Atmospheric pollutant three-dimensional monitoring system and method based on multispectral laser radar

Abstract

The invention provides an atmospheric pollutant three-dimensional monitoring system and method based on a multispectral laser radar. The invention discloses an atmospheric pollutant three-dimensional monitoring method based on a multispectral laser radar. The method comprises the following steps: S1, data acquisition; s2, data preprocessing; s3, parameter inversion is carried out; s4, pollutant identification and classification; s5, performing data fusion and three-dimensional modeling; and S6, carrying out real-time analysis and early warning. The atmospheric pollutant three-dimensional monitoring system and method based on the multispectral laser radar have the advantage that high-precision, high-resolution and three-dimensional real-time monitoring of regional atmospheric pollutants can be realized.

Description

Technical Field

[0001] This invention relates to the field of atmospheric environment monitoring technology, and in particular to a three-dimensional monitoring system and method for atmospheric pollutants based on multispectral lidar. Background Technology

[0002] Aerosols, as a significant component of air pollutants, have optical properties (such as aerosol optical depth, AOD) that are crucial for assessing air pollution levels and studying aerosol climate effects. Currently, regional aerosol monitoring primarily relies on satellite remote sensing retrieval technology. Traditional passive satellite remote sensing methods, such as the dark target method, the Deep Blue algorithm, and the multi-angle algorithm, are mostly based on one-dimensional radiative transfer models using the planar parallel approximation and the multiple scattering assumption. These methods have significant limitations: firstly, they lack consideration for remote sensing proximity effects (especially the three-dimensional proximity effect of clouds), leading to a significant decrease in retrieval accuracy in areas surrounding clouds; secondly, their temporal resolution is low (typically only 1-2 passes per day), failing to meet the needs of real-time monitoring; and thirdly, their vertical resolution is insufficient, making it difficult to obtain information on the vertical distribution of pollutants and preventing nighttime monitoring.

[0003] While active lidar technology can provide high-precision vertical profile information, traditional single-wavelength or dual-wavelength lidar has shortcomings in pollutant identification, classification, and quantitative inversion. Moreover, it is usually a fixed-point observation, making it difficult to achieve three-dimensional real-time dynamic monitoring over a wide area.

[0004] Therefore, it is necessary to provide a three-dimensional monitoring system and method for atmospheric pollutants based on multispectral lidar to solve the above-mentioned technical problems. Summary of the Invention

[0005] The technical problem solved by this invention is to provide a three-dimensional air pollutant monitoring system and method based on multispectral lidar that can achieve high-precision, high-resolution, three-dimensional real-time monitoring of regional air pollutants.

[0006] To address the aforementioned technical problems, the present invention provides a three-dimensional monitoring method for atmospheric pollutants based on multispectral lidar, comprising the following steps:

[0007] S1: Data Acquisition: Acquire multispectral lidar air pollution monitoring data;

[0008] S2: Data preprocessing: Background noise reduction, signal averaging, and distance correction are performed on the raw backscattered signal obtained in S1;

[0009] S3: Parameter Inversion: The preprocessed signal is processed using the Fernald inversion algorithm to obtain the vertical distribution of aerosol extinction coefficient and backscattering coefficient;

[0010] S4: Pollutant Identification and Classification: Based on multispectral characteristic parameters, different types of pollutants are identified and classified by calculating the extinction coefficient ratio and depolarization ratio at different wavelengths;

[0011] S5: Data Fusion and 3D Modeling: Integrate monitoring data from mobile monitoring platforms and fixed monitoring station networks, and use spatial interpolation methods to generate a 3D distribution field of air pollutants in the monitoring area;

[0012] S6: Real-time analysis and early warning: Calculate pollutant concentrations based on a three-dimensional distribution field, assess pollution levels, generate early warning information, and display and publish monitoring results in real time through a visualization platform.

[0013] Preferably, the laser wavelengths emitted by the multispectral lidar unit in S1 include 355nm, 532nm and 1064nm, and it can simultaneously receive backscattered signals of these three wavelengths.

[0014] Preferably, the Fernald inversion algorithm in S3 simultaneously utilizes signals at three wavelengths of 355nm, 532nm, and 1064nm for coordinated inversion, thereby improving inversion accuracy.

[0015] Preferably, the pollutant identification and classification in S4 specifically includes: distinguishing between sand and dust by calculating the ratio of the extinction coefficients of 355nm and 532nm wavelengths, determining the particle size distribution by calculating the ratio of the extinction coefficients of 532nm and 1064nm wavelengths, and identifying non-spherical particles by the depolarization ratio.

[0016] Preferably, the spatial interpolation method in S5 employs a three-dimensional Kriging interpolation algorithm, which integrates data from the mobile monitoring platform and the fixed monitoring station network to generate a continuous three-dimensional pollutant distribution field.

[0017] Preferably, the visualization platform in S6 adopts WebGIS technology to realize the three-dimensional dynamic visualization of monitoring results and support the reproduction and analysis of the spatiotemporal evolution process of pollutants.

[0018] Preferably, it also includes: S7: Monitoring result verification and optimization: The pollutant concentration and distribution data obtained by the data processing center are compared with the synchronous observation data of the ground-based solar photometer and the ground-based reference lidar. The accuracy of the monitoring results is verified based on the comparison results, and the parameters of the inversion algorithm are corrected and the system is optimized based on the verification results.

[0019] This invention also provides a three-dimensional monitoring system for atmospheric pollutants based on multispectral lidar, comprising:

[0020] A multispectral lidar unit is used to emit laser pulses with wavelengths of at least 355nm, 532nm and 1064nm, and to receive atmospheric backscattered signals.

[0021] A mobile monitoring platform is used to carry the multispectral lidar unit and is equipped with a positioning module and an environmental parameter acquisition module;

[0022] A fixed monitoring station network consists of several fixed monitoring stations deployed within the monitoring area;

[0023] The data processing center includes:

[0024] The data preprocessing module is used to perform background noise reduction, signal averaging, and distance correction on the backscattered signal;

[0025] The parameter inversion module uses the Fernald inversion algorithm to simultaneously process signals at three wavelengths: 355nm, 532nm, and 1064nm, and obtain the vertical distribution of aerosol extinction coefficient and backscattering coefficient.

[0026] The pollutant identification module is used to identify and classify pollutant types by calculating the extinction coefficient ratio and depolarization ratio at different wavelengths;

[0027] The data fusion module uses a three-dimensional Kriging interpolation algorithm to fuse data from mobile monitoring platforms and fixed monitoring station networks to generate a three-dimensional pollutant distribution field.

[0028] The early warning analysis module is used to calculate pollutant concentrations based on a three-dimensional distribution field, assess pollution levels, and generate early warning information.

[0029] The visualization module, using WebGIS technology, is used to achieve three-dimensional dynamic visualization of monitoring results.

[0030] Compared with related technologies, the three-dimensional monitoring system and method for atmospheric pollutants based on multispectral lidar provided by this invention has the following advantages:

[0031] This invention provides a three-dimensional atmospheric pollutant monitoring system and method based on multispectral lidar. By fusing multi-source data from multispectral lidar with mobile platforms and fixed station networks, and combining it with a three-dimensional Kriging interpolation algorithm, it can overcome the limitations of traditional single-point or two-dimensional monitoring. It can acquire a continuous three-dimensional distribution field of atmospheric pollutants from near-ground to high altitudes, with high spatial resolution and fast data update speed, enabling three-dimensional perception of atmospheric pollution. By synergistically utilizing optical information from three wavelengths (355nm, 532nm, and 1064nm), it calculates the absorption index and particle size index, and combines this with depolarization ratio information to effectively distinguish between different types of pollutants such as dust, soot, and secondary aerosols. The system identifies different types of pollutants and reveals their physicochemical properties. The Fernald collaborative inversion algorithm, through angstrom index constraints, effectively transfers information from the 1064nm channel to the visible light channel, reducing dependence on the assumed value of a single-wavelength lidar, minimizing inversion errors, and improving the accuracy and reliability of inversion for the vertical profile of aerosol optical parameters. The system can automatically assess pollution levels based on a three-dimensional pollution field and generate multi-level early warning information according to preset rules. Combined with a WebGIS three-dimensional visualization platform, it enables dynamic and intuitive display of monitoring results and rapid dissemination of early warning information, providing technical support and decision-making basis for precise prevention and control of air pollution and emergency command. Attached Figure Description

[0032] Figure 1 The flowchart shows the three-dimensional monitoring method for atmospheric pollutants based on multispectral lidar provided by the present invention. Detailed Implementation

[0033] The present invention will be further described below with reference to the accompanying drawings and embodiments.

[0034] Please refer to the following: Figure 1 ,in, Figure 1 A flowchart illustrating the three-dimensional monitoring method for atmospheric pollutants based on multispectral lidar provided by this invention. The method includes the following steps:

[0035] S1: Data Acquisition: Acquire atmospheric pollution monitoring data using multispectral lidar.

[0036] Specifically, a vehicle-mounted or fixed multispectral lidar system is used to simultaneously emit laser pulses with wavelengths of 355nm, 532nm, and 1064nm into the atmosphere at specific frequencies (e.g., 10-20Hz). The system's receiving telescope collects the backscattered light signals generated by the interaction of each wavelength of laser with aerosols and molecules in the atmosphere. After being separated by a beam splitter, the signals are converted into electrical signals by photodetectors (e.g., PMT, APD) of the corresponding wavelengths. These signals are then sampled and recorded by a high-speed data acquisition card, ultimately obtaining the original backscattered signal power sequence classified by wavelength. Its value varies with time and altitude (i.e., distance).

[0037] S2: Data preprocessing: Background noise reduction, signal averaging, and distance correction are performed on the raw backscattered signal obtained in S1;

[0038] Background noise subtraction: Data is collected during the no-signal period before / after laser pulse emission, and its average value is calculated as the background noise. Then, this noise value is subtracted from the entire backscattered signal sequence.

[0039] Signal averaging: The backscattered echo signals of multiple continuously transmitted pulses (such as hundreds to thousands of pulses) are superimposed and averaged to improve the signal-to-noise ratio by trading time for time.

[0040] Range correction: The averaged backscattered signal is multiplied by the square of the detection range to eliminate the influence that the echo power is inversely proportional to the square of the range, resulting in a range correction signal used for inversion. Where z is the distance from the radar to the atmospheric scatterer. This is a distance correction signal, which is the wavelength. A function of distance z; The power of the backscattered signal after background noise subtraction and signal averaging is given.

[0041] S3: Parameter Inversion: The preprocessed signal is processed using the Fernald inversion algorithm, which simultaneously utilizes signals at three wavelengths of 355nm, 532nm and 1064nm for coordinated inversion to improve inversion accuracy; and obtains the vertical distribution of aerosol extinction coefficient and backscattering coefficient.

[0042] The steps of the Fernald inversion algorithm are as follows:

[0043] Input preparation: Input the preprocessed distance correction signals at three wavelengths (355nm, 532nm, 1064nm). Combine real-time meteorological sounding data or standard atmospheric models (such as the US Standard Atmosphere) with the calculated atmospheric molecular extinction. and backscattering Profile; preset the lidar ratio for each wavelength according to the aerosol type. (Initial value);

[0044] Boundary condition setting: Select a clean upper atmosphere (usually above the tropopause or aerosol layer, such as 6-8 km) as the calibration altitude. Assuming the aerosol contribution at that location is zero ( Using this calibration height and molecular scattering data, the system constant for each channel is calculated. ;

[0045] Collaborative inversion iteration: Starting from the 1064nm channel: Taking advantage of its low molecular interference, the preliminary profile of the aerosol backscattering coefficient is first independently inverted. ;

[0046] Angstrom index constraint: obtained using inversion Based on the Angstrom formula, the estimated Angstrom exponent is obtained. Initial estimates of aerosol backscattering at 532 nm and 355 nm channels were derived. ;

[0047] Multi-channel joint optimization: Using the estimated values ​​as initial conditions, the Fernald iterative formula for each channel is substituted into the inversion formula, which is:

[0048]

[0049] Through optimization algorithms such as the least squares method, the inversion results of the three wavelengths are made physically consistent (conforming to the angstrom exponential law), and the lidar ratio is optimized. The value of ;

[0050] Iterative loop: Repeat the above steps until the inversion results of each channel converge (the change is less than the set threshold).

[0051] Output: Directly outputs the aerosol backscattering coefficients after the final iteration convergence. Vertical profile; using formulas Calculate and output the aerosol extinction coefficient. Vertical profile.

[0052] By using the angstrom index as a constraint to transfer and guide the inversion of the 532nm and 355nm channels, the over-reliance on the assumption of a single wavelength lidar ratio is effectively reduced, thereby significantly improving the accuracy and reliability of the inversion results.

[0053] S4: Pollutant Identification and Classification: Based on multispectral characteristic parameters, different types of pollutants are identified and classified by calculating the ratio of extinction coefficients and depolarization ratios at different wavelengths. Specifically, the pollutant identification and classification includes: distinguishing between dust and smoke by calculating the ratio of extinction coefficients at wavelengths of 355nm and 532nm; determining the particle size distribution by calculating the ratio of extinction coefficients at wavelengths of 532nm and 1064nm; and identifying non-spherical particles by calculating the depolarization ratio.

[0054] Specifically, this involves calculating key optical indices: for each altitude, using the extinction coefficient obtained from the inversion, calculating two angstrom indices: the absorption index (…). The ratio of the extinction coefficients of 355nm to 532nm ( ),and Non-light-absorbing aerosols are defined as those with a particle size >1.5, while those with a particle size >1.5 are strongly light-absorbing aerosols such as smoke and dust. This distinguishes between light-absorbing (e.g., smoke and dust) and non-light-absorbing aerosols; particle size index (P<0.05) The ratio of the extinction coefficients of 532nm to 1064nm ( ), 1.5-2.5 represents coarse-mode particles, and 1.5-2.5 represents fine-mode particles. The particle size is determined (the larger the value, the smaller the particle size); the depolarization ratio at 532 nm is obtained. ) ,in The power of the received signal component perpendicular to the polarization direction of the emitted laser is denoted as . The power of the received component signal parallel to the polarization direction of the emitted laser directly determines whether the particle is spherical (the higher the value, the greater the degree of non-sphericality). Spherical particles... Non-spherical particles such as sand and dust ;

[0055] Based on the numerical combination of the three optical parameters mentioned above, a simple decision tree is used for rapid classification: First, based on shape (depolarization ratio): if the depolarization ratio is high (e.g., ... If the particle size distribution is high, it is a non-spherical particle (dust); if the deflection ratio is low, it is a spherical particle; then, based on particle size and absorbency: for non-spherical particles, further based on the particle size index: if... Small, coarse-grained sand and dust; if Large particles indicate a mixture of sand / dust; for spherical particles, first consider the absorption index: if... Large, consisting of light-absorbing fine particles (smoke and dust); if Smaller, then based on the particle size index: if Large, non-absorbing fine particles (sulfates, etc.); if Smaller aerosols are mixed aerosols;

[0056] Output: Assign a type label (such as dust, smoke, etc.) to each height point; finally generate a vertical distribution profile of aerosol type, which can be visualized.

[0057] S5: Data Fusion and 3D Modeling: The monitoring data from the mobile monitoring platform and the fixed monitoring station network are fused together, and a 3D distribution field of air pollutants in the monitoring area is generated using a spatial interpolation method. The spatial interpolation method uses the 3D Kriging interpolation algorithm to generate a continuous 3D pollutant distribution field by fusing data from the mobile monitoring platform and the fixed monitoring station network.

[0058] Specifically, the data preparation and integration involves matching the trajectory data of the mobile platform (latitude, longitude, altitude, pollutant concentration obtained from S3 inversion, and type information identified by S4) with the ground concentration data of the fixed station within a unified time window, and integrating all data points into a sample point dataset containing three-dimensional coordinates (longitude, latitude, altitude) and multiple attributes (such as PM2.5 concentration, dust concentration, etc.).

[0059] Spatial structure analysis: Calculate the distance and concentration difference between all data point pairs and plot the experimental variogram; use theoretical models such as the spherical model for fitting to determine three key parameters characterizing spatial correlation: nugget value (representing random error or small-scale variation), sill value (representing total variation), and range (representing the maximum distance of spatial autocorrelation).

[0060] 3D interpolation prediction: The monitoring area is divided into a fine 3D grid; for each grid node, the optimal weight is calculated by solving the Kriging equation system using the known sample points around it, based on the variogram model; the concentration values ​​of the known points are weighted and averaged using these weights to obtain the concentration prediction value of the node; at the same time, the prediction variance of each node is output as a measure of the reliability of the result.

[0061] Generation and output results: After traversing all grid nodes, a continuous three-dimensional distribution data field of pollutants is generated. This data field contains multiple pollutant concentrations, aerosol types (from S4), and uncertainty information for each three-dimensional grid point. The results are visualized and output, such as horizontal slice plots, vertical profile plots, and three-dimensional isosurface plots, to intuitively show the spatial distribution and transport of pollutants.

[0062] S6: Real-time analysis and early warning: Calculate pollutant concentration based on three-dimensional distribution field, assess pollution level, generate early warning information, and display and publish monitoring results in real time through a visualization platform. The visualization platform adopts WebGIS technology to realize three-dimensional dynamic visualization of monitoring results and supports the reproduction and analysis of the spatiotemporal evolution process of pollutants.

[0063] Specifically, the system extracts ground-level and vertical column concentrations (AOD) from a 3D pollution field; maps ground-level concentrations to pollution levels (excellent, good, lightly polluted, etc.) according to national AQI standards and performs spatial statistics; sets early warning triggering rules based on concentration thresholds, spatial range, and duration; automatically determines the early warning level (e.g., yellow, orange, red) and generates structured early warning information including core areas, major pollutants, expected duration, and health recommendations; builds a web-based 3D Earth scene using engines such as Cesium.js; renders the 3D pollution field data into a visual cloud that can be rotated and scaled in 3D space using color mapping and transparency adjustment, integrating meteorological data layers such as wind fields and temperature fields to assist in analyzing pollutant sources and diffusion paths; integrates a timeline control to support playback of the spatial diffusion and evolution of pollutants over time; highlights the early warning area on the 3D map and displays detailed information; pushes early warning information to the government affairs platform via API; and releases real-time visualization and early warning results to the public via web links, mobile apps, or SMS.

[0064] S7: Monitoring Result Verification and Optimization: The pollutant concentration and distribution data obtained from the data processing center are compared with the synchronous observation data of the ground-based solar photometer and the ground-based reference lidar. The accuracy of the monitoring results is verified based on the comparison results, and the parameters of the inversion algorithm are corrected and the system is optimized based on the verification results.

[0065] This invention also provides a three-dimensional monitoring system for atmospheric pollutants based on multispectral lidar, comprising:

[0066] A multispectral lidar unit is used to emit laser pulses with wavelengths of at least 355nm, 532nm and 1064nm, and to receive atmospheric backscattered signals.

[0067] A mobile monitoring platform is used to carry the multispectral lidar unit and is equipped with a positioning module and an environmental parameter acquisition module;

[0068] A fixed monitoring station network consists of several fixed monitoring stations deployed within the monitoring area;

[0069] The data processing center includes:

[0070] The data preprocessing module is used to perform background noise reduction, signal averaging, and distance correction on the backscattered signal;

[0071] The parameter inversion module uses the Fernald inversion algorithm to simultaneously process signals at three wavelengths: 355nm, 532nm, and 1064nm, and obtain the vertical distribution of aerosol extinction coefficient and backscattering coefficient.

[0072] The pollutant identification module is used to identify and classify pollutant types by calculating the extinction coefficient ratio and depolarization ratio at different wavelengths;

[0073] The data fusion module uses a three-dimensional Kriging interpolation algorithm to fuse data from mobile monitoring platforms and fixed monitoring station networks to generate a three-dimensional pollutant distribution field.

[0074] The early warning analysis module is used to calculate pollutant concentrations based on a three-dimensional distribution field, assess pollution levels, and generate early warning information.

[0075] The visualization module, using WebGIS technology, is used to achieve three-dimensional dynamic visualization of monitoring results.

[0076] Compared with related technologies, the three-dimensional monitoring system and method for atmospheric pollutants based on multispectral lidar provided by this invention has the following advantages:

[0077] This invention provides a three-dimensional atmospheric pollutant monitoring system and method based on multispectral lidar. By fusing multi-source data from multispectral lidar with mobile platforms and fixed station networks, and combining it with a three-dimensional Kriging interpolation algorithm, it can overcome the limitations of traditional single-point or two-dimensional monitoring. It can acquire a continuous three-dimensional distribution field of atmospheric pollutants from near-ground to high altitudes, with high spatial resolution and fast data update speed, enabling three-dimensional perception of atmospheric pollution. By synergistically utilizing optical information from three wavelengths (355nm, 532nm, and 1064nm), it calculates the absorption index and particle size index, and combines this with depolarization ratio information to effectively distinguish between different types of pollutants such as dust, soot, and secondary aerosols. The system identifies different types of pollutants and reveals their physicochemical properties. The Fernald collaborative inversion algorithm, through angstrom index constraints, effectively transfers information from the 1064nm channel to the visible light channel, reducing dependence on the assumed value of a single-wavelength lidar, minimizing inversion errors, and improving the accuracy and reliability of inversion for the vertical profile of aerosol optical parameters. The system can automatically assess pollution levels based on a three-dimensional pollution field and generate multi-level early warning information according to preset rules. Combined with a WebGIS three-dimensional visualization platform, it enables dynamic and intuitive display of monitoring results and rapid dissemination of early warning information, providing technical support and decision-making basis for precise prevention and control of air pollution and emergency command.

[0078] The above description is merely an embodiment of the present invention and does not limit the patent scope of the present invention. Any equivalent structural or procedural transformations made based on the content of the present invention specification and drawings, or direct or indirect applications in other related technical fields, are similarly included within the patent protection scope of the present invention.

Claims

1. A method for three-dimensional monitoring of atmospheric pollutants based on multispectral lidar, characterized in that, Includes the following steps: S1: Data Acquisition: Acquire multispectral lidar air pollution monitoring data; S2: Data preprocessing: Background noise reduction, signal averaging, and distance correction are performed on the raw backscattered signal obtained in S1; S3: Parameter Inversion: The preprocessed signal is processed using the Fernald inversion algorithm to obtain the vertical distribution of aerosol extinction coefficient and backscattering coefficient; S4: Pollutant Identification and Classification: Based on multispectral characteristic parameters, different types of pollutants are identified and classified by calculating the extinction coefficient ratio and depolarization ratio at different wavelengths; S5: Data Fusion and 3D Modeling: Integrate monitoring data from mobile monitoring platforms and fixed monitoring station networks, and use spatial interpolation methods to generate a 3D distribution field of air pollutants in the monitoring area; S6: Real-time analysis and early warning: Calculate pollutant concentrations based on a three-dimensional distribution field, assess pollution levels, generate early warning information, and display and publish monitoring results in real time through a visualization platform.

2. The method for three-dimensional monitoring of atmospheric pollutants based on multispectral lidar according to claim 1, characterized in that, The laser wavelengths emitted by the multispectral lidar unit in S1 include 355nm, 532nm and 1064nm, and it can simultaneously receive backscattered signals of these three wavelengths.

3. The method for three-dimensional monitoring of atmospheric pollutants based on multispectral lidar according to claim 2, characterized in that, The Fernald inversion algorithm described in S3 simultaneously utilizes signals at three wavelengths—355nm, 532nm, and 1064nm—for coordinated inversion, thereby improving inversion accuracy.

4. The method for three-dimensional monitoring of atmospheric pollutants based on multispectral lidar according to claim 1, characterized in that, The pollutant identification and classification in S4 specifically includes: distinguishing between dust and smoke by calculating the ratio of the extinction coefficients of 355nm and 532nm wavelengths, determining the particle size distribution by calculating the ratio of the extinction coefficients of 532nm and 1064nm wavelengths, and identifying non-spherical particles by the depolarization ratio.

5. The method for three-dimensional monitoring of atmospheric pollutants based on multispectral lidar according to claim 1, characterized in that, The spatial interpolation method described in S5 employs a three-dimensional Kriging interpolation algorithm, which integrates data from mobile monitoring platforms and fixed monitoring station networks to generate a continuous three-dimensional pollutant distribution field.

6. The method for three-dimensional monitoring of atmospheric pollutants based on multispectral lidar according to claim 1, characterized in that, The visualization platform described in S6 uses WebGIS technology to realize the three-dimensional dynamic visualization of monitoring results and supports the reproduction and analysis of the spatiotemporal evolution of pollutants.

7. The method for three-dimensional monitoring of atmospheric pollutants based on multispectral lidar according to claim 1, characterized in that, Also includes: S7: Monitoring Result Verification and Optimization: The pollutant concentration and distribution data obtained from the data processing center are compared with the synchronous observation data of the ground-based solar photometer and the ground-based reference lidar. The accuracy of the monitoring results is verified based on the comparison results, and the parameters of the inversion algorithm are corrected and the system is optimized based on the verification results.

8. A three-dimensional monitoring system for atmospheric pollutants based on multispectral lidar, characterized in that, include: A multispectral lidar unit is used to emit laser pulses with wavelengths of at least 355nm, 532nm and 1064nm, and to receive atmospheric backscattered signals. A mobile monitoring platform is used to carry the multispectral lidar unit and is equipped with a positioning module and an environmental parameter acquisition module; A fixed monitoring station network consists of several fixed monitoring stations deployed within the monitoring area; The data processing center includes: The data preprocessing module is used to perform background noise reduction, signal averaging, and distance correction on the backscattered signal; The parameter inversion module uses the Fernald inversion algorithm to simultaneously process signals at three wavelengths: 355nm, 532nm, and 1064nm, and obtain the vertical distribution of aerosol extinction coefficient and backscattering coefficient. The pollutant identification module is used to identify and classify pollutant types by calculating the extinction coefficient ratio and depolarization ratio at different wavelengths; The data fusion module uses a three-dimensional Kriging interpolation algorithm to fuse data from mobile monitoring platforms and fixed monitoring station networks to generate a three-dimensional pollutant distribution field. The early warning analysis module is used to calculate pollutant concentrations based on a three-dimensional distribution field, assess pollution levels, and generate early warning information. The visualization module, using WebGIS technology, is used to achieve three-dimensional dynamic visualization of monitoring results.

Images