Atmospheric pollutant microscale traceability method based on park atmospheric concentration monitoring

Abstract

The invention discloses an atmospheric pollutant micro-scale traceability method based on park atmospheric concentration monitoring, and the method comprises the steps: carrying out the statistical standardization of atmospheric pollutant concentration monitoring data, carrying out the data screening through employing an abnormal region data correction algorithm, screening an abnormal high-value region, and marking the abnormal high-value region as a possible region; a probability density space distribution sample of the position where a possible strong emission source is located under the specific wind direction is judged by taking the standard deviation of the local abnormal high value region and the pollutant concentration mean value as a criterion; carrying out detrending processing on data samples under different wind direction backgrounds, carrying out probability summation on all probability density space distribution samples after detrending, and taking concentration as a weighting function to obtain an overall probability density distribution space function of the near-surface emission intensity of the park; and projecting and mapping the probability density spatial distribution probability graph to an enterprise map, and positioning the position of an emission source. According to the invention, the air pollutant concentration data is subjected to multi-dimensional analysis and processing, so that the position of the strong emission source can be identified and positioned more accurately.

Description

Technical Field

[0001] The invention belongs to the technical field of measurement data processing, and in particular relates to a micro-scale tracing method of atmospheric pollutants based on park atmospheric concentration monitoring. Background Art

[0002] Strong emission sources within the spatial scope of industrial parks have a significant impact on the environmental air quality and human health of the surrounding areas. Many related air pollution control and management work must be based on tracing the source of pollution.

[0003] The existing common methods for tracing the source of air pollutants mainly include:

[0004] Pollution source analysis model: Using the distribution and trend of pollutant concentrations, combined with geographic information system (GIS) data, the location and possible sources of pollution sources are inferred. This method relies on the analysis and pattern recognition of monitoring data.

[0005] Isotope tracer technology: uses the distribution characteristics of stable isotopes or radioactive isotopes to track the migration and source of pollutants in the atmosphere, thereby determining the location and nature of the pollution source.

[0006] Meteorological and mathematical model tracing: Through meteorological observation and monitoring means such as stations, combined with meteorological models, the transmission, transformation and accumulation of pollutants in the specified area are simulated to provide a physical model basis for tracing research. Statistical methods are used to model and analyze the spatiotemporal distribution of atmospheric pollutants, and the possible locations and characteristics of pollution sources are identified through correlation analysis and regression analysis.

[0007] Product method tracing: The type of pollution source and the concentration of pollutants are determined by analyzing the physical and chemical characteristics and formation process of atmospheric pollutants such as aerosol particles and NOx. At the same time, various technical means are used to detect and analyze basic parameters such as the material characteristics and isotopes of pollutants. However, the various reactions of pollutants in the atmosphere are relatively complex. It can be seen that the existing tracing methods have certain limitations.

[0008] CN116258101A proposes a method for quickly calculating emission source strength based on gas concentration monitoring data. The method relocates the location data of wind field measuring stations, obtains wind field data for wind field assimilation, obtains concentration data, and combines the wind field assimilation results to perform concentration field assimilation. Based on the corrected wind field measuring station data and the concentration field assimilation results, grid data of the spatial distribution of pollutant concentrations are generated. Diffusion mode analysis is performed in combination with the wind field assimilation results to calculate the total emission source strength of the target area. Summary of the invention

[0009] In order to solve the deficiencies in the prior art, the purpose of the present invention is to provide a micro-scale tracing method for atmospheric pollutants based on park atmospheric concentration monitoring.

[0010] To achieve the purpose of the present invention, the technical solution adopted by the present invention is:

[0011] A method for tracing the source of atmospheric pollutants at a microscale based on monitoring of atmospheric concentration in a park, comprising the following steps:

[0012] Step 1: Statistically standardize the atmospheric pollutant concentration monitoring data in the time dimension, use the abnormal area data correction algorithm to screen the data, screen the abnormally high value areas, and mark them as "possible areas";

[0013] Step 2: Using the standard deviation of the local abnormally high value area and the mean pollutant concentration as the criterion, the probability density spatial distribution samples of the possible strong emission sources under a specific wind direction are identified;

[0014] Step 3: Detrend the data samples under different wind direction backgrounds to remove the probability density deviation of the emission source location caused by the probability distribution of regional wind direction frequency;

[0015] Step 4: Probability summation is performed on all probability density spatial distribution samples after detrending, and concentration is used as the weight function to obtain the probability density distribution spatial function of the overall near-ground emission intensity of the park for multiple wind direction scenarios;

[0016] Step five: Project the gridded superimposed probability density spatial distribution probability map onto the enterprise map, and locate the emission source based on the actual enterprise operation and process conditions.

[0017] Furthermore, in step 1, the atmospheric pollutant concentration monitoring data is statistically standardized in the time dimension, specifically,

[0018] Subtract the mean from the data and divide by the standard deviation so that the data distribution approximates the standard normal distribution, which is defined as:

[0019]

[0020] in, is the standard deviation of the mean pollutant concentration, is any variable, is the mean value of the variable.

[0021] Furthermore, in step 1, an abnormal area data correction algorithm is used to screen the data, specifically,

[0022] in, is the monitoring data after standardized statistical analysis at the i-th spatial grid point at the j-th moment; C is the raw data matrix without the above processing, detC is the result of vector determinant operation on the C matrix, its significance is to obtain the modulus length of the C matrix, which is used as the normalization criterion of the whole formula; k is the variable subscript for time dimension data standardization, its range is 1 to m, m is the window width for time dimension data standardization, its value is greater than 1 and less than 1 / 2 of the total length of the data time series; is the quality control threshold function for the jth moment and the kth data normalization time window, and its value is 1 or 0. When k is the window width, the value of any jth moment and i-th spatial grid point involved is and When it is an orthogonal relationship, ;

[0023] is the correction matrix obtained when the value of the raw data at the jth moment is assimilated in the time dimension using a k-window, is the correction matrix obtained when the value of the raw data at the i-th spatial grid point is assimilated in the k-window space dimension. is the raw data value at the i-th spatial grid point at the j-th moment.

[0024] Furthermore, step 2 is specifically implemented by obtaining the site spatial distribution data of the regional near-ground concentration within a period of any background wind direction, and using the standard deviation of the mean pollutant concentration as the confidence criterion , the normalized concentration distribution is used as the probability density spatial distribution sample of the strong emission source area, and the extreme value area is identified as the strongly correlated area of ​​the emission source location with a higher probability.

[0025] Furthermore, in step three, detrending is specifically implemented by selecting different dominant wind direction scenario periods from the long-term series concentration monitoring data of the park, normalizing the concentration data samples of each period, and removing the probability density bias caused by the distribution spectrum of the regional dominant wind direction.

[0026] Furthermore, in step 4, the probability density distribution spatial function of the overall park near-ground emission intensity is calculated as follows:

[0027] Calculate the wind frequency distribution probability density:

[0028]

[0029] in, It is the local variable frequency average value of the filtered data; is the standard deviation of the sample in the local space and time; n is the sample size, Used to express confidence level; Z *It is a standard score defined in the Z-score way;

[0030] Calculate the probability of high emission area at each wind direction grid point :

[0031]

[0032] in, It represents the probability density of wind frequency distribution under the jth wind direction at any point x,y in space; is the pollutant concentration at that point, as an additive weight function;

[0033] Superimpose the probability of each grid and construct its overall probability density space distribution function:

[0034]

[0035] in, is the probability of high emission area in the final field, is the wind direction frequency distribution function, which is defined as the number of gusts in a specific wind direction divided by the total number of gusts in the entire measurement period; is the total number of wind direction categories; is the probability of high emission area at each wind direction grid point.

[0036] The beneficial effect of the present invention is that, compared with the prior art, the present invention can more accurately identify and locate the location of strong emission sources by performing multi-dimensional analysis and processing on atmospheric pollutant concentration data and utilizing methods such as the summation of probability density distribution, thereby reducing interference and improving the accuracy of tracing. Normalizing pollutant data under different meteorological scenarios to remove probability density deviations caused by the distribution spectrum of the dominant wind direction in the region can more comprehensively evaluate the impact range of emission sources and improve the credibility of tracing results. Through the combination of standardized processing and probability analysis, a distribution probability map of strong emission sources within industrial parks can be constructed. This method has great potential in the location and analysis of strong emission sources, and provides a new and effective tool for environmental protection. BRIEF DESCRIPTION OF THE DRAWINGS

[0037] Figure 1 It is a flow chart of the micro-scale source tracing method of atmospheric pollutants based on park atmospheric concentration monitoring according to the present invention;

[0038] Figure 2 It is the probability density map of the strong SO2 emission source distribution within a certain park;

[0039] Figure 3 It is the result of tracing the strong VOCs emission source points within a certain industrial park. DETAILED DESCRIPTION

[0040] The technical solution of the present invention is further described below in conjunction with the accompanying drawings and embodiments. The following embodiments are only used to more clearly illustrate the technical solution of the present invention, and cannot be used to limit the protection scope of this application.

[0041] like Figure 1 As shown, the microscale tracing method of atmospheric pollutants based on park atmospheric concentration monitoring according to the present invention specifically comprises the following steps:

[0042] (1) Perform statistical standardization of the time dimension of the multi-site air pollutant concentration monitoring data set up within the industrial park, and remove missing values ​​in the air pollutant concentration monitoring data; use the abnormal area data correction algorithm to screen the data, screen out abnormally high value areas, and mark them as "possible areas";

[0043] The purpose of standardizing the atmospheric pollutant concentration monitoring data is usually to make the data statistically comparable. A common method is to subtract the mean from the data and divide it by the standard deviation, so that the data distribution approximates the standard normal distribution. It is defined as:

[0044]

[0045] in, is the standard deviation of the mean pollutant concentration, is any variable, is the mean value of the variable.

[0046] Remove missing values ​​in the atmospheric pollutant concentration monitoring data to obtain a complete data set. Possible methods include interpolation or simply deleting the locations where the missing values ​​are located.

[0047] The abnormal area data correction algorithm is used to screen the data, identify areas with abnormally high concentrations, and mark them as “possible areas”.

[0048] Data screening method using abnormal area data correction algorithm:

[0049]

[0050] in, is the monitoring data after standardized statistical analysis at the i-th spatial grid point at the j-th moment; C is the raw data matrix without the above processing, detC is the result of vector determinant operation on the C matrix, its significance is to obtain the modulus length of the C matrix, which is used as the normalization criterion of the whole formula; k is the variable subscript for time dimension data standardization, its range is 1 to m, m is the window width for time dimension data standardization, its value is greater than 1 and less than 1 / 2 of the total length of the data time series; is the quality control threshold function for the jth moment and the kth data normalization time window, and its value is 1 or 0. When k is the window width, the value of any jth moment and i-th spatial grid point involved is and When it is an orthogonal relationship, .

[0051] is the correction matrix obtained when the value of the raw data at the jth moment is assimilated in the time dimension using a k-window, is the correction matrix obtained when the value of the raw data at the i-th spatial grid point is assimilated in the k-window space dimension. is the raw data value at the i-th spatial grid point at the j-th moment.

[0052] Therefore, the physical meaning of the above formula is: , first generate and The two correction matrices are used to calculate the possible correction amount in the spatial and temporal dimensions with the value k as the window width, that is, the change range between the data and the surrounding data becomes relatively "smooth" after the correction within the time and space range of the same number of grid points; however, if it is found that the above two corrections in the time and space dimensions make the data change orthogonal, then the correction is considered unnecessary or inaccurate, and then The value of is 0 to skip the correction here. When the correction is completed for all k values, all corrections are accumulated and normalized (to avoid affecting the absolute value of the data due to different corrections accumulated for data c at different i, j time and space points), thus generating the corrected standardized revised monitoring data .

[0053] (2) Using multiple criteria such as the local abnormally high-value areas and the standard deviation of the mean pollutant concentration within the park, the probability density spatial distribution samples of possible strong emission source areas within the park at specific time periods, that is, under specific meteorological background conditions with specified wind directions, are identified;

[0054] Obtain the site spatial distribution data of the regional near-ground concentration during any period of time when the background wind direction and wind speed are relatively stable, and use the standard deviation of the mean pollutant concentration as the confidence criterion , the normalized concentration distribution is used as the probability density spatial distribution sample of the strong emission source area, and the extreme value area is identified as the strongly correlated area of ​​the emission source location with a higher probability.

[0055] (3) Detrending pollutant concentration data under different wind direction backgrounds;

[0056] Different dominant wind direction scenario periods are selected from the long-term series concentration monitoring data of the park, and the concentration data samples of each period are normalized. In other words, the probability density deviation caused by the distribution spectrum of the regional dominant wind direction is removed. This process is called detrending.

[0057] (4) Probability summation is performed on all probability density spatial distribution samples after detrending, and concentration is used as the weight function to obtain the probability density distribution spatial function of the overall near-surface emission intensity of the park for multiple scenarios;

[0058] Calculate the wind frequency distribution probability density:

[0059]

[0060] in, is the local variable frequency average of the data after filtering in step 1; is the standard deviation of the sample in the local space and time; n is the sample size, Same as above, used to express confidence level; Z * It is a standard score defined in Z-score format.

[0061] Calculate the probability of high emission area at each wind direction grid point :

[0062]

[0063] In the above formula, It represents the probability density of wind frequency distribution under the jth wind direction at any point x,y in space; is the pollutant concentration at that point, which serves as the summed weight function.

[0064] The probability of each grid is further superimposed and the overall probability density spatial distribution function is constructed:

[0065]

[0066] in, is the probability of high emission area in the final field, is the wind direction frequency distribution function, which is defined as the number of gusts in a specific wind direction divided by the total number of gusts in the entire measurement period; is the total number of wind direction categories; is the probability of high emission area at each wind direction grid point.

[0067] (5) Finally, the gridded superimposed probability density spatial distribution probability map is projected onto the industrial area enterprise map, and the high-value probability density area is compared and verified with the enterprise location and process map. The spatial horizontal position of the high-probability strong emission source can be further determined from the high-probability emission area based on first-line experience.

[0068] like Figure 2 and Figure 3 As shown in the figure, the above method is used to calculate the SO2 and VOCs monitoring data within a certain park, and the calculated distribution probability density is used as the traceability result. The light gray area is the mechanical processing enterprise, the one with a right slash is the chemical enterprise, and the one with a left slash is the coating and dye enterprise. The calculated results are completely consistent with the actual situation.

[0069] The beneficial effect of the present invention is that, compared with the prior art, the present invention can more accurately identify and locate the location of strong emission sources by performing multi-dimensional analysis and processing on atmospheric pollutant concentration data and utilizing methods such as the summation of probability density distribution, thereby reducing interference and improving the accuracy of tracing. Normalizing pollutant data under different meteorological scenarios to remove probability density deviations caused by the distribution spectrum of the dominant wind direction in the region can more comprehensively evaluate the impact range of emission sources and improve the credibility of tracing results. Through the combination of standardized processing and probability analysis, a distribution probability map of strong emission sources within industrial parks can be constructed. This method has great potential in the location and analysis of strong emission sources, and provides a new and effective tool for environmental protection.

[0070] The applicant of the present invention has made a detailed explanation and description of the implementation examples of the present invention in conjunction with the drawings in the specification. However, those skilled in the art should understand that the above implementation examples are only preferred implementation schemes of the present invention, and the detailed description is only to help readers better understand the spirit of the present invention, but not to limit the scope of protection of the present invention. On the contrary, any improvements or modifications based on the inventive spirit of the present invention should fall within the scope of protection of the present invention.

Claims

1. A method for tracing the source of atmospheric pollutants at a microscale based on monitoring of atmospheric concentration in a park, characterized in that: Includes steps: Step 1: Statistically standardize the atmospheric pollutant concentration monitoring data in the time dimension, use the abnormal area data correction algorithm to screen the data, screen the abnormally high value areas, and mark them as "possible areas"; Step 2: Using the standard deviation of the local abnormally high value area and the mean pollutant concentration as the criterion, the probability density spatial distribution samples of the possible strong emission sources under a specific wind direction are identified; Step 3: Detrend the data samples under different wind direction backgrounds to remove the probability density deviation of the emission source location caused by the probability distribution of regional wind direction frequency; Step 4: Probability summation is performed on all probability density spatial distribution samples after detrending, and concentration is used as the weight function to obtain the probability density distribution spatial function of the overall near-ground emission intensity of the park for multiple wind direction scenarios; Step five: Project the gridded superimposed probability density spatial distribution probability map onto the enterprise map, and locate the emission source based on the actual enterprise operation and process conditions.

2. The microscale source tracing method of atmospheric pollutants based on park atmospheric concentration monitoring according to claim 1 is characterized in that: Step 1: Statistical standardization of atmospheric pollutant concentration monitoring data in the time dimension, specifically, Subtract the mean from the data and divide by the standard deviation so that the data distribution approximates the standard normal distribution, which is defined as: ； in, is the standard deviation of the mean pollutant concentration, is any variable, is the mean value of the variable.

3. The method for tracing the source of atmospheric pollutants based on park atmospheric concentration monitoring according to claim 1 is characterized in that: Step 1: Use the abnormal area data correction algorithm to screen the data, specifically, in, is the monitoring data after standardized statistical analysis at the i-th spatial grid point at the j-th moment; C is the raw data matrix without the above processing, detC is the result of vector determinant operation on the C matrix, its significance is to obtain the modulus length of the C matrix, which is used as the normalization criterion of the whole formula; k is the variable subscript for time dimension data standardization, its range is 1 to m, m is the window width for time dimension data standardization, its value is greater than 1 and less than 1 / 2 of the total length of the data time series; is the quality control threshold function for the jth moment and the kth data normalization time window, and its value is 1 or 0. When k is the window width, the value of any jth moment and i-th spatial grid point involved is and When it is an orthogonal relationship, ; is the correction matrix obtained when the value of the raw data at the jth moment is assimilated in the time dimension using a k-window. is the correction matrix obtained when the value of the raw data at the i-th spatial grid point is assimilated in the k-window space dimension. is the raw data value at the i-th spatial grid point at the j-th moment.

4. The method for tracing the source of atmospheric pollutants based on park atmospheric concentration monitoring according to claim 1 is characterized in that: Step 2 is specifically implemented by obtaining the site spatial distribution data of the regional near-ground concentration within any background wind direction period, and using the standard deviation of the mean pollutant concentration as the confidence criterion , the normalized concentration distribution is used as the probability density spatial distribution sample of the strong emission source area, and the extreme value area is identified as the strongly correlated area of ​​the emission source location with a higher probability.

5. The method for tracing the source of atmospheric pollutants based on park atmospheric concentration monitoring according to claim 1 is characterized in that: Step three, detrending is specifically achieved by selecting different dominant wind direction scenario periods from the long-term series concentration monitoring data of the park, normalizing the concentration data samples of each period, and removing the probability density bias caused by the distribution spectrum of the regional dominant wind direction.

6. The method for tracing the source of atmospheric pollutants based on the monitoring of atmospheric concentration in a park according to claim 1 is characterized in that: Step 4: The probability density distribution spatial function of the overall park near-ground emission intensity is calculated as follows: Calculate the wind frequency distribution probability density: ； in, It is the local variable frequency average value of the filtered data; is the standard deviation of the sample in the local space and time; n is the sample size, Used to express confidence level; Z * It is a standard score defined in the Z-score way; Calculate the probability of high emission area at each wind direction grid point : ； in, It represents the probability density of wind frequency distribution under the jth wind direction at any point x,y in space; is the pollutant concentration at that point, as an additive weight function; Superimpose the probability of each grid and construct its overall probability density space distribution function: ； in, is the probability of high emission area in the final field, is the wind direction frequency distribution function, which is defined as the number of gusts in a specific wind direction divided by the total number of gusts in the entire measurement period; is the total number of wind direction categories; is the probability of high emission area at each wind direction grid point.

Images