Method and system for intelligent source tracing of organic pollution of water body

Patent metadata

Field	Value
Publication number	US-11965871-B2
Application number	US-202218005565-A
Country	US
Kind code	B2
Filing date	Feb 24, 2022
Priority date	Dec 30, 2021
Publication date	Apr 23, 2024
Grant date	Apr 23, 2024

Abstract

Disclosed are a method and system for intelligent source tracing of organic pollution of a water body, which belongs to the technical field of environmental analytical chemistry. The method comprises: acquiring organic matter analysis and detection data from water samples of a polluted water body through high performance liquid chromatography-tandem mass spectrometry; performing high-throughput screening on the organic matter in the water samples according to said data to determine pollutants; identifying pollution sources by means of network analysis according to the determined pollutants; and according to the identified pollution sources and the organic pollutants, determining key pollutants and quantifying the pollution contributions thereof by using a machine learning classification model. The present invention can achieve intelligent tracing of the pollution sources and the key pollutants thereof in the case of the pollution sources being unknown, and provide technical support for investigation and control of organic pollution in a water environment.

First claim

The invention claimed is:

1. A method for intelligent source tracing of organic pollution of a water body, comprising:
acquiring organic matter analysis and detection data from a plurality of water samples of a polluted water body from upstream to downstream through high performance liquid chromatography-tandem mass spectrometry;
performing high-throughput screening on the organic matter in the water samples according to the analysis and detection data to determine pollutants in the water body;
identifying pollution sources by means of network analysis according to the determined pollutants; and
according to the identified pollution sources and the organic pollutants in the receiving water body at the pollution sources, determining key pollutants in the pollution sources and quantifying the pollution contributions of the key pollutants by using a machine learning classification model;
wherein identifying pollution sources by means of network analysis according to the determined pollutants comprises:
calculating the correlation of the peak areas of the organic pollutants, constructing a correlation-based pollutant network according to the correlation, and performing group classification on the pollutant network;
according to the pollutant groups, drawing change curves of the peak areas of the pollutants in the groups and change curves of mean peak areas of the pollutants in the groups; and
according to the sites where the change curves of the mean peak areas of the pollutants in the large groups sharply increase from upstream to downstream, determining location intervals of potential pollution sources, and meanwhile, considering the consistency of the peak areas of the pollutants in the groups at the sites, determining the pollution sources according to actual geographic information;
wherein calculating the correlation of the peak areas of the organic pollutants, constructing a correlation-based pollutant network according to the correlation, and performing group classification on the pollutant network, comprises:
calculating the correlation of the peak areas of the organic pollutants in the water samples of the receiving water body from upstream to downstream, taking the correlation relationship as an edge and the pollutants as nodes, inputting them into the network analysis software, such as Gephi or Cytoscape to construct the correlation-based pollutant network, and performing modular analysis to obtain a group classification result of the pollutant network;
wherein according to the categories of the pollutant groups, drawing change curves of the peak areas of the pollutants in the groups and change curves of mean peak areas of the pollutants in the groups comprises:
standardizing the peak areas of the pollutants in the water samples from upstream to downstream, and drawing change curve graphs of the peak areas of the pollutants in the groups according to the categories of the pollutant groups;
according to the categories of the pollutant groups, calculating mean values of the peak areas of the pollutants at each site to obtain the change curves of the mean peak areas of the pollutants in the groups.

2. The method of claim 1 , wherein according to the analysis and detection data, performing high-throughput screening on the organic matter in the water samples to determine pollutants in the water body comprises:
importing data files obtained by means of analysis and detection into analysis software to perform peak extraction and alignment, using a public large mass spectrometry database to perform the high-throughput screening on the organic matter in the water samples, manually checking the matching of secondary spectra to remove false positives, and determining the pollutants in the water body according to the matter classification information provided by PubChem.

3. The method of claim 2 , wherein the machine learning classification model is a random forest model.

4. The method of claim 1 , the calculating the correlation of the peak areas of the organic pollutants in the water samples of the receiving water body from upstream to downstream, retaining the correlation relationship with the significance p<0.05, where the p value is a positive value.

5. A system for intelligent source tracing of organic pollution of a water body, which uses the method for intelligent source tracing of organic pollution of a water body described in of claim 4 , the system comprising:
a data acquisition unit for acquiring organic matter analysis and detection data from a plurality of water samples of a polluted water body from upstream to downstream through high performance liquid chromatography-tandem mass spectrometry;
a pollutant determination unit for performing high-throughput screening on the organic matter in the water samples according to the analysis and detection data to determine pollutants in the water body;
a pollution source identification unit for identifying pollution sources by means of network analysis according to the determined pollutants; and
a pollution source evaluation unit for determining, according to the identified pollution sources and the organic pollutants in the receiving water body at the pollution sources, key pollutants in the pollution sources and quantifying the pollution contributions of the key pollutants by using a machine learning classification model.

6. The method of claim 4 , wherein the machine learning classification model is a random forest model.

7. The method of claim 1 , wherein the machine learning classification model is a random forest model.

8. The method of claim 7 , wherein according to the identified pollution sources and the organic pollutants in the receiving water body at the pollution sources, determining key pollutants in the pollution sources and quantifying the pollution contributions of the key pollutants by using a machine learning classification model comprises:
determining the organic pollutants in the water samples of the receiving water body into which the determined pollution sources flow, selecting the organic pollutants existing in both the pollution sources and the receiving water body, and taking the peak areas of the organic pollutants at each site of the receiving water body as inputs;
constructing the random forest classification model, and taking the samples of the receiving water body as criteria for binary classification relative to the upstream or the downstream of the pollution sources;
training the random forest classification model; and
outputting indexes indicating the importance of variables, and determining the key pollutants and the pollution contributions of the key pollutants according to values of the indexes.

9. The method of claim 8 , wherein outputting indexes indicating the importance of variables, and determining the key pollutants and the pollution contributions of the key pollutants according to values of the importance indexes, comprises:
if the values of the importance indexes are greater than variables of set thresholds, regarding the pollutants as potential pollution contribution factors, and judging the relationship between the maximum values of the peak areas of the factors in downstream samples and upstream samples of the receiving water body relative to the pollution sources; and
if the peak areas of the potential pollution contribution factors in the downstream samples are greater than the peak areas in the upstream samples, regarding the factors as the key pollutants in the pollution sources, and quantifying the pollution contributions of the key pollutants on the basis of the values of the importance indexes of these key pollutants.

10. The method of claim 9 , wherein acquiring organic matter analysis and det

Assignees

• Nanjing University Of Technology

Inventors

• Wu Jichun
• Liu Mengwen
• Wei Si
• Zhu Xiaobin
• Yu Nanyang

Classifications

• G01N33/1826 Primary

  Organic contamination in water · CPC title

• G01N30/8637

  Peak shape · CPC title

• G01N30/8693

  Models, e.g. prediction of retention times, method development and validation · CPC title

• G01N33/18

  Water · CPC title