Prediction method of dimethyl isoborneol in lake water during high temperature period

By collecting and analyzing the 2-MIB concentration and related factor data of water bodies in the lake during high temperature periods, a multivariate linear regression model was constructed, which solved the problem of 2-MIB prediction and forecasting of lake water bodies, and achieved accurate prediction of 2-MIB concentration changes and water Safety guarantee for plant water withdrawal.

CN117116335BInactive Publication Date: 2024-11-05TAIHU BASIN HYDROLOGY & WATER RESOURCES MONITORING CENT (TAIHU BASIN WATER ENVIRONMENT MONITORING CENT)
View PDF 0 Cites 0 Cited by

Patent Information

Application Number
CN202311051832.1
Authority / Receiving Office
CN · China
Patent Type
Patents(China)
Current Assignee / Owner
Filing Date
2023-08-21
Publication Date
2024-11-05
Estimated Expiration
Not applicable · inactive patent

AI Technical Summary

Technical Problem

The prediction and forecasting methods for 2-MIB in lake water during high temperatures have not been effectively solved, resulting in challenges in water quality problems and water plant water intake safety.

Method used

Through on-site collection and laboratory detection of 2-MIB concentrations in different water layers of water bodies, combined with water environment factors and hydrological meteorological factor data, qPCR and high-throughput sequencing technology were used to determine the species of 2-MIB olfactory cyanobacteria, and the prediction was constructed using multiple linear regression method. The model predicts the 2-MIB concentration and makes early warning judgment based on the prediction model.

Benefits of technology

It can accurately identify the species and influencing factors of 2-MIB olfactory cyanobacteria in the target waters, predict the changes in 2-MIB concentration in the next few days in advance, provide a scientific basis for water plants, and avoid suspending water withdrawal due to high 2-MIB concentration.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN117116335B_ABST
    Figure CN117116335B_ABST
Patent Text Reader

Abstract

The invention discloses a method for predicting dimethyl isoborneol in lake water during high temperature period, relates to the technical field of water quality early warning, establishes a 2-MIB concentration prediction model under the influence of meteorological factors according to historical 2-MIB concentration and meteorological data; calculates the 2-MIB concentration prediction value according to the weight factor and the meteorological data of the day; compares the 2-MIB concentration prediction value with the critical threshold, and judges whether a 2-MIB high value event will occur in the near future according to the comparison result. The present invention provides a 2-MIB high value early warning based on 2-MIB concentration monitoring, establishes a corresponding model according to historical 2-MIB concentration and meteorological data, calculates the corresponding 2-MIB concentration prediction value, and each data is obtained by calculation, rather than empirical data, which can effectively improve the accuracy of 2-MIB high value early warning.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The invention relates to a method for predicting dimethyl isoborneol in lake water during high temperature periods, and belongs to the technical field of water quality early warning. Background Art

[0002] Lakes are important surface water carriers and drinking water resource reservoirs. In recent years, the outbreak of cyanobacteria caused by eutrophication of lake water has led to a series of water quality problems, attracting widespread attention from scholars at home and abroad. The emergence of cyanobacteria blooms during the high temperature period (May-September) not only reduces the water quality of the water body and affects the urban water supply, but also the filamentous cyanobacteria often occur at a high incidence during this period, and the odor substances produced, mainly 2-methylisoborneol (2-MIB), seriously threaten the water ecological security and water resource quality. 2-MIB is a terpene compound with a musty smell. It has a low odor threshold (<10ng / L) and is difficult to remove by traditional water treatment processes. It not only makes the water body have an unpleasant odor and increases the treatment cost of the water plant, but also enriches in aquatic products such as fish, affecting their use value.

[0003] 2-MIB in lake water is mainly attributed to the growth and metabolism of filamentous cyanobacteria, such as Pseudo-Anabaena, Oscillatoria, Pseudo-Planthophore and Aphanizomenon. The sources of 2-MIB in different regions are different. It is necessary to find the main producers of 2-MIB in the target area to prevent and control the odor problem of lake water from the source. At the same time, the growth and odor production of 2-MIB-producing cyanobacteria are affected by many environmental factors such as hydrology, meteorology and nutrients. However, it can be seen from the existing research results that there are many environmental factors that affect 2-MIB, and there are great differences in the dominant factors of 2-MIB under different environmental conditions.

[0004] At present, there is no 2-MIB prediction and forecasting method for lakes during high temperature periods in China, which poses certain challenges to the safety of water extraction from lake water sources. Summary of the invention

[0005] In order to solve the above problems, the present invention discloses a method for predicting dimethyl isoborneol in lake water during high temperature periods, and its specific technical scheme is as follows:

[0006] The method for predicting dimethyl isoborneol in lake water during high temperature period includes the following steps:

[0007] Step 1: Determination of target water layer

[0008] Collect 2-MIB samples from different water layers on site, test them indoors, and compare and analyze the 2-MIB concentrations in different water layers to determine whether the 2-MIB concentration in the water body is stratified. If the water body is stratified, the water layer with the highest 2-MIB concentration will enter step 2;

[0009] Step 2: Detection of water environmental factors in target water layer

[0010] The water temperature, dissolved oxygen, and pH data of the target water layer were measured on site, and the water quality, phytoplankton, and 2-MIB samples of the target water layer were collected on site. The different forms of nitrogen and phosphorus, phytoplankton species composition, and 2-MIB concentration data were tested and analyzed in the laboratory;

[0011] Step 3: Collection and collation of hydrological and meteorological factor data

[0012] Collect hourly hydrological and meteorological factor data of rainfall, evaporation, humidity, light intensity, wind speed and direction, and air pressure in the area near the target water body from the meteorological website, including data from May to October, and calculate the hydrological and meteorological factor data of the previous day, the average of the day, or different time periods based on the hourly hydrological and meteorological factor data;

[0013] Step 4: 2-MIB-producing olfactory cyanobacteria identification

[0014] The mic gene copy number in the water sample was detected by qPCR technology, and the different odor-producing cyanobacteria species and proportions with mic genes in the water sample were detected by high-throughput sequencing technology. The mic gene copy number of different 2-MIB odor-producing cyanobacteria proportions obtained by high-throughput sequencing were multiplied to obtain the mic gene copy number of different 2-MIB odor-producing cyanobacteria, and the Pearson correlation analysis was performed with the 2-MIB concentration of the target water layer to obtain the 2-MIB odor-producing cyanobacteria species in the target water layer.

[0015] Step 5: Determination of key influencing factors of olfactory cyanobacteria

[0016] Through Pearson correlation analysis, the correlation between the water environmental factors in step 2 and the hydrological and meteorological factors in step 3 and the biomass of 2-MIB odor-producing cyanobacteria was statistically analyzed, and the indicators with correlation coefficients greater than 0.5 were selected for principal component analysis. Through principal component analysis, the environmental factors that contributed the most to the production of 2-MIB odor-producing cyanobacteria were identified;

[0017] Step 6: 2-MIB prediction model construction

[0018] From the environmental factors that contribute most to the 2-MIB odor-producing cyanobacteria identified in step 5, the factors that are most closely related to the growth of odor-producing cyanobacteria are selected to construct a prediction model, and the factors are selected as follows: temperature (T, unit: ℃), wind speed (S, unit: m / s) and light intensity (I, unit: W / m 2 )3 related indicators as independent variables, 2-MIB concentration as the dependent variable, and the multivariate linear regression method was used to build a prediction model. The equation of the prediction model is:

[0019] lg(2-MIB)=A×TB×SC×lg(I)+D

[0020] Among them, T is the temperature index, S is the wind speed index, I is the light intensity index, A, B, and C are the weight factors of T, S, and I respectively, and D is a constant term. The values ​​of A, B, C, and D are automatically given during the model construction process. At the same time, in order to eliminate the influence of the large fluctuation range of 2-MIB concentration and light intensity on the model prediction accuracy, the 2-MIB concentration and light intensity data can be logarithmically processed to further improve the model prediction accuracy;

[0021] Step 7: Determine the accuracy of the prediction model

[0022] The 2-MIB concentration values ​​calculated by the prediction model were compared with the actual monitoring values. The line graph was used to identify the intervals where the 2-MIB concentration values ​​calculated by the prediction model were inconsistent with the actual monitoring values. Targeted re-analysis was then conducted to optimize the indicators involved in building the 2-MIB prediction model to further improve the model accuracy.

[0023] Step 8: Target water layer 2-MIB high risk period threshold definition

[0024] The logarithmic transformation of the dependent variable reduces the impact of the fluctuation range of 2-MIB concentration on the prediction accuracy, but also reduces the prediction accuracy of the prediction model for abnormal peaks, which can cause the predicted value of 2-MIB concentration during the abnormal peak period to be lower than the actual value. During the period of 2-MIB high value, the model prediction value is greater than 2.0. Therefore, 2.0 is selected as the critical threshold for the occurrence of 2-MIB high value.

[0025] Step 9: Use the prediction model to make early warning judgments

[0026] Collect hourly data on temperature, light intensity, and wind speed in the area near the target water body for the next three days from the meteorological website, calculate the T, S, and I indicators in the 2-MIB prediction model and bring them into the prediction model to calculate the 2-MIB occurrence in the target water layer in the next three days. If the calculated result is greater than 2.0, it is necessary to promptly activate the emergency plan to ensure the safety of water intake at the water plant.

[0027] Furthermore, the step 1 is specifically as follows: 2-MIB is mainly distributed in the water depth range of 0-5 meters, and is stratified at intervals of 0.5 meters, that is, 2-MIB samples in the water column of 0-5 meters are collected on-site at 0.5 meters, 1.5 meters, 2.5 meters, 3.5 meters, and 4.5 meters below the water surface, and the 2-MIB concentrations in different water layers are tested in the laboratory. If the 2-MIB concentrations of all water layers differ by 20-30%, the water body is considered to be unstratified, and the water layer at 0.5 meters is focused on; if the 2-MIB concentrations of all water layers differ by more than 30%, the water body is considered to be stratified, and the water layer with the highest 2-MIB concentration is focused on.

[0028] Furthermore, in step 6, the reliability of the established prediction model is determined by the following criteria: first, the F test is used to test whether the overall model is significantly different from zero. If the F value is greater than 0.5, the prediction model can enter the next test. If the F value is less than 0.5, it means that the overall regression of the model is not significant, and it is necessary to return to the previous step to re-screen the temperature, wind speed, and light intensity indicators of the constructed model until the F value is greater than 0.5.

[0029] The beneficial effects of the present invention are:

[0030] Through the method of the present invention, the types and influencing factors of 2-MIB-producing cyanobacteria in the target waters can be determined, and a 2-MIB prediction and forecasting model for the target waters can be established. The changes in 2-MIB concentration in the target waters in the next few days can be predicted in advance, providing a scientific basis for water plants to choose the right time to draw water, and avoiding suspension of water drawing due to excessively high 2-MIB concentration in raw water. BRIEF DESCRIPTION OF THE DRAWINGS

[0031] Figure 1 It is a comparison between the predicted value of MIB concentration of the present invention and the actual value. DETAILED DESCRIPTION

[0032] The technical means disclosed in the scheme of the present invention are not limited to the technical means disclosed in the above-mentioned technical means, but also include technical schemes composed of any combination of the above technical features.

[0033] 1. Determination of target water layer

[0034] 2-MIB is mainly distributed in the water depth range of 0-5 meters. According to the 1-meter interval stratification, that is, 0.5 meters, 1.5 meters, 2.5 meters, 3.5 meters, and 4.5 meters below the water surface, 2-MIB samples in the 0-5 meter water column were collected on-site, and the 2-MIB concentrations in different water layers were tested in the laboratory. The 2-MIB concentrations in different water layers were compared and analyzed to determine whether the 2-MIB concentration in the water body was stratified: if the 2-MIB concentrations in different water layers differed by 20-30%, the water body was considered to be non-stratified, and the focus was on the water layer at 0.5 meters; if the 2-MIB concentrations in different water layers differed by more than 30%, the water body was considered to be stratified, and the focus was on the water layer with the highest 2-MIB concentration.

[0035] (II) Detection of water environmental factors in target water layers

[0036] The water temperature, dissolved oxygen, pH and other data of the target water layer are measured on-site, and the water quality, phytoplankton and 2-MIB samples of the target water layer are collected on-site. The different forms of nitrogen and phosphorus, phytoplankton species composition, and 2-MIB concentration data are tested and analyzed in the laboratory, and the relevant index tests are tested in accordance with existing national standard testing methods.

[0037] (III) Collection and collation of hydrological and meteorological factor data

[0038] Hourly hydrological and meteorological factor data such as rainfall, evaporation, humidity, light intensity, wind speed and direction, and air pressure in the area near the target water body are collected on the meteorological website, including data from May to October. Based on the hourly data, the hydrological and meteorological factor data of the previous day, the average of the day, or different time periods are calculated.

[0039] (IV) Determination of 2-MIB-producing olfactory cyanobacteria

[0040] The mic gene copy number in the water sample was detected by qPCR technology. The different types and proportions of odor-producing cyanobacteria with mic genes in the water sample were detected by high-throughput sequencing technology. According to the mic gene copy number and the proportion of different 2-MIB odor-producing cyanobacteria obtained by high-throughput sequencing, the mic gene copy number of different 2-MIB odor-producing cyanobacteria was multiplied, and the Pearson correlation analysis was performed with the 2-MIB concentration of the target water layer. The types of 2-MIB odor-producing cyanobacteria in the target water layer were finally confirmed according to the correlation coefficient.

[0041] qPCR technology is an existing mature technology. qPCR monitors the changes of PCR amplification products in real time through the fluorescence signal value and performs quantitative analysis during the exponential amplification period. For each cycle of PCR, DNA increases exponentially by 2 times and soon reaches a plateau. Assuming that the amount of starting DNA is A0, after n cycles, the amount of DNA product can theoretically be expressed as: An = A0 × 2n. Then, the more the starting DNA amount A0, the earlier the amount of amplified product reaches the detection value An, and the number of cycles when An is reached is the Ct value. In other words, the more the starting DNA amount A0, the earlier the amplification curve peaks, and the smaller the corresponding number of cycles n required. In theory, the Ct value is inversely proportional to the logarithmic value of the starting template number. If the standard of known concentration is diluted in a gradient and used as a template for Real Time PCR, a series of amplification curves will be obtained at equal intervals in the order of the starting DNA amount from large to small. According to the linear relationship between the Ct value and the logarithmic value of the starting template number, a standard curve can be made. By substituting the Ct value of the sample of unknown concentration into the standard curve, the starting template amount of the sample of unknown concentration can be calculated. This is the quantitative principle of qPCR.

[0042] High-throughput sequencing technology is an existing mature nucleic acid detection technology.

[0043] Pearson correlation analysis is a method to measure vector similarity, which is characterized by the size of the correlation coefficient. The output range is -1 to +1, 0 represents no correlation, negative values ​​represent negative correlation, and positive values ​​represent positive correlation. If the linear relationship between two variables increases, the correlation coefficient tends to 1 or -1. The specific calculation formula is as follows:

[0044]

[0045] According to the correlation coefficient, the one with the largest correlation coefficient is the target water layer 2-MIB olfactory cyanobacteria species.

[0046] qPCR technology, high-throughput sequencing technology and Pearson correlation analysis are all direct uses of existing mature technologies. This patent does not improve the technology, but only uses the technology to perform detection and obtain test results. This patent will conduct subsequent analysis and processing of the test results.

[0047] (V) Determination of key influencing factors of olfactory cyanobacteria

[0048] Pearson correlation analysis was used to screen out factors closely related to the biomass of 2-MIB odor-producing cyanobacteria, and then these factors were subjected to principal component analysis with 2-MIB odor-producing cyanobacteria to identify the environmental factors with the largest contribution. The following criteria were used to determine whether the screened environmental factors were effective: the eigenvalues ​​of the first two principal components in the principal component analysis results were both greater than 1, and the total variance contribution rate was 80.0%. If the above criteria are not met, the environmental factors need to be rescreened until they are met. Based on the results of the principal component analysis, the main factor composition of principal component 1 and principal component 2 was further analyzed to determine the key influencing factors.

[0049] Principal component analysis is a data dimensionality reduction algorithm that maps n-dimensional features to k-dimensional features. This k-dimensional feature is a new orthogonal feature, namely the principal component. The specific process is as follows:

[0050] 1. Input data set X = {x1, x2, x3, ..., x n}, need to be reduced to k dimensions,

[0051] 2. Decentralize, subtract the average value from each feature.

[0052] 3. Calculate the covariance matrix

[0053] 4. Use the eigenvalue decomposition method to find the covariance matrix The eigenvalues ​​and eigenvectors of

[0054] 5. Sort the eigenvalues ​​from large to small, select the largest k, and then use the corresponding k eigenvectors as row vectors to form the eigenvector matrix P.

[0055] 6. Convert the data into a new space constructed by k eigenvectors, that is, Y = PX.

[0056] Principal component analysis is also a mature technology. There are some introductions in existing patents, such as the application number 201811479966.2. Formula 3 also uses the canonical correspondence analysis method to analyze the relationship between dominant species and environmental factors and determine the main driving factors of dominant species. It is also an application example of principal component analysis.

[0057] (VI) Construction of 2-MIB prediction model

[0058] From the identified key influencing factors, the factors most closely related to the growth of odor-producing cyanobacteria are selected to construct a prediction model. Based on existing experience, temperature (T, unit: °C), wind speed (S, unit: m / s) and light intensity (I, unit: W / m 2 )3 related indicators were used as independent variables, 2-MIB concentration was used as the dependent variable, and the prediction model was constructed using the multivariate linear regression method.

[0059] lg(2-MIB)=A×TB×SC×lg(I)+D

[0060] Among them, T is the temperature index, S is the wind speed index, I is the light intensity index, A, B, C are the weight factors of T, S, and I respectively, and D is a constant term. The values ​​of A, B, C, and D are automatically given during the model construction process. At the same time, in order to eliminate the impact of the large fluctuation range of 2-MIB concentration and light intensity on the model prediction accuracy, the two can be logarithmically processed to further improve the model prediction accuracy.

[0061] The reliability of the established model is determined by the following standards: First, the F test is used to test whether the overall model is significantly different from zero. If the p value of the F test is less than 0.05, it means that the overall regression of the model is significant, that is, the model can enter the next step of testing. If the p value of the F test is greater than 0.05, it means that the overall regression of the model is not significant, and it is necessary to return to the previous step to re-screen the temperature, wind speed, and light intensity indicators of the constructed model until the p value of the F test is less than 0.05. Secondly, look at the significance of each coefficient, that is, the t test. If the t value is greater than 2 or the p value is less than 0.05, it means that the coefficient in front of the variable is significantly not 0, that is, the selected variable is valid. If the t value is less than 2 or the p value is greater than 0.05, it means that the selected variable is invalid, and the temperature, wind speed, and light intensity indicators need to be re-screened again until the t value is greater than 2 or the p value is less than 0.05.

[0062] The F test is a test that the statistical value follows the F distribution under the null hypothesis (H0). It is usually used to analyze statistical models with more than one parameter to determine whether all or part of the parameters in the model are suitable for estimating the population. The specific process is as follows:

[0063] 1. Calculate the mean of the two groups of sample data respectively

[0064] 2. Calculate the square of the standard deviation of the two sets of sample data respectively

[0065] 3. Calculate the square ratio of the standard deviation of two sets of sample data Take the larger square as the numerator and the smaller square as the denominator.

[0066] Where: F represents the square ratio of the standard deviation of two groups of sample data, S1 and S2 represent two groups of sample data.

[0067] The F test is also an existing mature detection method. This patent does not make any changes to the F test, but directly uses the F test to further improve the accuracy of the model.

[0068] (VII) Model Accuracy Determination

[0069] The 2-MIB concentration values ​​calculated by the prediction model were compared with the actual monitoring values. The comparison results are shown in Figure 1 . The line graph is used to identify the intervals where the 2-MIB concentration values ​​calculated by the prediction model are inconsistent with the actual monitoring values, and targeted re-analysis is carried out to optimize the indicators involved in building the 2-MIB prediction model to further improve the model accuracy.

[0070] (VIII) Target water layer 2-MIB high risk period threshold definition

[0071] The logarithmic transformation of the dependent variable reduces the impact of the fluctuation range of 2-MIB concentration on the prediction accuracy, but also reduces the prediction accuracy of the prediction model for abnormal peaks, which may cause the predicted value of 2-MIB concentration during the abnormal peak period to be lower than the actual value. Based on existing experience, during the period of high 2-MIB values, the model prediction value is greater than 2.0, so 2.0 is selected as the critical threshold for the occurrence of high 2-MIB values.

[0072] 9. Using prediction models for early warning judgment

[0073] Collect hourly data on temperature, light intensity, and wind speed in the area near the target water body for the next three days from the meteorological website, calculate the T, S, and I indicators in the 2-MIB prediction model and bring them into the prediction model to calculate the 2-MIB occurrence in the target water layer in the next three days. If the calculated result is greater than 2.0, it is necessary to promptly activate the emergency plan to ensure the safety of water intake at the water plant.

[0074] An example of the practical application of this patent is given below:

[0075] East Taihu Lake is a bay of Taihu Lake, with an average water depth of less than 2 meters. It is an important water source for Suzhou City. In recent years, the concentration of 2-MIB in East Taihu Lake has often increased abnormally from May to September, which has had a great impact on the safe water intake of the water source. By testing the concentration of 2-MIB in water samples at 0.5m, 1.0m and 1.5m below the surface of East Taihu Lake, it was found that the concentration of 2-MIB in different water layers is basically the same, that is, there is no stratification of 2-MIB concentration in East Taihu Lake.

[0076] Water samples were collected from the East Taihu Lake at a depth of 0.5 m, and indicators such as water temperature, dissolved oxygen, and pH were measured on site. Phytoplankton, different forms of nitrogen and phosphorus, phytoplankton species composition, and 2-MIB concentration were analyzed in the laboratory.

[0077] Hourly hydrological and meteorological factor data such as rainfall, evaporation, humidity, light intensity, wind speed and direction, and air pressure around East Taihu Lake were collected from the windy website (https: / / www.windy.com).

[0078] Use in the laboratory Spin Kit for Soil (MP Biomedicals, USA) was used for DNA extraction. The extracted DNA samples were partially used for qPCR experiments to detect the absolute quantitative mic gene copy number. Meanwhile, the remaining DNA samples were sent to Shanghai Lingen Biotechnology Co., Ltd. for high-throughput sequencing. The sequencing primers used were mic genes (MIBQSF: 5'-GACAGCTTCTACACCTCCATGA-3' and MIBQSR: 5'-CAATCTGTAGCACCATGTTGAC-3'). The sequencing results showed that there were 6 species of cyanobacteria in the water samples of East Taihu Lake that contained mic genes, among which Planktothricoides had the highest relative abundance. The mic gene copy number and the proportion of the six olfactory cyanobacteria were multiplied to obtain the mic gene copy number of different 2-MIB olfactory cyanobacteria, and a Pearson correlation analysis was performed between it and the 2-MIB concentration in East Taihu Lake. The results showed that there was a strong correlation between the mic gene copy number of Pseudohylochromocysteine ​​and the 2-MIB concentration (r=0.743, P<0.05), confirming that Pseudohylochromocysteine ​​was the 2-MIB olfactory cyanobacteria in East Taihu Lake.

[0079] Table 1 Relative abundance of cyanobacteria containing mic genes

[0080]

[0081]

[0082] Through Pearson correlation analysis and principal component analysis, temperature, wind speed, and light intensity indicators were finally selected to construct a prediction model, where temperature is the temperature at 10 a.m. on the day, wind speed is the average wind speed on the day, and light intensity is the average light intensity on the day. According to the prediction model, the high value of 2-MIB in East Taihu Lake is predicted to occur on cloudy days with high temperature, no wind or poor wind conditions. According to the statistical results of meteorological data over the years, mid-to-late July 2023 may be the period when the high value of 2-MIB occurs.

[0083] Based on the above ideal embodiments of the present invention, the relevant staff can make various changes and modifications without departing from the technical concept of the present invention through the above description. The technical scope of the present invention is not limited to the contents of the specification, and its technical scope must be determined according to the scope of the claims.

Claims

1. A method for predicting dimethyl isoborneol in lake water during high temperature periods, characterized in that: The steps include: Step 1: Determination of target water layer Collect 2-MIB samples from different water layers on site, test them indoors, and compare and analyze the 2-MIB concentrations in different water layers to determine whether the 2-MIB concentration in the water body is stratified. If the water body is stratified, the water layer with the highest 2-MIB concentration will enter step 2; According to the 1-meter interval, that is, 0.5 meters, 1.5 meters, 2.5 meters, 3.5 meters, and 4.5 meters below the water surface, 2-MIB samples in the 0-5 meter water column were collected on site, and the 2-MIB concentrations in different water layers were tested in the laboratory. If the 2-MIB concentrations of all water layers differed by 20-30%, the water body was considered to be non-stratified, and the water layer at 0.5 meters was focused on; if the 2-MIB concentrations of all water layers differed by more than 30%, the water body was considered to be stratified, and the water layer with the highest 2-MIB concentration was focused on; Step 2: Detection of water environmental factors in target water layer The water temperature, dissolved oxygen, and pH data of the target water layer were measured on site, and the water quality, phytoplankton, and 2-MIB samples of the target water layer were collected on site. The different forms of nitrogen and phosphorus, phytoplankton species composition, and 2-MIB concentration data were tested and analyzed in the laboratory; Step 3: Collection and collation of hydrological and meteorological factor data Collect hourly hydrological and meteorological factor data of rainfall, evaporation, humidity, light intensity, wind speed and direction, and air pressure in the area near the target water body from the meteorological website, including data from May to October, and calculate the hydrological and meteorological factor data of the previous day, the average of the day, or different time periods based on the hourly hydrological and meteorological factor data; Step 4: 2-MIB-producing olfactory cyanobacteria identification The mic gene copy number in the water sample was detected by qPCR technology, and the different odor-producing cyanobacteria species and proportions with mic genes in the water sample were detected by high-throughput sequencing technology. The mic gene copy number of different 2-MIB odor-producing cyanobacteria proportions obtained by high-throughput sequencing were multiplied to obtain the mic gene copy number of different 2-MIB odor-producing cyanobacteria, and the Pearson correlation analysis was performed with the 2-MIB concentration of the target water layer to obtain the 2-MIB odor-producing cyanobacteria species in the target water layer. Step 5: Determination of key influencing factors of olfactory cyanobacteria Through Pearson correlation analysis, the correlation between the water environmental factors in step 2 and the hydrological and meteorological factors in step 3 and the biomass of 2-MIB odor-producing cyanobacteria was statistically analyzed, and the indicators with correlation coefficients greater than 0.5 were selected for principal component analysis. Through principal component analysis, the environmental factors that contributed the most to the production of 2-MIB odor-producing cyanobacteria were identified; Step 6: 2-MIB prediction model construction From the environmental factors that contribute most to the 2-MIB odor-producing cyanobacteria identified in step 5, the factors that are most closely related to the growth of odor-producing cyanobacteria are selected to construct a prediction model, and the factors are selected as follows: temperature (T, unit: ℃), wind speed (S, unit: m / s) and light intensity (I, unit: W / m 2 )3 related indicators as independent variables, 2-MIB concentration as the dependent variable, and the multivariate linear regression method was used to build a prediction model. The equation of the prediction model is: lg(2-MIB)=A×TB×SC×lg(I)+D Among them, T is the temperature index, S is the wind speed index, I is the light intensity index, A, B, and C are the weight factors of T, S, and I respectively, and D is a constant term. The values ​​of A, B, C, and D are automatically given during the model construction process. At the same time, in order to eliminate the influence of the large fluctuation range of 2-MIB concentration and light intensity on the model prediction accuracy, the 2-MIB concentration and light intensity data can be logarithmically processed to further improve the model prediction accuracy; The reliability of the established prediction model is determined by the following criteria: First, the F test is used to test whether the overall model is significantly different from zero. If the F value is greater than 0.5, the prediction model can enter the next step of testing. If the F value is less than 0.5, it means that the overall regression of the model is not significant, and it is necessary to return to the previous step to re-screen the temperature, wind speed, and light intensity indicators of the constructed model until the F value is greater than 0.5; Step 7: Determine the accuracy of the prediction model The 2-MIB concentration values ​​calculated by the prediction model were compared with the actual monitoring values. The line graph was used to identify the intervals where the 2-MIB concentration values ​​calculated by the prediction model were inconsistent with the actual monitoring values. Targeted re-analysis was then conducted to optimize the indicators involved in building the 2-MIB prediction model to further improve the model accuracy. Step 8: Target water layer 2-MIB high risk period threshold definition The logarithmic transformation of the dependent variable reduces the impact of the fluctuation range of 2-MIB concentration on the prediction accuracy, but also reduces the prediction accuracy of the prediction model for abnormal peaks, which can cause the predicted value of 2-MIB concentration during the abnormal peak period to be lower than the actual value. During the period of 2-MIB high value, the model prediction value is greater than 2.

0. Therefore, 2.0 is selected as the critical threshold for the occurrence of 2-MIB high value. Step 9: Use the prediction model to make early warning judgments Collect hourly data on temperature, light intensity, and wind speed in the area near the target water body for the next three days from the meteorological website, calculate the T, S, and I indicators in the 2-MIB prediction model and bring them into the prediction model to calculate the 2-MIB occurrence in the target water layer in the next three days. If the calculated result is greater than 2.0, it is necessary to promptly activate the emergency plan to ensure the safety of water intake at the water plant.