewater treatment plant. Select 71 kinds of organic pollutants that exist in both wastewater and rivers, and use their peak areas at various points in the river as input.
  • Step S420 Construct a random forest classification model, and use the upstream or downstream of the receiving water sample relative to the pollution source as the standard for binary classification of samples.
  • a random forest classification model is constructed using R language, and the river samples are divided into 8 upstream and 3 downstream samples relative to pollution sources.
  • Step S430 Training a random forest classification model.
  • the optimal value of the model parameter mtry was determined to be 16 by cyclically calculating the mean of the out-of-bag error rate (OOB), and the confusion matrix evaluated the model classification error rate to 0%.
  • OOB out-of-bag error rate
  • Step S440 output the index representing the importance of the variable, and determine the key pollutants and their pollution contribution according to the value of the importance index.
  • the metric used is the reduction in average accuracy. For the 32 pollutants whose average accuracy reduction value is greater than 0 (set threshold), judge whether their peak area in the downstream river sample closest to the pollution source is greater than the maximum value among all upstream river samples, and finally find 25 pollutants
  • the pollutants are the key pollution contribution factors in the pollution point source of the sewage treatment plant, and their pollution contribution is quantified based on the reduction value of their average accuracy rate, as shown in Figure 4.
  • the present invention provides an intelligent traceability system for water body organic pollution.
  • the system can trace the source of water body pollution according to the above-mentioned water body organic pollution traceability method.
  • the system includes:
  • the data acquisition unit is used to acquire the organic matter analysis and detection data of high performance liquid chromatography-tandem mass spectrometry of several water samples from the upstream to the downstream of the polluted water body;
  • the pollutant determination unit is used to perform high-throughput screening of organic matter in the water sample and confirm the pollutants in the water body according to the analysis and detection data;
  • a pollution source identification unit is used to identify pollution sources through network analysis based on the determined pollutants
  • the pollution source evaluation unit is used to determine the key pollutants in the pollution source and quantify their pollution contribution according to the identified pollution source and the organic pollutants in the receiving water of the pollution source, using a machine learning classification model.

Landscapes

  • Health & Medical Sciences (AREA)
  • Life Sciences & Earth Sciences (AREA)
  • Chemical & Material Sciences (AREA)
  • General Health & Medical Sciences (AREA)
  • Physics & Mathematics (AREA)
  • Analytical Chemistry (AREA)
  • Biochemistry (AREA)
  • General Physics & Mathematics (AREA)
  • Immunology (AREA)
  • Pathology (AREA)
  • Food Science & Technology (AREA)
  • Medicinal Chemistry (AREA)
  • Engineering & Computer Science (AREA)
  • Other Investigation Or Analysis Of Materials By Electrical Means (AREA)

Abstract

一种水体有机污染智能化溯源方法及系统,属于环境分析化学技术领域。包括获取受污染水体自上游至下游的若干水样的高效液相色谱-串联质谱的有机物分析检测数据;根据所述分析检测数据,对水样中的有机物进行高通量筛查,确认水体中的污染物;根据所确定的污染物,通过网络分析识别污染源;根据所识别的污染源以及所述污染源的受纳水体中的有机污染物,利用机器学习分类模型,确定污染源中的关键污染物,量化其污染贡献。所述方法能够在未知污染源的情况下,实现对污染源及其中关键污染物的智能化追溯,为水环境中有机污染的调查和管控提供技术支撑。

Description

一种水体有机污染智能化溯源方法及系统 技术领域
本发明属于环境分析化学技术领域,具体涉及一种水体有机污染智能化溯源方法及系统。
背景技术
随着社会经济发展,各类有机污染物层出不穷,包括但不限于农药及其转化产物、药物及其转化产物、表面活性剂、增塑剂和阻燃剂等有毒有害在产在用化学物质。这些有机污染物不断进入水环境中,其数量庞大、种类繁多,并造成潜在的生态和健康风险,因此需要对水环境中的有机污染物加以关注。
有机污染物进入水环境中的途径多种多样,已有研究表明,通过污水厂出水、工业废水排放、地表径流和地下径流等方式污染物流入地表水中,导致来源广泛且复杂的水环境有机污染问题,因此水环境的污染溯源工作是一项巨大的挑战。根据文献调研,目前对水环境中的有机污染溯源工作,一方面基于在水体和污染源间三维荧光信号的比较来追溯污染排放源,另一方面基于污染源本身独特的质谱信号所指示的特征污染物来开发源指纹追踪技术,除此之外,水生微生物群落对污染排放源引起的水质变化的响应也被用于追溯污染源。值得注意的是,目前这些溯源方法仍在明显的不足:首先,需要预先设定可能的污染来源,根据污染源本身的特征来确定水体污染的来源;其次,对污染来源中的关键污染因子及其贡献程度的识别工作仍较少。
鉴于现有方法的缺陷,需要开发一种在未知任何污染排放源的条件下,对水环境中大量有机污染物进行来源追溯的方法。
发明内容
技术问题:针对目前水环境中有机污染物数量和种类繁多,来源广泛且复杂,导致溯源困难,且目前溯源技术依赖于预先知道污染来源的问题,本发明提供一种水体有机污染智能化溯源方法及系统,能够在未知污染源的情况下,实现对污染源及其中关键污染物的智能化追溯,为水环境中有机污染的调查和管控提供技术支撑。
技术方案:第一方面,本发明提供一种水体有机污染智能化溯源方法,包括:
获取受污染水体自上游至下游的若干水样的高效液相色谱-串联质谱的有机物分析检测数据;
根据所述分析检测数据,对水样中的有机物进行高通量筛查,确认水体中的污染物;
根据所确定的污染物,通过网络分析识别污染源;
根据所识别的污染源以及所述污染源的受纳水体中的有机污染物,利用机器学习分类模型,确定污染源中的关键污染物,量化其污染贡献。
进一步地,根据所述分析检测数据,对水样中的有机物进行高通量筛查,确认水体中的污染物包括:
将分析检测得到的数据文件导入分析软件,例如MS-DIAL、PeakView、Compound Discover等,进行峰提取和对齐,使用公开的大型质谱数据库对水样中的有机物进行高通量筛查,手动检查二级谱图的匹配情况去除假阳性,根据PubChem提供的物质分类信息确认其中的污染物。
进一步地,所述根据所确定的污染物,通过网络分析识别污染源包括:
计算有机污染物的峰面积的相关性,并根据所述相关性构建基于相关性的污染物网络,对污染物网络进行类群划分;
根据污染物类群,绘制各类群中污染物的峰面积的变化曲线以及各类群污染物的平均峰面积的变化曲线;
根据大型类群中污染物的平均峰面积的变化曲线从上游至下游剧烈增加的点位确定潜在污染源位置区间,同时考虑该类群污染物峰面积在此点位剧烈增加的一致性,结合实际地理信息,确定污染源。
进一步地,所述计算有机污染物的峰面积的相关性,并根据所述相关性构建基于相关性的污染物网络,对污染物网络进行类群划分包括:
计算从上游至下游的受纳水体水样中有机污染物峰面积的相关性,保留显著性p<0.05且为正值的相关关系,将其作为边、将污染物作为节点输入网络分析软件,例如Gephi、Cytoscape,构建基于相关性的污染物网络,进行模块化分析,得到污染物网络的类群划分结果。
进一步地,所述根据污染物类群划分,绘制各类群中污染物的峰面积的变化曲线以及各类群污染物的平均峰面积的变化曲线包括:
将上游至下游的水样中污染物的峰面积标准化,根据污染物类群划分,绘制各类群中污染物的峰面积的变化曲线图;
根据污染物类群划分,计算各点位污染物峰面积的平均值,获得各类群污染物的平均峰面积的变化曲线。
进一步地,所述机器学习分类模型为随机森林模型。
进一步地,根据所识别的污染源以及所述污染源处受纳水体中的有机污染物,利用机器学习分类模型,确定污染源中关键的污染物,量化其污染贡献包括:
针对确定的污染源,测定其汇入受纳水体的水样中的有机污染物;选择同时在污染源和受纳水体中存在的有机污染物,将其在受纳水体各点位的峰面积作为输入;
构建随机森林分类模型,将受纳水体样本相对于污染源的上游或下游作为样本二分类的标准;
训练随机森林分类模型;
输出表征变量重要性的指标,并根据所述重要性指标的数值大小确定关键的污染物及其污染贡献。
进一步地,输出表征变量重要性的指标,并根据所述重要性指标的数值大小确定关键的污染物及其污染贡献包括:
对平均准确率的减少值大于设定阈值的变量,认为其是潜在的污染贡献因子,判断其峰面积在受纳水体距离污染源最近的下游样本和所有上游样本的最大值之间的大小关系;
若潜在污染贡献因子的峰面积在距离污染源最近的下游样本中比在上游样本中大,即其相对丰度发生增加,则认为是污染源中关键污染物,基于这些关键污染物的平均准确率的减少值的大小量化它们的污染贡献。
进一步地,获取受污染水体自上游至下游的若干水样的高效液相色谱-串联质谱的有机物分析检测数据包括:
采集接受污染的水体从上游至下游的水样;
对水样进行处理,尽可能全面地提取和富集其中的有机物,得到待测液;
使用高效液相色谱-串联质谱对样品进行有机物的分析检测。
第二方面,本发明提供一种水体有机污染智能化溯源系统,利用本发明提供的任一项所述的水体有机污染智能化溯源方法对水体污染进行溯源,包括:
数据获取单元,用于获取受污染水体自上游至下游的若干水样的高效液相色谱-串联质谱的有机物分析检测数据;
污染物确定单元,用于根据所述分析检测数据,对水样中的有机物进行高通量筛查,确认水体中的污染物;
污染源识别单元,用于根据所确定的污染物,通过网络分析识别污染源;
污染源评价单元,用于根据所识别的污染源以及所述污染源的受纳水体中的有机污染物,利用机器学习分类模型,确定污染源中的关键污染物,量化其污染贡献。
有益效果:本发明与现有技术相比,具有以下优点:
本发明利用有机污染物在受纳水体中峰面积,即相对丰度的相关关系,构建污染物的相关性网络图,可以将在受纳水体中相对丰度变化趋势相似的污染物进行可视化聚类,高效、快速找到大量污染物中具有相似分布趋势的类群,同时可以快速聚焦至污染物数量较多的类群,根据类群平均相对丰度的变化趋势,有效追踪到污染源所在的地理位置区间,结合地理信息,实现污染源的识别。以受纳水体水样相对于污染源位置的上游和下游进行样本分类,构建机器学习分类模型,发现潜在的污染贡献因子,根据其在上游和下游的相对丰度进行筛选,可以在大量污染物中找到对受纳水体的关键污染贡献因子,在识别污染源的基础上有效识别其中的关键污染物,同时依据重要性指标数值实现对污染物污染贡献的量化。
整个过程综合利用了高通量筛查、网络分析以及机器学习技术,从而能够在未知污染源的情况下,实现对污染源及其中关键污染物的智能化追溯,为水环境中有机污染的调查和管控提供技术支撑。
附图说明
图1为本发明的实施例中水体有机污染智能化溯源方法的流程图;
图2本发明实施例中网络分析得到的基于相关性的污染物网络图;
图3为发明实施例中网络分析得到的8个类群的污染物峰面积的变化曲线图;
图4为发明实施例中机器学习模型输出的污染物的重要性指标数值图。
具体实施方式
下面结合实施例和说明书附图对本发明作进一步的说明。
第一方面,本发明提供一种水体有机污染智能化溯源方法,图1示出了本发明的实施例中水体有机污染智能化溯源方法的流程图,结合图1所示,实施例中该方法包括:
步骤S100:获取受污染水体自上游至下游的若干水样的高效液相色谱-串联质谱的有机物分析检测数据。在本发明的实施例中,结合具体实例,该步骤可按照分步骤S110~S140进行:
步骤S110:水样的收集。采集从上游至下游共11个点位的河流水样。
步骤S120:河流水样中有机物的提取。使用1μm的玻璃纤维滤膜对1L河流水样进行过滤,随后采用固相萃取(SPE)法富集样品中的有机物,水样依次流经Oasis MAX、 MCX、HLB固相萃取小柱,控制过柱流速保持约3mL/min。使用小柱前需进行活化:对于MAX小柱,依次加入10ml 2%甲酸甲醇溶液、10ml甲醇、10ml Fisher水;对于MCX小柱,依次加入10ml 5%氨水甲醇溶液、10ml甲醇、10ml Fisher水;对于HLB小柱,依次加入10ml甲醇、10ml Fisher水。过完水样后,先使用离心机对萃取小柱进行离心脱水(3000rpm,5min),再进行洗脱:对MAX柱,加入12ml 2%甲酸甲醇溶液;对MCX柱,加入12ml 5%氨水甲醇溶液;对HLB柱,加入12ml甲醇。合并同一个水样的洗脱液得到最终的萃取洗脱液。
步骤S130:浓缩与定容。将洗脱液氮吹至近干,用甲醇定容至1ml,随后离心取上清液,保存至进样小瓶。
步骤S140:上机检测。通过高效液相色谱-组合型四极杆轨道离子阱质谱仪联用对样品进行有机物的分析检测,仪器条件如下:
高效液相色谱仪:Thermo UltiMate 3000;
色谱柱:Acquity UPLC BEH C18柱(2.1×150mm,1.7μm);
柱温:40℃;
流速:0.3ml/min;
流动相:(A相)2mM乙酸铵水溶液、(B相)甲醇;
洗脱梯度:
Figure PCTCN2022077587-appb-000001
质谱仪:Q Exactive Focus,Thermo Fisher;
离子源:电喷雾;
离子模式:正离子模式和负离子模式;
一级(MS)全扫描范围:80-1000m/z;
一级(MS)分辨率:70000
二级(MS/M)分辨率:17500;
喷雾电压:3500V(正离子模式);2500V(负离子模式);
碰撞能量:35±15eV(正离子模式);-35±15eV(负离子模式);
步骤S200:根据所述分析检测数据,对水样中的有机物进行高通量筛查,确认水体中的污染物。具体的,在本发明的实施例中,将步骤S100中所获得的数据文件导入MS-DIAL软件,进行峰提取和对齐,使用公开的大型质谱数据库对水样中的有机物进行高通量筛查,手动检查二级谱图的匹配情况去除假阳性,根据PubChem提供的物质分类信息确认其中的污染物。在本发明的其他实施例中,也可以利用PeakView、Compound Discover等软件进行,公开的大型质谱数据库如MS-DIAL、NIST、MassBank、GNPS质谱数据库等。
参数设置如下:提峰响应:≥30000;对齐保留时间误差:≤0.2min;对齐质量误差:≤0.01Da;筛查质量误差:一级≤0.01Da,二级≤0.002Da。
手动检查去除假阳性的标准如下:若数据库中二级谱图只有一个碎片离子信息,去除没有碎片离子匹配上的物质;若数据库中二级谱图有两个及以上碎片离子信息,去除少于两个碎片离子匹配上的物质;若数据库中二级谱图没有碎片离子信息,去除该物质。
最终筛查鉴定出了河流中132种有机污染物。
步骤S300:根据所确定的污染物,通过网络分析识别污染源。具体的,该步骤包括:
步骤S310:计算有机污染物的峰面积的相关性,并根据所述相关性构建基于相关性的污染物网络,对污染物网络进行类群划分。在本发明的实施例中,计算从上游至下游的受纳水体水样中有机污染物峰面积的皮尔逊相关性,保留显著性p<0.05且为正值的相关关系,将其作为边、将污染物作为节点输入Gephi软件中,构建基于相关性的污染物网络,进行模块化分析,得到污染物网络的类群划分结果。在本发明所提供的示例中,基于相关性的污染物网络如图2所示,污染物网络被划分为8个不同的类群,如图3所示。在本发明的其他实施例中,也可以采用Cytoscape软件,来构建基于相关性的污染物网络。
步骤S320:根据污染物类群,绘制各类群中污染物的峰面积的变化曲线以及各类群污染物的平均峰面积的变化曲线;具体的,可将上游至下游的水样中污染物的峰面积标准化,根据污染物类群划分,绘制各类群中污染物的峰面积的变化曲线图。在此基础上,计算各点位污染物峰面积的平均值,获得各类群污染物的平均峰面积的变化曲线。在所提供的示例中,如图3所示。
步骤S330:根据大型类群中污染物的平均峰面积的变化曲线从上游至下游剧烈增加的点位确定潜在污染源位置区间,同时考虑该类群污染物峰面积在此点位剧烈增加的 一致性,结合实际地理信息,确定污染源。
如果追溯到受纳水体的支流是污染来源,则应重复步骤S100~S300,直到识别到具体的污染点源。
具体的,结合图3,黑色的曲线为污染物的平均峰面积的变化曲线。灰色的曲线为污染物的峰面积的变化曲线,在确定污染源时,可以看到,污染物的平均峰面积的变化曲线陡增的地方,大多数的污染物的峰面积的变化曲线也发生陡增的情况,然后结合实际的地理信息,可以确定污染源。所提供示例中,根据前3个大型类群中该曲线从上游至下游剧烈增加的点位确定2个潜在污染源的位置区间,同时考虑这3个类群污染物峰面积在这2个点位剧烈增加的一致性,结合实际地理信息,确定附近的污水处理厂和一条支流是其污染来源。
步骤S400:根据所识别的污染源以及所述污染源的受纳水体中的有机污染物,利用机器学习分类模型,确定污染源中的关键污染物,量化其污染贡献。
在本发明的一个实施例中,采用的机器学习分类模型为随机森林模型,当然,在其他的实施例中,也可以采用其他的分类模型,例如决策树、支撑向量机等等。此处结合随机森林模型进行具体说明,具体的,可按照步骤S410~S440进行。
步骤S410:针对确定的污染源,测定其汇入受纳水体的水样中的有机污染物;选择同时在污染源和受纳水体中存在的有机污染物,将其在受纳水体各点位的峰面积作为输入。测定污染源汇入受纳水体中的有机污染物时,可按照步骤S110~S140的方法进行检测,然后按照步骤S200的方法筛查鉴定其中有机污染物。在所提供的示例中,最终筛查鉴定出了污水处理厂的废水样中76种有机污染物。选择同时在废水和河流中存在的71种有机污染物,将其在河流各点位的峰面积作为输入。
步骤S420:构建随机森林分类模型,将受纳水体样本相对于污染源的上游或下游作为样本二分类的标准。
在所提供示例中,利用R语言构建随机森林分类模型,将河流样本划分为相对于污染源的8个上游和3个下游样本。
步骤S430:训练随机森林分类模型。在所提供的示例中,通过循环计算袋外错误率(OOB)的平均值来确定模型参数mtry的最佳值为16,混淆矩阵评估模型分类错误率为0%。
步骤S440:输出表征变量重要性的指标,并根据所述重要性指标的数值大小确定关键的污染物及其污染贡献。在所提供的示例中,采用的指标为平均准确率的减少值。 对平均准确率的减少值大于0(设定阈值)的32种污染物,判断它们在距离污染源最近的下游河流样本中的峰面积是否大于所有上游河流样本中的最大值,最终找到25种污染物是污水处理厂这一污染点源中关键污染贡献因子,基于它们的平均准确率的减少值的大小量化它们的污染贡献,如图4所示。
第二方面,本发明提供一种水体有机污染智能化溯源系统,该系统可根据上述水体有机污染溯源方法对水体污染进行溯源,该系统包括:
数据获取单元,用于获取受污染水体自上游至下游的若干水样的高效液相色谱-串联质谱的有机物分析检测数据;
污染物确定单元,用于根据所述分析检测数据,对水样中的有机物进行高通量筛查,确认水体中的污染物;
污染源识别单元,用于根据所确定的污染物,通过网络分析识别污染源;
污染源评价单元,用于根据所识别的污染源以及所述污染源的受纳水体中的有机污染物,利用机器学习分类模型,确定污染源中的关键污染物,量化其污染贡献。
上述各个单元实现相应的功能,与本发明所提供的方法中相应的步骤对应,此处就不再赘述。
上述实施例仅是本发明的优选实施方式,应当指出:对于本技术领域的普通技术人员来说,在不脱离本发明原理的前提下,还可以做出若干改进和等同替换,这些对本发明权利要求进行改进和等同替换后的技术方案,均落入本发明的保护范围。

Claims (10)

  1. 一种水体有机污染智能化溯源方法,其特征在于,包括:
    获取受污染水体自上游至下游的若干水样的高效液相色谱-串联质谱的有机物分析检测数据;
    根据所述分析检测数据,对水样中的有机物进行高通量筛查,确认水体中的污染物;
    根据所确定的污染物,通过网络分析识别污染源;
    根据所识别的污染源以及所述污染源的受纳水体中的有机污染物,利用机器学习分类模型,确定污染源中的关键污染物,量化其污染贡献。
  2. 根据权利要求1所述的方法,其特征在于,根据所述分析检测数据,对水样中的有机物进行高通量筛查,确认水体中的污染物包括:
    将分析检测得到的数据文件导入分析软件,进行峰提取和对齐,使用公开的大型质谱数据库对水样中的有机物进行高通量筛查,手动检查二级谱图的匹配情况去除假阳性,根据PubChem提供的物质分类信息确认其中的污染物。
  3. 根据权利要求2所述的方法,其特征在于,所述根据所确定的污染物,通过网络分析识别污染源包括:
    计算有机污染物的峰面积的相关性,并根据所述相关性构建基于相关性的污染物网络,对污染物网络进行类群划分;
    根据污染物类群,绘制各类群中污染物的峰面积的变化曲线以及各类群污染物的平均峰面积的变化曲线;
    根据大型类群中污染物的平均峰面积的变化曲线从上游至下游剧烈增加的点位确定潜在污染源位置区间,同时考虑该类群污染物峰面积在此点位剧烈增加的一致性,结合实际地理信息,确定污染源。
  4. 根据权利要求3所述的方法,其特征在于,所述计算有机污染物的峰面积的相关性,并根据所述相关性构建基于相关性的污染物网络,对污染物网络进行类群划分包括:
    计算从上游至下游的受纳水体水样中有机污染物峰面积的相关性,保留显著性p<0.05且为正值的相关关系,将其作为边、将污染物作为节点输入网络分析软件中,构建基于相关性的污染物网络,进行模块化分析,得到污染物网络的类群划分结果。
  5. 根据权利要求4所述的方法,其特征在于,所述根据污染物类群,划分绘制各类群中污染物的峰面积的变化曲线以及各类群污染物的平均峰面积的变化曲线包括:
    将上游至下游的水样中污染物的峰面积标准化,根据污染物类群划分,绘制各类群 中污染物的峰面积的变化曲线图;
    根据污染物类群划分,计算各点位污染物峰面积的平均值,获得各类群污染物的平均峰面积的变化曲线。
  6. 根据权利要求1-5任一项所述的方法,其特征在于,所述机器学习分类模型为随机森林模型。
  7. 根据权利要求6所述的方法,其特征在于,根据所识别的污染源以及所述污染源处受纳水体中的有机污染物,利用机器学习分类模型,确定污染源中的关键污染物,量化其污染贡献包括:
    针对确定的污染源,测定其汇入受纳水体的水样中的有机污染物;选择同时在污染源和受纳水体中存在的有机污染物,将其在受纳水体各点位的峰面积作为输入;
    构建随机森林分类模型,将受纳水体样本相对于污染源的上游或下游作为样本二分类的标准;
    训练随机森林分类模型;
    输出表征变量重要性的指标,并根据所述重要性指标的数值大小确定关键的污染物及其污染贡献。
  8. 根据权利要求7所述的方法,其特征在于,输出表征变量重要性的指标,并根据所述重要性指标的数值大小确定关键污染物及其污染贡献包括:
    对重要性指标的数值大于设定阈值的变量,认为其是潜在的污染贡献因子,判断其峰面积在受纳水体相对于污染源的下游样本和上游样本之间的大小关系;
    若潜在污染贡献因子的峰面积在下游样本中比在上游样本中大,则认为是污染源中关键污染物,基于这些关键污染物的重要性指标的数值的大小量化它们的污染贡献。
  9. 根据权利要求8所述的方法,其特征在于,获取受污染水体自上游至下游的若干水样的高效液相色谱-串联质谱的有机物分析检测数据包括:
    采集接受污染的水体从上游至下游的水样;
    对水样进行处理,尽可能全面地提取和富集其中的有机物,得到待测液;
    使用高效液相色谱-串联质谱对样品进行有机物的分析检测。
  10. 一种水体有机污染智能化溯源系统,利用权利要求1-9任一项所述的水体有机污染智能化溯源方法对水体污染进行溯源,其特征在于,包括:
    数据获取单元,用于获取受污染水体自上游至下游的若干水样的高效液相色谱-串联质谱的有机物分析检测数据;
    污染物确定单元,用于根据所述分析检测数据,对水样中的有机物进行高通量筛查,确认水体中的污染物;
    污染源识别单元,用于根据所确定的污染物,通过网络分析识别污染源;
    污染源评价单元,用于根据所识别的污染源以及所述污染源的受纳水体中的有机污染物,利用机器学习分类模型,确定污染源中的关键污染物,量化其污染贡献。
PCT/CN2022/077587 2021-12-30 2022-02-24 一种水体有机污染智能化溯源方法及系统 WO2023024463A1 (zh)

Priority Applications (1)

Application Number Priority Date Filing Date Title
US18/005,565 US11965871B2 (en) 2021-12-30 2022-02-24 Method and system for intelligent source tracing of organic pollution of water body

Applications Claiming Priority (2)

Application Number Priority Date Filing Date Title
CN202111650481.7 2021-12-30
CN202111650481.7A CN114295749B (zh) 2021-12-30 2021-12-30 一种水体有机污染智能化溯源方法及系统

Publications (1)

Publication Number Publication Date
WO2023024463A1 true WO2023024463A1 (zh) 2023-03-02

Family

ID=80973790

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/CN2022/077587 WO2023024463A1 (zh) 2021-12-30 2022-02-24 一种水体有机污染智能化溯源方法及系统

Country Status (3)

Country Link
US (1) US11965871B2 (zh)
CN (1) CN114295749B (zh)
WO (1) WO2023024463A1 (zh)

Cited By (9)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN116148400A (zh) * 2023-04-20 2023-05-23 北京大学 基于污染源和污染受体高分辨质谱数据的定量源解析方法
CN116263444A (zh) * 2023-04-20 2023-06-16 北京大学 高分辨质谱非靶向分析水体污染源识别与溯源方法
CN116340723A (zh) * 2023-05-22 2023-06-27 安徽中科大国祯信息科技有限责任公司 基于大数据的乡村水污染快速溯源方法及系统
CN116630812A (zh) * 2023-07-21 2023-08-22 四川发展环境科学技术研究院有限公司 基于可见光图像分析的水体特征检测方法及系统
CN116881749A (zh) * 2023-09-01 2023-10-13 北京建工环境修复股份有限公司 一种污染场地施工监测方法及系统
CN117272070A (zh) * 2023-11-21 2023-12-22 北京大学 一种同时利用有机和无机污染物的定量源解析方法
CN117373557A (zh) * 2023-12-05 2024-01-09 山东清控生态环境产业发展有限公司 基于多维数据的溯源方法及系统
CN117709208A (zh) * 2024-02-05 2024-03-15 四川国蓝中天环境科技集团有限公司 一种人为排放污染物的大气环境容量计算方法
CN117808173A (zh) * 2024-02-29 2024-04-02 四川省水利科学研究院 水田肥度检测方法、相关产品及基于其的种植方法

Families Citing this family (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN115309845B (zh) * 2022-08-23 2023-05-02 北京万维盈创科技发展有限公司 基于水系编译器与机器学习处理的水污染溯源方法及系统
CN116735276A (zh) * 2023-06-25 2023-09-12 深圳市汉宇环境科技有限公司 一种废水中新污染物采样方法
CN117649156B (zh) * 2024-01-30 2024-04-12 水利部交通运输部国家能源局南京水利科学研究院 高城镇化河网全自动水环境监测方法及系统

Citations (10)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN102661939A (zh) * 2012-05-15 2012-09-12 北京化工大学 一种快速实现水污染溯源的方法
WO2016122042A1 (ko) * 2015-01-29 2016-08-04 계명대학교 산학협력단 인공위성 영상과 랜덤포레스트 분류기 결합을 이용한 자동 하천 검출 시스템 및 방법
CN107480839A (zh) * 2017-10-13 2017-12-15 深圳市博安达信息技术股份有限公司 基于主成分分析和随机森林的高危污染源的分类预测方法
CN108008048A (zh) * 2017-12-08 2018-05-08 环境保护部华南环境科学研究所 一种地下水系统中新型污染物的污染源识别方法与应用
JP2019105502A (ja) * 2017-12-12 2019-06-27 株式会社東芝 学習モデル製造方法、汚染密度算出方法および汚染密度算出装置
CN110004239A (zh) * 2019-04-03 2019-07-12 河海大学 一种基于微生物溯源解析水体污染源的方法
CN110308255A (zh) * 2019-05-08 2019-10-08 宁波大学 一种基于污染指示菌群对近海水体污染程度定量预测方法
CN111118128A (zh) * 2020-01-06 2020-05-08 华东理工大学 一种利用微生物群落图谱解析河流水体污染来源的方法
CN111161802A (zh) * 2020-01-06 2020-05-15 华东理工大学 一种利用微生物生物标记的湖泊水体污染来源解析方法
CN112613675A (zh) * 2020-12-29 2021-04-06 南开大学 一种分析污染源与气象因素对不同程度pm2.5污染影响贡献和效应的机器学习模型

Family Cites Families (7)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20190065674A1 (en) * 2017-08-22 2019-02-28 Noblis, Inc. Nucleic acid sample analysis
US20220229036A1 (en) * 2019-06-12 2022-07-21 B.G. Negev Technologies & Applications Ltd., At Ben-Gurion University System and method for determining aqueous nitrate concentration in solution containing dissolved organic carbon
CN111461167B (zh) * 2020-03-02 2024-06-07 平安国际智慧城市科技股份有限公司 基于大数据的污染源定位方法、装置、设备及存储介质
US20220157408A1 (en) * 2020-11-13 2022-05-19 Wynd Technologies, Inc. Pollution type sensing
CN112986497B (zh) * 2021-02-09 2022-04-01 浙江大学 基于气体传感器阵列指纹识别的污染气体溯源方法
CN113128129B (zh) * 2021-05-07 2023-03-24 大连理工大学 一种突发水污染正逆耦合溯源方法及系统
US20230176027A1 (en) * 2021-12-08 2023-06-08 Maytronics Ltd. Apparatus for measuring characteristics of a water facility

Patent Citations (10)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN102661939A (zh) * 2012-05-15 2012-09-12 北京化工大学 一种快速实现水污染溯源的方法
WO2016122042A1 (ko) * 2015-01-29 2016-08-04 계명대학교 산학협력단 인공위성 영상과 랜덤포레스트 분류기 결합을 이용한 자동 하천 검출 시스템 및 방법
CN107480839A (zh) * 2017-10-13 2017-12-15 深圳市博安达信息技术股份有限公司 基于主成分分析和随机森林的高危污染源的分类预测方法
CN108008048A (zh) * 2017-12-08 2018-05-08 环境保护部华南环境科学研究所 一种地下水系统中新型污染物的污染源识别方法与应用
JP2019105502A (ja) * 2017-12-12 2019-06-27 株式会社東芝 学習モデル製造方法、汚染密度算出方法および汚染密度算出装置
CN110004239A (zh) * 2019-04-03 2019-07-12 河海大学 一种基于微生物溯源解析水体污染源的方法
CN110308255A (zh) * 2019-05-08 2019-10-08 宁波大学 一种基于污染指示菌群对近海水体污染程度定量预测方法
CN111118128A (zh) * 2020-01-06 2020-05-08 华东理工大学 一种利用微生物群落图谱解析河流水体污染来源的方法
CN111161802A (zh) * 2020-01-06 2020-05-15 华东理工大学 一种利用微生物生物标记的湖泊水体污染来源解析方法
CN112613675A (zh) * 2020-12-29 2021-04-06 南开大学 一种分析污染源与气象因素对不同程度pm2.5污染影响贡献和效应的机器学习模型

Cited By (18)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN116148400A (zh) * 2023-04-20 2023-05-23 北京大学 基于污染源和污染受体高分辨质谱数据的定量源解析方法
CN116263444A (zh) * 2023-04-20 2023-06-16 北京大学 高分辨质谱非靶向分析水体污染源识别与溯源方法
CN116148400B (zh) * 2023-04-20 2023-06-27 北京大学 基于污染源和污染受体高分辨质谱数据的定量源解析方法
CN116263444B (zh) * 2023-04-20 2023-07-07 北京大学 高分辨质谱非靶向分析水体污染源识别与溯源方法
CN116340723A (zh) * 2023-05-22 2023-06-27 安徽中科大国祯信息科技有限责任公司 基于大数据的乡村水污染快速溯源方法及系统
CN116340723B (zh) * 2023-05-22 2023-08-01 安徽中科大国祯信息科技有限责任公司 基于大数据的乡村水污染快速溯源方法及系统
CN116630812A (zh) * 2023-07-21 2023-08-22 四川发展环境科学技术研究院有限公司 基于可见光图像分析的水体特征检测方法及系统
CN116630812B (zh) * 2023-07-21 2023-09-26 四川发展环境科学技术研究院有限公司 基于可见光图像分析的水体特征检测方法及系统
CN116881749A (zh) * 2023-09-01 2023-10-13 北京建工环境修复股份有限公司 一种污染场地施工监测方法及系统
CN116881749B (zh) * 2023-09-01 2023-11-17 北京建工环境修复股份有限公司 一种污染场地施工监测方法及系统
CN117272070A (zh) * 2023-11-21 2023-12-22 北京大学 一种同时利用有机和无机污染物的定量源解析方法
CN117272070B (zh) * 2023-11-21 2024-02-02 北京大学 一种同时利用有机和无机污染物的定量源解析方法
CN117373557A (zh) * 2023-12-05 2024-01-09 山东清控生态环境产业发展有限公司 基于多维数据的溯源方法及系统
CN117373557B (zh) * 2023-12-05 2024-02-27 山东清控生态环境产业发展有限公司 基于多维数据的溯源方法及系统
CN117709208A (zh) * 2024-02-05 2024-03-15 四川国蓝中天环境科技集团有限公司 一种人为排放污染物的大气环境容量计算方法
CN117709208B (zh) * 2024-02-05 2024-04-16 四川国蓝中天环境科技集团有限公司 一种人为排放污染物的大气环境容量计算方法
CN117808173A (zh) * 2024-02-29 2024-04-02 四川省水利科学研究院 水田肥度检测方法、相关产品及基于其的种植方法
CN117808173B (zh) * 2024-02-29 2024-04-30 四川省水利科学研究院 水田肥度检测方法、相关产品及基于其的种植方法

Also Published As

Publication number Publication date
CN114295749A (zh) 2022-04-08
US20240027419A1 (en) 2024-01-25
CN114295749B (zh) 2022-10-25
US11965871B2 (en) 2024-04-23

Similar Documents

Publication Publication Date Title
WO2023024463A1 (zh) 一种水体有机污染智能化溯源方法及系统
CN109709057B (zh) 水质指标预测模型构建方法及水质指标监测方法
CN111368401A (zh) 污染源的溯源方法、装置和存储介质
CN109470667A (zh) 一种结合水质参数和三维荧光光谱进行污染物溯源的方法
Cuccia et al. Size-resolved comprehensive characterization of airborne particulate matter
CN104730053B (zh) 一种利用三维荧光光谱反映城市污水厂运行状态的监测方法
Vlachou et al. Development of a versatile source apportionment analysis based on positive matrix factorization: a case study of the seasonal variation of organic aerosol sources in Estonia
CN112198144B (zh) 一种快速污水溯源的方法及系统
CN109633094B (zh) 一种臭气浓度在线监测方法
Zhu et al. Online recognition of drainage type based on UV-vis spectra and derivative neural network algorithm
CN104212804A (zh) 用于铅离子定量快速检测的核酸适配体序列及其检测方法
CN114563381A (zh) 一种水体污染溯源方法
CN114660030A (zh) 污染源解析方法、装置及存储介质
CN116263444B (zh) 高分辨质谱非靶向分析水体污染源识别与溯源方法
CN116148400B (zh) 基于污染源和污染受体高分辨质谱数据的定量源解析方法
CN112630202A (zh) 一种城市排水系统的雨天溢流污水来源的识别方法
CN115950864A (zh) 一种基于三维荧光光谱法的水污染溯源检测方法
CN103076387A (zh) 一种基于dart-tof技术的书写时间鉴别方法
Di Lecce et al. Online wastewater monitoring system using solid state sensor
López-Kleine et al. UV-vis in situ spectrometry data mining through linear and non linear analysis methods
CN115575363A (zh) 一种获取生态影响机制的方法及系统
Danczak et al. Riverine organic matter functional diversity increases with catchment size
CN111007220B (zh) 一种生猪养殖污水水质敏捷监测的方法
CN104280491B (zh) 一种检测土壤中11种离子液体阳离子的分析方法
CN117272070B (zh) 一种同时利用有机和无机污染物的定量源解析方法

Legal Events

Date Code Title Description
WWE Wipo information: entry into national phase

Ref document number: 18005565

Country of ref document: US

121 Ep: the epo has been informed by wipo that ep was designated in this application

Ref document number: 22859822

Country of ref document: EP

Kind code of ref document: A1