A VOCs spatiotemporal distribution regularity analysis system and method

Abstract

The application discloses a VOCs space-time distribution rule analysis system and method, relates to the technical field of air pollution detection, and comprises the following steps: deploying a VOCs monitoring point network; performing multi-time period sampling; adopting a pre-concentration-gas chromatography mass spectrometry combined technology to perform component analysis; constructing a VOCs concentration space distribution model to generate a space-time distribution atlas; analyzing VOCs component sources to identify main pollution source types; and constructing a VOCs space-time distribution characteristic index system. The application adopts a space point arrangement strategy to ensure that the monitoring points are representative. The sampling time sequence and frequency arrangement are optimized to improve the accuracy and integrity of component analysis. An optimized space interpolation algorithm is introduced to improve the accuracy and reliability of the space-time distribution atlas. Multiple source analysis methods are integrated to eliminate the limitations of a single model through model cross verification.

Description

Technical Field

[0001] This invention relates to the field of air pollution detection technology, specifically to a system and method for analyzing the spatiotemporal distribution patterns of VOCs. Background Technology

[0002] Volatile organic compounds (VOCs) refer to a general term for organic compounds with high saturated vapor pressure at normal temperature and pressure. They mainly include alkanes, alkenes, alkynes, aromatic hydrocarbons, and heteroatoms containing oxygen, nitrogen, and sulfur. VOCs not only directly harm human health but are also important precursors to the formation of ozone (O3) and secondary organic aerosols (SOA), playing a crucial role in air pollution.

[0003] Currently, VOCs research mainly focuses on pollutant concentration levels, component distribution, emission source apportionment, and health risk assessment. Existing studies are mostly limited to large cities or specific regions, with a relative lack of systematic research on the VOCs pollution characteristics of small and medium-sized cities and regional areas. Furthermore, research on the spatiotemporal distribution of VOCs typically faces the following technical challenges:

[0004] The monitoring points were not scientifically located, making it difficult to fully reflect the regional VOCs distribution characteristics;

[0005] The sampling method lacks timeliness and is difficult to capture dynamic changes in VOCs concentration;

[0006] The accuracy of component analysis is not high, especially for the identification and quantification of low-concentration components;

[0007] Spatial interpolation methods are limited and cannot accurately depict the spatial transport patterns of VOCs;

[0008] The results from multiple source resolution models vary greatly, and a comprehensive evaluation framework is lacking.

[0009] The incomplete characteristic indicator system makes it difficult to effectively guide VOCs pollution prevention and control.

[0010] Currently, there are VOCs analysis methods for specific scenarios such as pesticide remediation sites, such as "A method for analyzing the characteristics of VOCs in the air of pesticide remediation sites and its application" (Announcement No. CN117269418B). This method mainly focuses on the analysis of VOCs components and source apportionment during the remediation process of contaminated sites, but it still has limitations such as narrow applicability, single analysis method, and difficulty in promoting it to regional VOCs monitoring.

[0011] With the acceleration of urbanization and industrial restructuring, regional atmospheric VOCs pollution is exhibiting new spatiotemporal distribution characteristics. How to construct a VOCs spatiotemporal distribution analysis system applicable to different regions, seasons, and pollution sources has become an urgent technical problem to be solved. Therefore, developing a systematic, widely applicable, and highly accurate method for analyzing VOCs spatiotemporal distribution patterns is of great significance for comprehensively understanding regional VOCs pollution characteristics, formulating precise prevention and control measures, and ensuring ambient air quality. Summary of the Invention

[0012] To address the aforementioned technical shortcomings, the present invention aims to provide a system and method for analyzing the spatiotemporal distribution patterns of VOCs.

[0013] To solve the above-mentioned technical problems, the present invention adopts the following technical solution: The present invention provides a method for analyzing the spatiotemporal distribution patterns of VOCs, comprising the following steps:

[0014] A hierarchical spatial deployment strategy was constructed. Based on the topographic features, meteorological parameters, pollution source intensity distribution and sensitive receptor distribution data of the target area, representative monitoring areas were determined, and a VOCs monitoring point network was deployed in the target area and its surrounding environment.

[0015] Design a time gradient sampling scheme to perform multi-season, multi-time period, and multi-frequency sampling at each of the deployed monitoring points, collect VOCs samples, and record meteorological parameters and surrounding emission activities during the sampling period;

[0016] The collected VOCs samples were analyzed using pre-concentration-gas chromatography-mass spectrometry (GC-MS). Multiple quality control standards were applied to screen the analytical data, a VOCs component concentration matrix was established, and the VOCs component spectrum and concentration distribution characteristics in the target area were determined.

[0017] Based on the geospatial coordinates of monitoring points and the VOCs concentration characteristics of each point, combined with geographic information system technology, an optimized inverse distance weighted interpolation algorithm is used to construct a spatial distribution model of VOCs concentration and generate a spatiotemporal distribution map.

[0018] By integrating principal component analysis and positive definite matrix factorization model, a multi-model cross-validation framework is established to analyze the source characteristics of VOCs components, calculate the contribution rate of each VOCs component to different pollution sources, and identify the main pollution source types.

[0019] By using the characteristic species ratio method and pollution source indicator analysis, we can identify the spatiotemporal evolution patterns of VOCs emission and construct a spatiotemporal distribution characteristic index system for VOCs, including component concentration change rate, spatial aggregation degree, temporal fluctuation index and source contribution index.

[0020] Furthermore, the method for deploying the VOCs monitoring point network includes:

[0021] The monitoring area was divided using a hierarchical design and spatial balance principle;

[0022] By applying the optimal spatial coverage algorithm in spatial statistics, a network of VOCs monitoring points is deployed in the target area.

[0023] Furthermore, the time gradient sampling scheme includes:

[0024] Weekly gradient: Sampling is conducted for 3 days per week, consisting of 2 weekdays and 1 rest day;

[0025] Sampling was conducted four times a day: morning, noon, evening, and night.

[0026] Special periods: During seasonal production peaks and periods of heavy pollution, increase the frequency of sampling.

[0027] Furthermore, the method for component analysis includes:

[0028] Sample pretreatment: The gas sample in the sampling vessel is pre-concentrated using a thermal desorption instrument;

[0029] Chromatographic separation: Gas chromatograph equipped with capillary column, temperature program executed;

[0030] Mass spectrometry detection: A mass spectrometer is used with an electron impact ionization source to acquire a full scan mass spectrum;

[0031] Data processing: Data processing was performed using chromatography workstation software. Compounds were identified by comparison with the NIST mass spectrometry library, and quantitative analysis was performed using the internal standard method.

[0032] Furthermore, the method for screening and analyzing data using multiple quality control standards includes:

[0033] VOCs components with a detection rate of less than 50% were screened out;

[0034] For components that are not detected, the concentration is calculated as half of the detection limit;

[0035] Samples that are outside the linear range should be diluted before component analysis.

[0036] Furthermore, the method for generating a high-precision spatiotemporal distribution map includes:

[0037] Data preparation: Import the coordinate information of the monitoring points and VOCs concentration data into the geographic information system software;

[0038] Parameter optimization: The optimal distance power p, terrain adjustment parameter β, and wind field adjustment parameter γ were determined by cross-validation.

[0039] Gridding: The study area is divided into a regular grid of 100m×100m, which serves as the basic unit for interpolation calculation;

[0040] Interpolation calculation: Apply an optimized inverse distance weighted interpolation algorithm to each grid center point to calculate...

[0041] Estimated VOCs concentrations;

[0042] The optimization of the inverse distance weighted interpolation algorithm is as follows: Topographic factors and wind field factors are introduced, as expressed below:

[0043]

[0044] Where Z(s0) represents the estimated VOCs concentration at the interpolation point s0; Z(s i ) represents a known monitoring point s i Measured VOCs concentration at the location; w i (s0) represents the monitoring point s i The weight of the interpolation point s0 is calculated using the following formula:

[0045]

[0046] Where, d i0 Indicates monitoring point s i Euclidean distance to the point s0 to be interpolated; d j0 Indicates monitoring point s j The Euclidean distance to the interpolation point s0; p is the distance power; n is the total number of monitoring points; T i T j monitoring point s i and monitoring points s j Topographical influence factors at the location; W i W j monitoring point s i and monitoring points s j The wind field influence factor at a given location is calculated using the following formula:

[0047] T i =exp(-β·|h i -h0|);

[0048] W i =exp(-γ·cos(θ) i -θ w )·v w );

[0049] Among them, h i h0 and h0 represent monitoring points s and s, respectively. iThe elevation of the point s0 to be interpolated; β is the terrain adjustment parameter; θ i Indicates from monitoring point s i The direction angle to the interpolation point s0; θ w Indicates wind direction angle; v w Indicates wind speed; γ is the wind field regulation parameter;

[0050] Visual representation: Based on the interpolation results, generate visualization results such as VOCs concentration isosurface maps and three-dimensional surface maps.

[0051] Furthermore, the principal component analysis is applied to preliminary source analysis, and the method includes:

[0052] Data preprocessing: Standardize the VOCs concentration data to eliminate the influence of different dimensions;

[0053] Principal component extraction: Based on the correlation matrix, eigenvalues ​​and eigenvectors are calculated, and principal components are extracted according to the Kaiser criterion;

[0054] Factor rotation: Orthogonal rotation is performed using the maximum variance method to make factor loadings easier to interpret;

[0055] Source type identification: Based on the characteristic species with high loading in each principal component, combined with the characteristics of VOC emission source spectrum, the pollution source type is identified.

[0056] Furthermore, the positive definite matrix factorization model is applied to:

[0057] Data preparation: Constructing the concentration matrix and uncertainty matrix;

[0058] Setting the number of factors: Initially try 3-8 factors, and determine the optimal number of factors by analyzing the objective function value, residual distribution and the physical meaning of the factor interpretation;

[0059] Model execution: The model was run multiple times using PMF software, and the solution with the minimum objective function value and reasonable physical interpretation was selected.

[0060] Source spectrum identification: By comparing the characteristic species composition of each factor with known source spectra, the type of pollution source represented by the factor can be determined.

[0061] Furthermore, the formulas for calculating the component concentration change rate, spatial aggregation degree, temporal fluctuation index, and source contribution index are as follows:

[0062] Component concentration change rate:

[0063]

[0064] Among them, CVR i C represents the rate of change in the concentration of the i-th VOC component; i,t and Ci,t-1 Let represent the concentrations of the i-th VOC component at time t and time t-1, respectively;

[0065] Spatial clustering:

[0066]

[0067] Wherein, SCI represents the spatial aggregation index of VOCs; C i C represents the VOCs concentration at the i-th monitoring point; min and C max These represent the minimum and maximum VOCs concentrations at all monitoring points, respectively.

[0068] Time Fluctuation Index:

[0069]

[0070] Where TFI represents the time fluctuation index of VOCs; σ t This represents the standard deviation of VOCs concentration within a time series. This represents the average concentration of VOCs over a time series.

[0071] Source Contribution Index:

[0072]

[0073] Among them, SCI j S represents the contribution index of the j-th pollution source; j denoted by j, representing the contribution of the j-th pollution source; p represents the total number of pollution sources.

[0074] Ozone formation contribution rate:

[0075]

[0076] Among them, OFPR i OFP represents the contribution rate of the i-th VOC component to ozone formation; i denoted as the ozone generation potential of the i-th VOC component; m is the total number of VOC components.

[0077] A VOCs spatiotemporal distribution pattern analysis system, which is used to perform any of the above-described methods.

[0078] Methods for analyzing the spatiotemporal distribution patterns of VOCs include:

[0079] The monitoring point determination module is used to construct a hierarchical spatial distribution strategy and deploy a network of VOCs monitoring points.

[0080] The multi-period sampling module is used to design time gradient sampling schemes and execute multi-period sampling.

[0081] The component analysis module is used for component analysis using pre-concentration-gas chromatography-mass spectrometry technology;

[0082] The spatiotemporal distribution map generation module is used to construct a spatial distribution model of VOCs concentration using an optimized inverse distance weighted interpolation algorithm and generate a spatiotemporal distribution map.

[0083] The main pollution source type identification module is used to integrate a multi-model source apportionment framework to analyze the sources of VOCs components and identify the main pollution source types.

[0084] The module for constructing a spatiotemporal distribution characteristic index system is used to identify VOCs emission patterns and spatiotemporal evolution laws, and to construct a spatiotemporal distribution characteristic index system for VOCs.

[0085] The beneficial effects of this invention are as follows:

[0086] Based on the topography, meteorological conditions, pollution source distribution, and sensitive receptor distribution characteristics of the target area, a systematic spatial deployment strategy is adopted to ensure that monitoring points cover representative polluted areas and background areas, thereby improving the scientific validity and representativeness of monitoring data from the source.

[0087] By optimizing the sampling time series and frequency arrangement, and combining it with high-precision component analysis technology, a comprehensive analytical method covering multiple categories of VOCs components such as alkanes, alkenes, and aromatic hydrocarbons has been established, which improves the accuracy and completeness of component analysis, especially the ability to identify key components at low concentrations.

[0088] To address the differences in spatial transport patterns among different types of VOCs components, an optimized spatial interpolation algorithm was introduced. Combined with high-precision geographic information system technology, this enabled high-precision spatial distribution simulation of VOCs concentrations, significantly improving the accuracy and reliability of spatial distribution maps.

[0089] It integrates multiple source analysis methods such as multivariate statistical models, positive definite matrix factorization models, and characteristic species ratio methods, and establishes a comprehensive evaluation framework. Through model cross-validation, it eliminates the limitations of single models, improves the accuracy and reliability of pollution source identification, and provides a scientific basis for precise source tracing.

[0090] Based on the spatiotemporal analysis results, a set of VOCs spatiotemporal distribution characteristic index system was constructed, including component concentration change rate, spatial aggregation degree, temporal fluctuation index and source contribution index, which provides a quantitative assessment tool for regional VOCs pollution prevention and control. Attached Figure Description

[0091] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0092] Figure 1 This is a schematic diagram of the implementation steps of the method of the present invention. Detailed Implementation

[0093] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0094] Example 1

[0095] according to Figure 1 As shown, this invention provides a method for analyzing the spatiotemporal distribution patterns of VOCs, comprising the following steps:

[0096] Step 100: Construct a hierarchical spatial deployment strategy and deploy a network of VOCs monitoring points;

[0097] In this embodiment, the spatiotemporal distribution patterns of VOCs in an industrial park and a surrounding area within a 10-kilometer radius are analyzed. First, based on the topographic features, meteorological parameters, pollution source intensity distribution, and sensitive receptor distribution data of the target area, a hierarchical spatial sampling strategy is constructed to determine representative monitoring areas.

[0098] In practice, the first step is to collect topographic data of the target area, including information such as altitude, slope, and aspect; obtain meteorological condition data for the past five years, including parameters such as wind speed, wind direction, temperature, humidity, and atmospheric stability; determine the spatial distribution and intensity of VOCs emission sources in the area through emission inventory surveys or pollution source census data; and collect location data of sensitive receptors such as population distribution, schools, and hospitals.

[0099] Based on the above data, and using a hierarchical design and spatial balance principle, the monitoring area is divided into the following categories:

[0100] (1) High emission areas from industrial sources: Set up monitoring points in the areas with the highest VOCs emission intensity in industrial parks;

[0101] (2) Sensitive points in residential areas: Set up monitoring points in densely populated surrounding areas;

[0102] (3) Traffic source impact area: Set up monitoring points near major traffic arteries;

[0103] (4) Upwind reference point: Set up a background point upwind of the industrial park according to the prevailing wind direction;

[0104] (5) Downwind impact points: Set up monitoring points in the potential impact area downwind of the industrial park.

[0105] By applying the optimal spatial coverage algorithm from spatial statistics, 15 VOCs monitoring points were ultimately deployed within the target area. These included 5 points in high-emission industrial areas, 4 points in sensitive residential areas, 3 points in traffic-affected areas, 1 upwind control point, and 2 downwind affected points. Precise coordinates were recorded at each monitoring point using GPS positioning devices to ensure representativeness and balanced spatial coverage.

[0106] Step 200: Design a time gradient sampling scheme and perform multi-time period sampling;

[0107] Based on the climate characteristics and emission activity patterns of the target area, a temporal gradient sampling scheme was designed. In this embodiment, the sampling work covers four seasons: spring, summer, autumn, and winter. A typical month was selected for sampling in each season, namely April, July, October, and January.

[0108] Within each sampling month, sampling is conducted according to the following time gradient:

[0109] (1) Weekly gradient: Sampling is conducted for 3 days per week, consisting of 2 weekdays and 1 rest day;

[0110] (2) Daily gradient: Sampling was conducted 4 times a day, namely in the morning (7:00-9:00), noon (12:00-14:00), evening (17:00-19:00) and night (22:00-24:00);

[0111] (3) Special periods: Increase sampling frequency for special emission activities (such as seasonal production peaks, heavy pollution weather, etc.).

[0112] During the sampling process, a calibrated portable vacuum sampling container was used to collect VOCs samples, with the sampling flow rate controlled at 100 ml / min and each sampling time lasting 1 hour. Simultaneously, a portable weather station was used to record meteorological parameters during the sampling period, including temperature, humidity, wind speed, and wind direction, as well as information on surrounding emission activities, such as traffic flow and industrial production status.

[0113] After sampling, each sampling container was numbered and sealed, and information such as sampling time, location, and meteorological conditions was recorded. The containers were then sent to the laboratory for analysis within 72 hours. The entire sampling process was strictly carried out in accordance with the "Technical Specification for Manual Monitoring of Ambient Air Quality" (HJ194-2017) to ensure the representativeness and accuracy of the samples.

[0114] Step 300: Component analysis is performed using pre-concentration-gas chromatography-mass spectrometry (GC-MS).

[0115] The collected VOCs samples were analyzed using pre-concentration gas chromatography-mass spectrometry (TD-GC / MS). Before analysis, the instrument was calibrated using a standard gas mixture (containing 56 VOCs components) at five points to ensure the linear range and accuracy of the detection.

[0116] The specific analysis process is as follows:

[0117] (1) Sample pretreatment: The gas sample in the sampling vessel is pre-concentrated by a thermal desorption instrument (such as Markes TD100). The sample volume is 400 ml, the desorption temperature is 300 °C, and the holding time is 10 minutes.

[0118] (2) Chromatographic separation: A gas chromatograph (such as Agilent 7890B) was used, equipped with a DB-624 capillary column (60m×0.25mm×1.4μm). The temperature program was as follows: initial temperature 35℃, hold for 5 minutes, increase to 180℃ at 5℃ / min, hold for 5 minutes, and the total running time was 40 minutes.

[0119] (3) Mass spectrometry detection: A mass spectrometer (such as Agilent 5977A) was used with an electron impact ionization source (EI) at an electron energy of 70 eV and a scanning range of m / z 35-300 to obtain a full scan mass spectrum;

[0120] (4) Data processing: Data processing was performed using chromatography workstation software. The compounds were identified by comparison with the NIST mass spectrometry library, and the internal standard method (using deuterated toluene as an internal standard) was used for quantitative analysis.

[0121] Multiple quality control standards were applied to screen and analyze the data, including:

[0122] (1) Screen out VOCs components with a detection rate of less than 50%;

[0123] (2) For components that are not detected, the concentration shall be calculated as half of the detection limit;

[0124] (3) Dilute and reanalyze samples that are outside the linear range;

[0125] (4) Each batch of sample analysis includes laboratory blank, transport blank, parallel sample and spike recovery test to ensure the quality of analysis.

[0126] Based on the above analysis, the VOCs composition spectrum of the target area was determined, and a total of 92 VOCs compounds were detected, mainly including: alkanes (such as n-hexane, cyclohexane, etc.), alkenes (such as ethylene, propylene, etc.), aromatic hydrocarbons (such as benzene, toluene, xylene, etc.), oxygen-containing VOCs (such as aldehydes, ketones, alcohols, esters, etc.), and halogenated hydrocarbons. A VOCs composition concentration matrix was established to record the VOCs composition concentrations at different monitoring points at different time periods, providing basic data for subsequent spatiotemporal distribution analysis.

[0127] Step 400: Use the optimized inverse distance weighted interpolation algorithm to construct a spatial distribution model of VOCs concentration and generate a spatiotemporal distribution map;

[0128] Based on the geospatial coordinates of monitoring points and the VOCs concentration characteristics of each point, combined with geographic information system technology, an optimized inverse distance weighted interpolation algorithm is used to construct a spatial distribution model of VOCs concentration and generate a spatiotemporal distribution map.

[0129] The basic principle of the Inverse Distance Weighted Interpolation (IDW) algorithm is that the attribute value of an unknown point is determined by the weighted average of the surrounding known points, with the weight inversely proportional to the distance. In this embodiment, the traditional IDW algorithm is optimized by introducing terrain and wind field factors to improve interpolation accuracy.

[0130] The optimized inverse distance weight interpolation algorithm is expressed as follows:

[0131]

[0132] Where Z(s0) represents the estimated VOCs concentration at the interpolation point s0; Z(s i ) represents a known monitoring point s i Measured VOCs concentration at the location; w i (s0) represents the monitoring point s i The weight of the interpolation point s0 is calculated using the following formula:

[0133]

[0134] Where, d i0 Indicates monitoring point s i Euclidean distance to the point s0 to be interpolated; d j0 Indicates monitoring point s j The Euclidean distance to the interpolation point s0; p is the power of the distance, usually taken as 2; T i T j monitoring point s i and monitoring points s jTopographical factors at the location, taking into account the impact of elevation differences on pollutant diffusion; W i W j monitoring point s i and monitoring points s j The wind field influence factor at a given location, considering the impact of wind direction and speed on pollutant transport, is calculated using the following formula:

[0135] T i =exp(-β·|h i -h0|);

[0136] W i =exp(-γ·cos(θ) i -θ w )·v w );

[0137] Among them, h i h0 and h0 represent monitoring points s and s, respectively. i The elevation of the point s0 to be interpolated; β is the terrain adjustment parameter; θ i Indicates from monitoring point s i The direction angle to the interpolation point s0; θ w Indicates wind direction angle; v w γ represents wind speed; γ is the wind field regulation parameter.

[0138] The interpolation process is as follows:

[0139] (1) Data preparation: Import the coordinate information of the monitoring points and VOCs concentration data into geographic information system software (such as ArcGIS);

[0140] (2) Parameter optimization: The optimal distance power p, terrain adjustment parameter β, and wind field adjustment parameter γ were determined by cross-validation.

[0141] (3) Gridding: The study area is divided into a regular grid of 100m×100m as the basic unit for interpolation calculation;

[0142] (4) Interpolation calculation: Apply the optimized IDW algorithm to each grid center point to calculate the estimated VOCs concentration;

[0143] (5) Visualization: Based on the interpolation results, generate visualization results such as VOCs concentration isosurface maps and three-dimensional surface maps.

[0144] Using the above interpolation method, spatial distribution maps of total VOCs concentration and major components (such as BTEX, alkanes, alkenes, etc.) for different seasons and time periods are generated, forming a complete set of spatiotemporal distribution maps of VOCs concentration, which intuitively shows the spatiotemporal variation patterns of VOCs pollution.

[0145] Step 500: Integrate the multi-model source apportionment framework to calculate the contribution rate of each VOCs component to different pollution sources, analyze the sources of VOCs components, and identify the main pollution source types.

[0146] By integrating principal component analysis (PCA) and positive definite matrix factorization (PMF) models, a multi-model cross-validation framework was established to analyze the source characteristics of VOCs components, calculate the contribution rate of each VOCs component to different pollution sources, identify the main pollution source types, and quantify the spatiotemporal contribution values ​​of each pollution source.

[0147] First, principal component analysis (PCA) is applied for preliminary source analysis:

[0148] (1) Data preprocessing: Standardize the VOCs concentration data to eliminate the influence of different dimensions;

[0149] (2) Principal component extraction: Based on the correlation matrix, calculate the eigenvalues ​​and eigenvectors, and extract the principal components according to the Kaiser criterion (eigenvalues ​​> 1);

[0150] (3) Factor rotation: Orthogonal rotation is performed using the Varimax method to make the factor loadings easier to interpret;

[0151] (4) Source type identification: Based on the characteristic species with high loading in each principal component, combined with the characteristics of VOC emission source spectrum, the pollution source type is identified.

[0152] Secondly, a positive definite matrix factorization (PMF) model is applied for in-depth source analysis:

[0153] The basic principle of the PMF model is to decompose the observed data matrix into the product of two non-negative matrices: the factor contribution matrix and the factor spectrum matrix. The PMF model can be represented as:

[0154] X = G × F + E;

[0155] Where X is an n×m dimensional concentration matrix, representing the concentration of m VOCs components in n samples; G is an n×p dimensional factor contribution matrix, representing the contribution of p pollution sources to n samples; F is a p×m dimensional factor spectrum matrix, representing the relative content of m VOCs components in p pollution sources; and E is an n×m dimensional residual matrix.

[0156] The PMF model optimizes the decomposition results by minimizing the objective function Q:

[0157]

[0158] Among them, e ij σ is an element in the residual matrix E, representing the difference between the model-fitted values ​​and the observed values; ij This represents the corresponding uncertainty.

[0159] In one embodiment of the present invention, the PMF model implementation process is as follows:

[0160] (1) Data preparation: Construct the concentration matrix X and the uncertainty matrix σ;

[0161] (2) Setting the number of factors: Initially try 3-8 factors. By analyzing the Q value, residual distribution and the physical meaning of factor interpretation, determine that the optimal number of factors is 5.

[0162] (3) Model running: Run the model multiple times (≥20 times) using PMF software (such as EPA-PMF 5.0) and select the solution with the smallest Q value and reasonable physical interpretation;

[0163] (4) Source spectrum identification: By comparing the characteristic species composition of each factor with the known source spectrum, the type of pollution source represented by the factor is determined;

[0164] (5) Uncertainty analysis: The stability and uncertainty of the results are evaluated by methods such as Bootstrap, permutation and DISP.

[0165] Finally, a multi-model cross-validation framework was established to integrate the analytical results of PCA and PMF:

[0166] (1) Results comparison: Analyze the consistency of the pollution source types and contribution rates identified by the two models;

[0167] (2) Cross-validation: For results that differ, the characteristic species ratio method is used for validation;

[0168] (3) Results fusion: Based on the validation results, the contribution rate estimates of the two models are integrated using the weighted average method.

[0169] It should be noted that the main pollution source types are obtained by comparing the estimated contribution rate with a set threshold.

[0170] Through the above source apportionment process, five major sources of VOCs in the target area were identified: industrial process emissions, solvent use, vehicle exhaust, fuel volatilization, and biomass combustion. The contribution rate of each pollution source in different seasons and regions was quantified, providing a scientific basis for VOCs pollution control.

[0171] Step 600: Construct a spatiotemporal distribution characteristic index system for VOCs.

[0172] By using the characteristic species ratio method and pollution source indicator analysis, we can identify the spatiotemporal evolution patterns of VOCs emission and construct a spatiotemporal distribution characteristic index system for VOCs, including component concentration change rate, spatial aggregation degree, temporal fluctuation index and source contribution index.

[0173] First, the characteristic species ratio method was used to analyze the characteristics of VOC emissions:

[0174] (1) Benzene / Toluene (B / T) ratio: B / T≈0.5 indicates that the main source is transportation, and B / T>1 indicates that coal or biomass combustion contributes significantly;

[0175] (2) Toluene / ethylbenzene (T / E) ratio: T / E≈3 indicates that the emission source is relatively fresh, and T / E<3 indicates that the pollutants have undergone a longer period of photochemical reaction;

[0176] (3) hexane / 2,2-dimethylbutane (nHex / 22DMB) ratio: A high ratio indicates a significant contribution from gasoline evaporation;

[0177] (4) Isopentane / n-pentane (i-C5 / n-C5) ratio: A ratio close to 1 indicates that vehicle exhaust emissions are dominant, while a ratio >3 indicates that gasoline evaporation has a significant impact.

[0178] Secondly, the contribution of VOCs to ozone formation is analyzed by combining ozone formation potential (OFP) calculations:

[0179] OFP i =C i ×MIR i ;

[0180] Among them, OFP i This represents the ozone formation potential (μg / m³) of the i-th VOC component. 3 );C i Represents the mass concentration (μg / m³) of the i-th VOC component. 3 MIR i The maximum incremental reactivity coefficient (g O3 / g VOC) of the i-th VOC component is represented by the MIR value proposed by Carter.

[0181] Based on the above analysis results, a spatiotemporal distribution characteristic index system for VOCs is constructed:

[0182] Component concentration change rate (CVR):

[0183]

[0184] Among them, CVR i C represents the rate of change in the concentration of the i-th VOC component; i,t and C i,t-1 Let represent the concentrations of the i-th VOC component at time t and time t-1, respectively.

[0185] Spatial Clustering (SCI):

[0186]

[0187] Wherein, SCI represents the spatial aggregation index of VOCs; C i C represents the VOCs concentration at the i-th monitoring point; min and C max These represent the minimum and maximum VOCs concentrations at all monitoring points, respectively; n is the total number of monitoring points.

[0188] Time Fluctuation Index (TFI):

[0189]

[0190] Where TFI represents the time fluctuation index of VOCs; σ t This represents the standard deviation of VOCs concentration within a time series. This represents the average concentration of VOCs within a time series.

[0191] Source Contribution Index (SCI):

[0192]

[0193] Among them, SCI j S represents the contribution index of the j-th pollution source; j Let represent the contribution of the j-th pollution source; p is the total number of pollution sources.

[0194] Ozone formation contribution rate (OFPR):

[0195]

[0196] Among them, OFPR i OFP represents the contribution rate of the i-th VOC component to ozone formation; i denoted as the ozone generation potential of the i-th VOC component; m is the total number of VOC components.

[0197] The spatiotemporal distribution characteristics of VOCs pollution in the target area are systematically evaluated using the above indicator system:

[0198] (1) Spatial distribution characteristics: Identify high-value areas and low-value areas of VOCs and their distribution patterns, and analyze their relationship with emission sources and geographical environment;

[0199] (2) Temporal variation characteristics: Analyze the diurnal, weekly and seasonal variation patterns of VOCs concentrations to reveal the influencing factors;

[0200] (3) Source contribution characteristics: Clarify the magnitude and spatiotemporal distribution of the contribution of different pollution sources to VOCs, so as to provide a basis for pollution control;

[0201] (4) Environmental impact characteristics: assess the contribution of VOCs to ozone formation, identify key active components, and provide support for synergistic control.

[0202] Example 2

[0203] The difference from Example 1 lies in the monitoring point deployment strategy and sampling scheme design.

[0204] In this embodiment, the monitoring area is a certain urban cluster, covering core cities and satellite cities, with a total area of ​​approximately 3,000 square kilometers. The deployment of monitoring points adopts a hierarchical and categorized strategy:

[0205] (1) First-level zoning: Based on administrative divisions and urban functions, the entire urban agglomeration is divided into five categories: core urban area, industrial area, commercial area, residential area and background area;

[0206] (2) Two-level stratification: Within each functional zone, high, medium and low levels are divided according to the intensity of VOCs emission sources;

[0207] (3) Location layout: The grid method and representative sampling method are combined to lay out the basic monitoring grid at a density of 1 point / 25 square kilometers, and the number of points is increased in VOCs emission hotspot areas.

[0208] Regarding the sampling scheme, this embodiment adopts a stratified sampling strategy different from that in Embodiment 1:

[0209] (1) Routine sampling: Sampling is conducted once per quarter, with each sampling lasting for 3 consecutive days and 4 time periods per day;

[0210] (2) Increased sampling frequency: For areas with high VOCs concentrations and large VOCs emissions, an additional sampling will be conducted every month;

[0211] (3) Vertical sampling: Vertical sampling points were set up in typical areas, and samples were taken at ground level, 10 meters and 30 meters above ground to analyze the vertical distribution characteristics of VOCs.

[0212] The remaining steps are basically the same as in Example 1. Through systematic VOCs monitoring, analysis and simulation, a spatiotemporal distribution characteristic map of VOCs pollution in urban agglomerations is constructed to provide a scientific basis for regional collaborative control.

[0213] The above description is merely an example and illustration of the concept of the present invention. Those skilled in the art can make various modifications or additions to the specific embodiments described or use similar methods to replace them, as long as they do not deviate from the concept of the invention or exceed the scope defined in this specification, they should all fall within the protection scope of the present invention.

Claims

1. A method for analyzing the spatiotemporal distribution patterns of VOCs, characterized in that, include: A hierarchical spatial deployment strategy was constructed. Based on the topographic features, meteorological parameters, pollution source intensity distribution and sensitive receptor distribution data of the target area, representative monitoring areas were determined, and a VOCs monitoring point network was deployed in the target area and its surrounding environment. Design a time gradient sampling scheme to perform multi-season, multi-time period, and multi-frequency sampling at each of the deployed monitoring points, collect VOCs samples, and record meteorological parameters and surrounding emission activities during the sampling period; The collected VOCs samples were analyzed using pre-concentration-gas chromatography-mass spectrometry. Multiple quality control standards were applied to screen the analytical data, a VOCs component concentration matrix was established, and the VOCs component spectrum and concentration distribution characteristics of the target area were determined. Based on the geospatial coordinates of monitoring points and the VOCs concentration characteristics of each point, combined with geographic information system technology, an optimized inverse distance weighted interpolation algorithm is used to construct a spatial distribution model of VOCs concentration and generate a spatiotemporal distribution map. Methods for generating spatiotemporal distribution maps include: Data preparation: Import the coordinate information of the monitoring points and VOCs concentration data into the geographic information system software; Parameter optimization: Determine the optimal distance power using cross-validation. Terrain adjustment parameters Wind field regulation parameters ; Gridding: The study area is divided into a regular grid of 100m×100m, which serves as the basic unit for interpolation calculation; Interpolation calculation: Apply an optimized inverse distance weighted interpolation algorithm to each grid center point to calculate the estimated VOCs concentration; The optimization of the inverse distance weighted interpolation algorithm is as follows: Topographic factors and wind field factors are introduced, as expressed below: ； in, Indicates the point to be interpolated Estimated VOCs concentration at the location; Indicates known monitoring points Measured VOCs concentration at the location; Indicates monitoring point For interpolation points The weight is calculated using the following formula: ； in, Indicates monitoring point to the interpolation point The Euclidean distance; Indicates monitoring point to the interpolation point The Euclidean distance; The distance is a power of the power; This represents the total number of monitoring points. , monitoring points and monitoring points Topographical factors at the location; , monitoring points and monitoring points The wind field influence factor at a given location is calculated using the following formula: ； ； in, and They represent monitoring points respectively. and interpolation points Elevation; For terrain adjustment parameters; Indicates from monitoring point to the interpolation point Direction angle; Indicates the wind direction angle; Indicates wind speed; These are the wind field regulation parameters; Visual representation: Based on the interpolation results, generate visualization results such as VOCs concentration isosurface maps and three-dimensional surface maps; By integrating principal component analysis and positive definite matrix factorization model, a multi-model cross-validation framework is established to analyze the source characteristics of VOCs components, calculate the contribution rate of each VOCs component to different pollution sources, and identify the main pollution source types. By using the characteristic species ratio method and pollution source indicator analysis, we can identify the spatiotemporal evolution patterns of VOCs emission and construct a spatiotemporal distribution characteristic index system for VOCs, including component concentration change rate, spatial aggregation degree, temporal fluctuation index and source contribution index.

2. The method for analyzing the spatiotemporal distribution of VOCs according to claim 1, characterized in that, The method for deploying a VOCs monitoring network includes: The monitoring area was divided using a hierarchical design and spatial balance principle; By applying the optimal spatial coverage algorithm in spatial statistics, a network of VOCs monitoring points is deployed in the target area.

3. The method for analyzing the spatiotemporal distribution of VOCs according to claim 2, characterized in that, The time gradient sampling scheme includes: Weekly gradient: Sampling is conducted for 3 days per week, consisting of 2 weekdays and 1 rest day; Sampling was conducted four times a day: morning, noon, evening, and night. Special periods: During seasonal production peaks and periods of heavy pollution, increase the frequency of sampling.

4. The method for analyzing the spatiotemporal distribution of VOCs according to claim 3, characterized in that, The method for component analysis includes: Sample pretreatment: The gas sample in the sampling vessel is pre-concentrated using a thermal desorption instrument; Chromatographic separation: Gas chromatograph equipped with capillary column, temperature program executed; Mass spectrometry detection: A mass spectrometer is used with an electron impact ionization source to acquire a full scan mass spectrum; Data processing: Data processing was performed using chromatography workstation software. Compounds were identified by comparison with the NIST mass spectrometry library, and quantitative analysis was performed using the internal standard method.

5. The method for analyzing the spatiotemporal distribution of VOCs according to claim 4, characterized in that, The method for screening and analyzing data using multiple quality control standards includes: VOCs components with a detection rate of less than 50% are screened out; For components that are not detected, the concentration is calculated as half of the detection limit; Samples that are outside the linear range should be diluted before component analysis.

6. The method for analyzing the spatiotemporal distribution of VOCs according to claim 5, characterized in that, The principal component analysis is applied to preliminary source analysis, and the method includes: Data preprocessing: Standardize the VOCs concentration data to eliminate the influence of different dimensions; Principal component extraction: Based on the correlation matrix, eigenvalues ​​and eigenvectors are calculated, and principal components are extracted according to the Kaiser criterion; Factor rotation: Orthogonal rotation is performed using the maximum variance method to make factor loadings easier to interpret; Source type identification: Based on the characteristic species with high loading in each principal component, combined with the characteristics of VOC emission source spectrum, the pollution source type is identified.

7. The method for analyzing the spatiotemporal distribution of VOCs according to claim 6, characterized in that, The positive definite matrix factorization model is applied to: Data preparation: Constructing the concentration matrix and uncertainty matrix; Setting the number of factors: Initially try 3-8 factors, and determine the optimal number of factors by analyzing the objective function value, residual distribution and the physical meaning of the factor interpretation; Model execution: The model was run multiple times using PMF software, and the solution with the minimum objective function value and reasonable physical interpretation was selected. Source spectrum identification: By comparing the characteristic species composition of each factor with known source spectra, the type of pollution source represented by the factor can be determined.

8. The method for analyzing the spatiotemporal distribution of VOCs according to claim 7, characterized in that, The formulas for calculating the component concentration change rate, spatial aggregation degree, temporal fluctuation index, and source contribution index are as follows: Component concentration change rate: ； in, Indicates the first The rate of change in concentration of each VOC component; and They represent the first VOCs components in Time and Concentration at any given time; Spatial clustering: ； in, Indicates the spatial concentration index of VOCs; Indicates the first VOCs concentration at each monitoring point; and These represent the minimum and maximum VOCs concentrations at all monitoring points, respectively. Time Fluctuation Index: ； in, This represents the time fluctuation index of VOCs; This represents the standard deviation of VOCs concentration within a time series. This represents the average concentration of VOCs within a time series. Source Contribution Index: ； in, Indicates the first Contribution index of various pollution sources; Indicates the first The contribution of each pollution source; The total number of pollution sources; Ozone formation contribution rate: ； in, Indicates the first The contribution rate of each VOC component to ozone formation; Indicates the first Ozone formation potential of various VOC components; This represents the total number of VOC components.

9. A system for analyzing the spatiotemporal distribution patterns of VOCs, characterized in that, It is used to perform a method for analyzing the spatiotemporal distribution of VOCs as described in any one of claims 1-8, comprising: The monitoring point determination module is used to construct a hierarchical spatial distribution strategy and deploy a network of VOCs monitoring points. The multi-period sampling module is used to design time gradient sampling schemes and execute multi-period sampling. The component analysis module is used for component analysis using pre-concentration-gas chromatography-mass spectrometry technology; The spatiotemporal distribution map generation module is used to construct a spatial distribution model of VOCs concentration using an optimized inverse distance weighted interpolation algorithm and generate a spatiotemporal distribution map. The main pollution source type identification module is used to integrate a multi-model source apportionment framework to analyze the sources of VOCs components and identify the main pollution source types. The module for constructing a spatiotemporal distribution characteristic index system is used to identify VOCs emission patterns and spatiotemporal evolution laws, and to construct a spatiotemporal distribution characteristic index system for VOCs.

Images