Aerosol micro-physical property profile determination method, device, equipment and medium

Abstract

The invention provides an aerosol micro-physical property profile determination method, device and equipment and a medium, and the method comprises the steps: obtaining atmospheric whole-layer column average aerosol volume scale spectral distribution and multi-wavelength complex refractive indexes of a plurality of ground sites in different time periods, and obtaining monitoring data including a real part and an imaginary part; constructing a coarse and fine modal aerosol six-parameter distribution database according to the monitoring data, and clustering the database to generate a plurality of coarse and fine modal aerosol scale spectrum distribution models; and each model calculates a preset wavelength backscattering coefficient calculation value based on a Mie scattering theory, performs layer-by-layer iterative comparison on the preset wavelength backscattering coefficient calculation value and corresponding wavelength observation values of different altitudes of the spaceborne laser radar by using a preset cost function, and inverts an aerosol micro-physical property profile. By adopting the scheme, high-precision and large-scale aerosol micro-physical property profile inversion can be realized, so that atmospheric aerosol monitoring research is effectively supported.

Description

Technical Field

[0001] This invention relates to the field of aerosol detection technology, and more specifically, to a method, apparatus, equipment, and medium for determining the microphysical property profile of aerosols. Background Technology

[0002] With rapid economic growth and accelerated urbanization, industrial emissions and traffic exhaust have increased significantly, making particulate matter pollution one of the most prominent types of air pollution. It not only directly causes health problems related to the respiratory and cardiovascular systems, but also interferes with climate change, accelerates building corrosion, reduces atmospheric visibility, and impacts transportation and the ecological environment. Therefore, accurate monitoring and research of particulate matter is a crucial prerequisite for pollution control.

[0003] Currently, monitoring methods for atmospheric particulate matter are mainly divided into two categories: ground-based monitoring and satellite remote sensing. Ground-based monitoring includes ground sampling methods and ground-based lidar methods: ground sampling methods directly collect and analyze particulate matter by deploying equipment at fixed sites; ground-based lidar utilizes the scattering characteristics of laser light to obtain the vertical profile information of particulate matter. Satellite remote sensing technology, with its advantages of wide coverage and high observation efficiency, can achieve synchronous observation at regional and even global scales, effectively supplementing the spatial coverage deficiencies of ground-based monitoring.

[0004] Existing technologies still have significant shortcomings: ground sampling methods are costly to deploy, have limited monitoring range, struggle to achieve uniform spatial coverage, and cannot provide large-scale continuous monitoring data; while ground-based lidar can acquire vertical profiles, it lacks multi-angle and multi-band observation capabilities, only ensuring the reliability of extinction coefficient profiles, resulting in low accuracy in retrieving aerosol microphysical property profiles. Although satellite remote sensing offers wide coverage, it also suffers from insufficient accuracy in retrieving aerosol microphysical properties, making it difficult to support refined studies of global pollution exposure and long-distance pollution transmission, and thus failing to meet the needs of precise pollution control. Summary of the Invention

[0005] In view of this, the purpose of this invention is to provide a method, apparatus, equipment and medium for determining the microphysical property profile of aerosols, which can realize high-precision, large-scale inversion of the microphysical property profile of aerosols, thereby effectively supporting atmospheric aerosol monitoring research.

[0006] In a first aspect, embodiments of this application provide a method for determining the microphysical property profile of aerosols, the method comprising: The monitoring data of atmospheric column average aerosol volume-scale spectral distribution, real and imaginary parts of multi-wavelength complex refractive index were obtained from several ground stations at different time periods. A six-parameter distribution database of coarse and fine mode aerosols was constructed based on the monitoring data. Clustering is performed on the six-parameter distribution database to generate several coarse-modal aerosol scale spectral distribution models and several fine-modal aerosol scale spectral distribution models; For each aerosol-scale spectral distribution model, the backscattering coefficient at a preset wavelength is calculated based on Mie scattering theory. Using a preset cost function, the calculated backscattering coefficient is compared layer by layer with the observed backscattering coefficient values ​​of the corresponding wavelength measured by the spaceborne lidar at different altitudes, and the microphysical property profile of the aerosol is obtained by inversion.

[0007] Optionally, the construction of the six-parameter distribution database of coarse-mode and fine-mode aerosols includes: Using the preset aerosol particle radius as the boundary, the aerosol volume scale distribution in each monitoring data is separated into coarse mode volume distribution data and fine mode volume distribution data; A dual-modal log-normal distribution function is used to fit the coarse-mode volume distribution data and the fine-mode volume distribution data respectively to obtain six-parameter distribution data; The six-parameter distribution data include peak radius, peak height, standard deviation, particle size distribution, and the real and imaginary parts of the complex refractive index.

[0008] Optionally, the dual-modal log-normal distribution function is:

[0009] in, This represents the particle volume-scale spectral distribution. Indicates the first Modal volume peak height, Indicates particle radius, Indicates the first Peak radius of the mode, Indicates the first The standard deviation of the modes =1 corresponds to the fine mode. =2 corresponds to the coarse mode.

[0010] Optionally, the step of clustering the six-parameter distribution database to generate several coarse-modal aerosol scale spectral distribution models and several fine-modal aerosol scale spectral distribution models includes: The peak radius, peak height, and standard deviation in the six-parameter distribution data are normalized. Based on the normalized parameters, a predetermined number of aerosol-scale spectral distribution models are generated by using a clustering algorithm that aims to minimize the total dissimilarity J to calculate the dissimilarity between the parameters and the cluster centers using Euclidean distance. Wherein, the total dissimilarity J is expressed as:

[0011] in, The preset number of models, These are the normalized parameter values. For the first Cluster centers, For belonging to the first The parameter set of each cluster.

[0012] Optionally, the calculation of the backscattering coefficient at a preset wavelength based on Mie scattering theory for each aerosol-scale spectral distribution model includes: The backscattering coefficient for the preset wavelength is calculated using the following formula. ;

[0013] in, This represents the particle volume-scale spectral distribution. The size parameter representing the particle. Indicates particle radius, The backscattering efficiency factor of particulate matter. Indicates the preset wavelength Preset wavelength The real part of the complex refractive index below, Preset wavelength The imaginary part of the complex refractive index, superscript l represents the calculated value.

[0014] Optionally, the preset wavelength includes 355nm, 532nm and 1064nm; The cost function is:

[0015] in, This represents the calculated value of the cost function. , , These are the observed backscattering coefficients measured by the spaceborne lidar at wavelengths of 355nm, 532nm, and 1064nm, respectively. , , These are the calculated values ​​of the backscattering coefficient at the corresponding wavelengths. , , The uncertainty of the observed backscattering coefficients at the corresponding wavelengths is expressed as the standard deviation of the observed values ​​per hour, indicated by the superscript. This represents the observed value.

[0016] Optionally, the step of using a preset cost function to iteratively compare the calculated backscattering coefficient with the observed backscattering coefficient at corresponding wavelengths measured by a spaceborne lidar at different altitudes to obtain the aerosol microphysical property profile includes: The iterative comparison is performed layer by layer from low to high altitude. At each altitude level, the coarse-mode aerosol scale distribution model, the fine-mode aerosol scale distribution model, and the corresponding complex refractive index that minimize the value of the cost function are selected as the inversion results for that layer. The aerosol microphysical profile is constructed from the inversion results of all altitude layers.

[0017] Secondly, embodiments of this application provide an apparatus for determining the microphysical profile of aerosols, the apparatus comprising: The monitoring data acquisition module is used to acquire monitoring data of atmospheric column average aerosol volume scale spectral distribution, real part and imaginary part of multi-wavelength complex refractive index collected from several ground stations at different time periods. The database construction module is used to construct a six-parameter distribution database of coarse and fine mode aerosols based on various monitoring data; The distribution model construction module is used to cluster the six-parameter distribution database to generate several coarse-mode aerosol scale spectral distribution models and several fine-mode aerosol scale spectral distribution models. The calculation value determination module is used to calculate the backscattering coefficient value at a preset wavelength for each of the aerosol scale spectral distribution models based on Mie scattering theory. The microphysical property profile determination module is used to use a preset cost function to iteratively compare the calculated backscattering coefficient with the observed backscattering coefficient at the corresponding wavelength measured by the spaceborne lidar at different altitudes, and invert to obtain the aerosol microphysical property profile.

[0018] Optionally, the construction of the six-parameter distribution database of coarse-mode and fine-mode aerosols includes: Using the preset aerosol particle radius as the boundary, the aerosol volume scale distribution in each monitoring data is separated into coarse mode volume distribution data and fine mode volume distribution data; A dual-modal log-normal distribution function is used to fit the coarse-mode volume distribution data and the fine-mode volume distribution data respectively to obtain six-parameter distribution data; The six-parameter distribution data include peak radius, peak height, standard deviation, particle size distribution, and the real and imaginary parts of the complex refractive index.

[0019] Optionally, the dual-modal log-normal distribution function is:

[0020] in, This represents the particle volume-scale spectral distribution. Indicates the first Modal volume peak height, Indicates particle radius, Indicates the first Peak radius of the mode, Indicates the first The standard deviation of the modes =1 corresponds to the fine mode. =2 corresponds to the coarse mode.

[0021] Optionally, the step of clustering the six-parameter distribution database to generate several coarse-modal aerosol scale spectral distribution models and several fine-modal aerosol scale spectral distribution models includes: The peak radius, peak height, and standard deviation in the six-parameter distribution data are normalized. Based on the normalized parameters, a predetermined number of aerosol-scale spectral distribution models are generated by using a clustering algorithm that aims to minimize the total dissimilarity J to calculate the dissimilarity between the parameters and the cluster centers using Euclidean distance. Wherein, the total dissimilarity J is expressed as:

[0022] in, The preset number of models, These are the normalized parameter values. For the first Cluster centers, For belonging to the first The parameter set of each cluster.

[0023] Optionally, the calculation of the backscattering coefficient at a preset wavelength based on Mie scattering theory for each aerosol-scale spectral distribution model includes: The backscattering coefficient for the preset wavelength is calculated using the following formula. ;

[0024] in, This represents the particle volume-scale spectral distribution. The size parameter representing the particle. Indicates particle radius, The backscattering efficiency factor of particulate matter. Indicates the preset wavelength Preset wavelength The real part of the complex refractive index below, Preset wavelength The imaginary part of the complex refractive index, superscript l represents the calculated value.

[0025] Optionally, the preset wavelength includes 355nm, 532nm and 1064nm; The cost function is:

[0026] in, This represents the calculated value of the cost function. , , These are the observed backscattering coefficients measured by the spaceborne lidar at wavelengths of 355nm, 532nm, and 1064nm, respectively. , , These are the calculated values ​​of the backscattering coefficient at the corresponding wavelengths. , , The uncertainty of the observed backscattering coefficients at the corresponding wavelengths is expressed as the standard deviation of the observed values ​​per hour, indicated by the superscript. This represents the observed value.

[0027] Optionally, the step of using a preset cost function to iteratively compare the calculated backscattering coefficient with the observed backscattering coefficient at corresponding wavelengths measured by a spaceborne lidar at different altitudes to obtain the aerosol microphysical property profile includes: The iterative comparison is performed layer by layer from low to high altitude. At each altitude level, the coarse-mode aerosol scale distribution model, the fine-mode aerosol scale distribution model, and the corresponding complex refractive index that minimize the value of the cost function are selected as the inversion results for that layer. The aerosol microphysical profile is constructed from the inversion results of all altitude layers.

[0028] Thirdly, embodiments of this application provide a computer device, including: a processor, a memory, and a bus. The memory stores machine-readable instructions executable by the processor. When the computer device is running, the processor communicates with the memory via the bus. When the machine-readable instructions are executed by the processor, the steps of the aerosol microphysical property profile determination method described in any of the optional embodiments of the first aspect are performed.

[0029] Fourthly, embodiments of this application provide a computer-readable storage medium storing a computer program, which, when executed by a processor, performs the steps of the aerosol microphysical property profile determination method described in any of the optional embodiments of the first aspect.

[0030] The technical solution provided in this application includes, but is not limited to, the following beneficial effects: Firstly, by collecting monitoring data from several ground stations at different times, we can obtain richer and more representative aerosol baseline data in terms of time and space. This provides sufficient and diverse sample support for the subsequent construction of databases and models, avoiding bias in subsequent analysis results due to limited data and ensuring the comprehensiveness and reliability of subsequent steps.

[0031] Then, based on the monitoring data, a six-parameter distribution database of coarse and fine aerosols is constructed. This database can systematically store and manage key microphysical parameters of aerosols (such as peak radius and complex refractive index), forming a structured and standardized data foundation. This facilitates efficient subsequent retrieval and analysis, and clearly reflects the characteristic differences between coarse and fine aerosols, providing a data prerequisite for accurately distinguishing aerosol types.

[0032] Next, the six-parameter distribution database is clustered to generate several coarse and fine modal aerosol scale distribution models. These models can group aerosol parameters based on their inherent similarity, summarize complex and diverse aerosol data into typical models, capture the characteristic patterns of different types of aerosols, achieve refined classification of aerosols, make the subsequent inversion process more targeted, and improve the model's coverage of different aerosol types in the actual atmosphere.

[0033] Then, based on the Mie scattering theory, the backscattering coefficient of the preset wavelength is calculated. The Mie scattering theory is a classic and reliable theory that describes the light scattering of spherical particles. It can accurately simulate the scattering process of light by aerosol particles, thereby obtaining accurate theoretical calculations of the backscattering coefficient. This provides a reliable theoretical basis for subsequent comparison with the observations of spaceborne lidar and is a key link between aerosol microphysics models and optical observations.

[0034] Finally, the cost function is used to iteratively compare the calculated backscattering coefficient with the observation values ​​of the spaceborne lidar at different altitudes. This allows for the optimization of the inversion results at each altitude level in the vertical direction, ensuring that the aerosol microphysical properties at each altitude can accurately match the observations of the spaceborne lidar. At the same time, the large-scale observation capability of the spaceborne lidar compensates for the spatial limitations of ground monitoring, ultimately enabling the inversion of high-precision, large-scale coverage aerosol microphysical property profiles.

[0035] In summary, the technical solution of this application starts with data collection from multiple sites and time periods. By constructing a database and generating typical models through clustering, and then combining Mie scattering theory with iterative comparison with spaceborne lidar, it not only ensures the richness of the data foundation and the typicality of the model, but also achieves high-precision, large-scale aerosol microphysical property profile inversion by means of precise optical calculation and large-scale observation. This provides an effective technical means for the monitoring and research of atmospheric aerosols, and demonstrates advantages in terms of data support, model accuracy, and spatial coverage.

[0036] To make the above-mentioned objects, features and advantages of the present invention more apparent and understandable, preferred embodiments are described below in detail with reference to the accompanying drawings. Attached Figure Description

[0037] To more clearly illustrate the technical solutions of the embodiments of the present invention, the accompanying drawings used in the embodiments will be briefly introduced below. It should be understood that the following drawings only show some embodiments of the present invention and should not be regarded as a limitation on the scope. For those skilled in the art, other related drawings can be obtained based on these drawings without creative effort.

[0038] Figure 1 The flowchart of a method for determining the microphysical profile of aerosols provided in Embodiment 1 of the present invention is shown. Figure 2 A flowchart of a distributed database construction method provided in Embodiment 1 of the present invention is shown; Figure 3 The flowchart of a method for generating an aerosol-scale spectral distribution model provided in Embodiment 1 of the present invention is shown. Figure 4 The flowchart of an aerosol microphysical property profile inversion method provided in Embodiment 1 of the present invention is shown; Figure 5 The flowchart of a specific inversion method for aerosol microphysical property profiles provided in Embodiment 1 of the present invention is shown. Figure 6 This diagram illustrates a theoretical reference value for a combination of fine-mode and coarse-mode volumetric spectral distribution models provided in Embodiment 1 of the present invention. Figure 7 This diagram illustrates the concentration distribution of an aerosol at a vertical height, as provided in Embodiment 1 of the present invention. Figure 8 This diagram shows a schematic diagram of an aerosol microphysical property profile determination device provided in Embodiment 2 of the present invention; Figure 9 A schematic diagram of the structure of a computer device provided in Embodiment 3 of the present invention is shown. Detailed Implementation

[0039] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, 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. The components of the embodiments of the present invention described and shown in the accompanying drawings can generally be arranged and designed in various different configurations. Therefore, the following detailed description of the embodiments of the present invention provided in the accompanying drawings is not intended to limit the scope of the claimed invention, but merely to illustrate selected embodiments of the invention. All other embodiments obtained by those skilled in the art based on the embodiments of the present invention without inventive effort are within the scope of protection of the present invention.

[0040] Example 1 To facilitate understanding of this application, the following is combined with... Figure 1 The flowchart of the method for determining the microphysical profile of aerosols provided in Embodiment 1 of the present invention is shown below, and the content of the flowchart describes Embodiment 1 of this application in detail.

[0041] See Figure 1 As shown, Figure 1 The flowchart of a method for determining the microphysical profile of aerosols according to Embodiment 1 of the present invention is shown, wherein the method includes steps S101 to S105: S101: Acquire monitoring data on the atmospheric column average aerosol volumetric spectral distribution, real and imaginary parts of the multi-wavelength complex refractive index collected from several ground stations at different time periods.

[0042] Specifically, the atmospheric column-averaged aerosol volumetric spectral distribution (PVSD), the corresponding real and imaginary parts of the multi-wavelength complex refractive index, and other data for multiple time periods at multiple stations are obtained through a ground-based observation network. The atmospheric column-averaged aerosol volumetric spectral distribution (PVSD), the real and imaginary parts of the multi-wavelength complex refractive index for each time period at each station are stored as a single monitoring data point.

[0043] S102: Construct a six-parameter distribution database of coarse and fine mode aerosols based on the monitoring data.

[0044] Specifically, the aerosol volumetric distribution in each monitoring data point is first separated into coarse-mode volumetric distribution data and fine-mode volumetric distribution data, with a 1.0 μm aerosol particle radius as the dividing line. Then, a dual-mode log-normal distribution function is used to fit the coarse and fine-mode volumetric distribution data respectively, and six types of parameters are extracted: peak radius, peak height, standard deviation, particle size distribution, real part and imaginary part of complex refractive index. Finally, the six parameters of coarse and fine modes corresponding to all monitoring data are integrated to form a six-parameter distribution database.

[0045] S103: Cluster the six-parameter distribution database to generate several coarse-mode aerosol scale distribution models and several fine-mode aerosol scale distribution models.

[0046] Specifically, the peak radius, peak height, and standard deviation of the six-parameter distribution data are first normalized; then, based on the normalized parameters, a predetermined number of models are generated by grouping them using a clustering algorithm that aims to minimize the total dissimilarity. The coarse-mode and fine-mode aerosol scale spectral distribution models need to be clustered separately.

[0047] S104: For each of the aerosol-scale spectral distribution models, calculate the backscattering coefficient at a preset wavelength based on Mie scattering theory.

[0048] Specifically, the preset wavelengths are first determined to be 355nm, 532nm and 1064nm; then, for each coarse and fine mode aerosol scale spectral distribution model, combined with the corresponding particle volume scale spectral distribution, real part and imaginary part of complex refractive index, the backscattering coefficient at each preset wavelength is calculated based on Mie scattering theory and corresponding formula.

[0049] S105: Using a preset cost function, the calculated backscattering coefficient is compared layer by layer with the observed backscattering coefficient of the corresponding wavelength measured by the spaceborne lidar at different altitudes, and the aerosol microphysical property profile is obtained by inversion.

[0050] Specifically, in order of increasing altitude, the calculated backscattering coefficients of each altitude layer are substituted with the observation values ​​of the spaceborne lidar at the same altitude and wavelength into a preset cost function; the coarse and fine mode models and their corresponding complex refractive indices that minimize the cost function value are selected for each altitude layer as the inversion result for that layer; finally, the inversion results of all altitude layers are integrated to form a complete profile of aerosol microphysical properties.

[0051] In one feasible implementation plan, see Figure 2 The above, Figure 2 The flowchart of a distribution database construction method provided in Embodiment 1 of the present invention is shown, wherein the construction of a six-parameter distribution database of coarse-mode and fine-mode aerosols includes steps S201-S202: S201: Using the preset aerosol particle radius as the boundary, the aerosol volume scale distribution in each monitoring data is separated into coarse mode volume distribution data and fine mode volume distribution data.

[0052] Specifically, the monitoring data from all ground stations and different time periods acquired by S101 are integrated to construct an aerosol volume distribution database. Each data point contains the aerosol volume scale spectral distribution of the corresponding station and time period.

[0053] Then, using a particle radius of 1.0 μm as the boundary between fine and coarse modes, mathematical truncation was performed on each aerosol volumetric distribution monitoring data in the database. Specifically, the part with a radius less than 1.0 μm was classified as fine mode volumetric distribution data, and the part with a radius greater than or equal to 1.0 μm was classified as coarse mode volumetric distribution data. Finally, the coarse and fine mode volumetric distribution data corresponding to each monitoring data were obtained.

[0054] S202: Using a dual-modal log-normal distribution function, the coarse-mode volume distribution data and the fine-mode volume distribution data are fitted respectively to obtain six-parameter distribution data.

[0055] Specifically, for each monitoring data point obtained from S201, the coarse-mode volume distribution data and the fine-mode volume distribution data were fitted using a dual-mode log-normal distribution function. Through fitting, the six-parameter distribution data of the coarse and fine-mode aerosols, conforming to a log-normal distribution, were extracted. These six parameters include peak radius, peak height, standard deviation, particle size distribution, and the real part of the complex refractive index. and the virtual part Finally, the coarse and fine mode six-parameter distribution data corresponding to all monitoring data are integrated to construct a six-parameter distribution database of coarse and fine mode aerosols.

[0056] In addition, linear fitting can be performed on the real and imaginary parts of the complex refractive index of multi-wavelength aerosols corresponding to each monitoring data point to obtain linear functions of the real and imaginary parts of the complex refractive index with respect to wavelength changes, which can be used for subsequent inversion calculations.

[0057] The six-parameter distribution data include peak radius, peak height, standard deviation, particle size distribution, and the real and imaginary parts of the complex refractive index.

[0058] Specifically, peak height represents the volume concentration of aerosols per unit volume; peak radius is the particle radius corresponding to the peak height, used to describe the average size of aerosol particles; standard deviation reflects the dispersion of aerosol particle size distribution, with a larger value indicating more significant particle size differences; the real part of the complex refractive index is related to the optical properties of aerosol particles, mainly affecting the scattering of light by particles, while the imaginary part of the complex refractive index also affects optical properties, primarily determining the particle's ability to absorb light; and particle size distribution visually reflects the distribution pattern of aerosol particles in different size ranges.

[0059] In a feasible implementation, the dual-modal log-normal distribution function is:

[0060] in, This represents the particle volume-scale spectral distribution. Indicates the first Modal volume peak height, Indicates particle radius, Indicates the first Peak radius of the mode, Indicates the first The standard deviation of the modes =1 corresponds to the fine mode. =2 corresponds to the coarse mode.

[0061] Specifically, using the bimodal log-normal function as an ideal approximation of the aerosol scale spectral distribution, the following formula is used to fit the bimodal log-normal distribution based on the coarse and fine mode aerosol volume distribution data corresponding to each monitoring data point, obtaining the peak radius, peak height, and standard deviation for each model:

[0062] in, This represents the particle volume-scale spectral distribution. This represents the peak height of the i-th mode aerosol volume. Indicates particle radius, This represents the peak radius of the i-th mode of aerosol. Let m represent the standard deviation of the i-th mode of aerosol, and m be the number of aerosol modes. In this embodiment, m is taken as 2, that is, two modes of aerosol, coarse and fine, are used for fitting.

[0063] In one feasible implementation plan, see Figure 3 The above, Figure 3 The flowchart illustrates a method for generating an aerosol-scale spectral distribution model according to Embodiment 1 of the present invention. The method involves clustering the six-parameter distribution database to generate several coarse-mode aerosol-scale spectral distribution models and several fine-mode aerosol-scale spectral distribution models, comprising steps S301-S302: S301: Normalize the peak radius, peak height, and standard deviation in the six-parameter distribution data.

[0064] Specifically, for the coarse-mode aerosol six-parameter distribution data and the fine-mode aerosol six-parameter distribution data stored in the six-parameter distribution database, the three core parameters of peak radius, peak height and standard deviation in the two types of data need to be normalized to ensure that the dimensions of each parameter are consistent in the subsequent clustering process and to avoid interference with the clustering results due to differences in parameter magnitude.

[0065] S302: Based on the normalized parameters, a predetermined number of aerosol-scale spectral distribution models are generated by using a clustering algorithm that aims to minimize the total dissimilarity J, which calculates the dissimilarity between the parameters and the cluster centers using Euclidean distance.

[0066] Specifically, for the normalized coarse and fine modal parameters, a predetermined number of cluster centers are randomly selected for each data point (corresponding to the number of aerosol-scale spectral distribution models generated in the final generation). Then, Euclidean distance is used to calculate the dissimilarity between each parameter and each cluster center value. Subsequently, the parameter with the smallest dissimilarity to a given cluster center is grouped into the same cluster, and the average value of all parameters within that cluster is taken to recalculate new cluster center values.

[0067] Finally, repeat the above steps of "calculating dissimilarity - classifying - updating cluster centers" for all cluster center values ​​until the total dissimilarity J corresponding to each cluster center value reaches the minimum value. At this point, each cluster center value corresponds to an aerosol scale spectral distribution model, and finally a predetermined number of coarse mode and fine mode aerosol scale spectral distribution models are generated.

[0068] Wherein, the total dissimilarity J is expressed as:

[0069] in, The preset number of models, These are the normalized parameter values. For the first Cluster centers, For belonging to the first The parameter set of each cluster.

[0070] Specifically, the parameters include peak radius, peak height, and standard deviation.

[0071] Specifically, this embodiment uses a random forest clustering method to obtain five coarse-mode and five fine-mode aerosol scale spectral distribution models. The coarse-mode and fine-mode aerosol volumetric scale spectral distribution models require separate clustering, with each mode divided into five groups. The initial cluster centers for each group are randomly selected, and the final cluster centers represent the clustering results. The dissimilarity between all parameters and all cluster centers is calculated. This embodiment uses standardized Euclidean distance as the dissimilarity metric for data clustering, obtaining five microphysical characteristic parameters (including volumetric scale spectral distribution and complex refractive index) models for each of the coarse and fine-mode aerosols. This allows for a more refined classification of aerosols based on the intrinsic relationships between the microphysical properties of different modes, enabling a more accurate characterization of local aerosol properties.

[0072] This embodiment uses clustering to obtain aerosol microphysical characteristic parameter models for five scales each of coarse and fine modes. The complex refractive index and volumetric scale distribution in the model can be used to analyze chemical composition and characterize information such as the size, absorbency, and chemical composition of particulate matter. Furthermore, based on inherent physical relationships (i.e., common knowledge in atmospheric physics), such as hygroscopic particles generally being small-scale spherical particles, mainly composed of soluble inorganic salts, exhibiting strong scattering (large real part of complex refractive index), and generally occurring more frequently in urban areas, which is related to local industrial patterns and geographical features. In combination, coarse particles in arid or semi-arid regions are generally dust, while coarse particles in built-up areas are generally fly ash. By combining coarse and fine modal models, most aerosols in nature can be characterized. In this embodiment, five fine-modal aerosol microphysical characteristic parameter models (including urban pollution type, terrestrial background type, secondary pollution type, mixed pollution type, and polluted fly ash type) and five coarse and fine-modal aerosol microphysical characteristic parameter models (including summer fly ash type, winter fly ash type, primary dust type, transported dust type, and background dust type) are obtained through clustering. The corresponding parameters are shown in Table 1.

[0073] Table 1 Where, r f and r c r represents the peak radius in the fine-mode and coarse-mode models, respectively; SMF and r SMC σ represents the peak radius in the sub-microfine mode and the super-microcoarse mode models, respectively; f and σ c σ represents the standard deviation in the fine-mode and coarse-mode models, respectively; SMF and σ SMC C represents the standard deviation in the submicron fine mode and ultramicron coarse mode models, respectively; f and C c C represents the peak height in the fine-mode and coarse-mode models, respectively; SMF and C SMC n represents the peak height in the submicron fine-mode and ultramicron coarse-mode models, respectively; f and n c k represents the real part of the complex refractive index (675 nm ~ 1020 nm) corresponding to the fine-mode and coarse-mode models, respectively; f and k c These represent the imaginary parts of the complex refractive index (675 nm ~ 1020 nm) corresponding to the fine and coarse mode models, respectively; k f,440 and k c,440 These represent the imaginary parts (440 nm) of the complex refractive index corresponding to the fine and coarse modes, respectively.

[0074] Preferably, the observation data in this embodiment comes from one year of observation data from eight ground-based Sun-Sky Radiometer Observation Network (SONET) stations, including: Beijing, Nanjing, Hefei, and Chengdu stations, which are dominated by urban / industrial aerosols and have seasonal biomass combustion aerosols; Guangzhou and Shanghai stations, which are dominated by urban / marine aerosols; and Xi'an and Zhangye stations, which are dominated by dust aerosols. Among them, Beijing, Nanjing, Hefei, and Chengdu stations respectively contain light-absorbing and non-absorbing urban / industrial aerosol particles; Guangzhou and Shanghai stations represent polluted and clean marine / urban aerosol particles, respectively; and Xi'an and Zhangye stations represent polluted and clean dust aerosol particles, respectively.

[0075] The acquired observational data includes the atmospheric column-averaged aerosol volumetric spectral distribution (PVSD) and the real part n of the multi-wavelength complex refractive index. λ and the imaginary part k λ The acquired observation data will be statistically analyzed to establish a database of aerosol microphysical characteristic parameters, including the aerosol volume-scale spectral distribution and the corresponding multi-wavelength aerosol complex refractive index.

[0076] Then, using a particle radius of 1.0 μm as the boundary between fine-mode and coarse-mode particles, mathematical truncation was performed on each observation result in the aerosol volume distribution database to obtain the coarse-mode aerosol volume distribution model and the fine-mode aerosol volume distribution model corresponding to each monitoring data, which respectively include aerosol volume scale spectral distribution and multi-wavelength aerosol complex refractive index.

[0077] The data used in this embodiment includes data from multiple stations over a long time scale, encompassing various weather conditions, even extreme weather. Different stations are dominated by different types of aerosols, which often exhibit different optical characteristics. Dust-type aerosol particles have larger particle sizes, strong scattering, and weak absorption. Marine aerosols also have larger particle sizes, strong scattering, and no absorption. Urban / industrial and biomass combustion aerosols have complex sources, closely related to human activities, typically including urban traffic exhaust emissions and factory emissions; they have smaller particle sizes and stronger absorption. By statistically analyzing the optical properties of various aerosol types, the resulting coarse-mode and fine-mode aerosol volume-scale spectral distribution models are universal and representative.

[0078] In one feasible implementation, the calculation of the backscattering coefficient at a preset wavelength based on Mie scattering theory for each aerosol-scale spectral distribution model includes: The backscattering coefficient for the preset wavelength is calculated using the following formula. ;

[0079] in, This represents the particle volume-scale spectral distribution. The size parameter representing the particle. Indicates particle radius, The backscattering efficiency factor of particulate matter. Indicates the preset wavelength. Preset wavelength The real part of the complex refractive index below, Preset wavelength The imaginary part of the complex refractive index, superscript This represents a calculated value.

[0080] In one feasible implementation, the preset wavelength includes 355nm, 532nm and 1064nm.

[0081] Specifically, the preset wavelengths include 355nm, 532nm, and 1064nm. The five coarse-mode and five fine-mode volume-scale spectral distribution models are combined in pairs and substituted with the corresponding complex refractive index into the following Mie scattering calculation formula to obtain the backscattering coefficients at 355nm, 532nm, and 1064nm.

[0082] The cost function is:

[0083] in, This represents the calculated value of the cost function. , , These are the observed backscattering coefficients measured by the spaceborne lidar at wavelengths of 355nm, 532nm, and 1064nm, respectively. , , These are the calculated values ​​of the backscattering coefficient at the corresponding wavelengths. , , The uncertainty of the observed backscattering coefficients at the corresponding wavelengths is expressed as the standard deviation of the observed values ​​per hour, indicated by the superscript. This represents the observed value.

[0084] In one feasible implementation plan, see Figure 4 The above, Figure 4The flowchart of an aerosol microphysical property profile inversion method provided in Embodiment 1 of the present invention is shown. The method involves using a preset cost function to iteratively compare the calculated backscattering coefficient with the observed backscattering coefficient values ​​at corresponding wavelengths measured by a spaceborne lidar at different altitudes to invert the aerosol microphysical property profile. This includes steps S401-S403. S401: Perform the iterative comparison layer by layer from low to high altitude.

[0085] Specifically, referring to the altitude coverage range (usually 0~10km) and height resolution (usually 500m~1km) of the spaceborne lidar observation data, in order of increasing altitude (e.g., starting from 0.5km and gradually increasing to 10km), an iterative comparison is carried out for each layer. That is, the calculated backscattering coefficient of each aerosol-scale spectral distribution model corresponding to the layer is compared with the backscattering coefficient observation value of the spaceborne lidar at the same altitude and wavelength. The comparison process of each layer is independent of each other and does not depend on the results of other altitude layers.

[0086] S402: At each altitude level, select the coarse-mode aerosol scale distribution model, the fine-mode aerosol scale distribution model, and the corresponding complex refractive index that minimize the value of the cost function as the inversion result for that layer.

[0087] Specifically, in each altitude layer, the calculated backscattering coefficients corresponding to the combination of all coarse and fine mode aerosol scale spectral distribution models are substituted into a preset cost function along with the observation values ​​of the spaceborne lidar for that layer. The coarse mode model and fine mode model combination that minimizes the cost function value are selected, and the complex refractive index (from the six-parameter distribution database) corresponding to the model combination is extracted. The above model and the complex refractive index are used together as the aerosol microphysical property inversion result for that altitude layer.

[0088] S403: The aerosol microphysical profile is composed of the inversion results of all altitude layers.

[0089] Specifically, the inversion results of all altitude layers (including the optimal coarse / fine modal scale spectral distribution models and corresponding complex refractive indices for each layer) are collected and integrated in order of altitude. The vertical distribution result formed after integration is the aerosol microphysical property profile, which includes the aerosol volume scale spectral distribution profile (reflecting the particle size and concentration at different altitudes) and the complex refractive index profile (reflecting the optical properties of particles at different altitudes), fully reflecting the distribution law of aerosol microphysical properties in the vertical direction.

[0090] See Figure 5 The above, Figure 5The flowchart illustrates a specific inversion method for aerosol microphysical property profiles provided in Embodiment 1 of the present invention. The flowchart shows the complete process of inverting aerosol microphysical property profiles, using SONET solar-sky radiometer observation network data and spaceborne lidar data as inputs. The final inversion result is obtained after multiple processing steps. The specific process is as follows: First, the SONET solar-sky radiometer observation network data is processed: on the one hand, the average volumetric spectral distribution (PVSD) data of the entire atmosphere is extracted; after mathematical truncation with a radius of 1 μm, the fine-mode PVSD of the volumetric spectral distribution is obtained. f and coarse-mode PVSD c Data; then, by approximating with a log-normal distribution function, a coarse and fine mode six-parameter model, namely PVSD, is constructed. f [C f ,r f ,σ f PVSD c[ C c ,r c ,σ c Subsequently, five fine-mode and five coarse-mode models were statistically analyzed using random forest clustering. On the other hand, the real and imaginary parts (n,k) of the average complex refractive index of the entire atmosphere were extracted; a function of the complex refractive index versus wavelength was generated as driving data. These two results were input into the dashed box region containing Mie scattering calculations. Simultaneously, coarse and fine-mode model combinations were selected within the dashed box to form a complete scale spectral distribution as driving data for Mie scattering calculations. Meanwhile, spaceborne lidar data was processed: first, backscattering coefficients at different altitudes in the 355nm, 532nm, and 1064nm bands, as well as observed depolarization ratios, were acquired; then, it was determined whether the cost function characterizing the deviation between the calculated and observed values ​​had reached its minimum. If not, the process returned to the dashed box stage, and the Mie scattering calculations and mode model combinations were repeated; if yes, the calculated backscattering coefficients in the 355nm, 532nm, and 1064nm bands were acquired, and finally, the coarse- and fine-mode aerosol volumetric scale spectral distributions and multi-wavelength complex refractive index profiles were output.

[0091] In establishing the volumetric spectral distribution models for coarse and fine aerosol modes, this application adopts the research experience of most scholars, using a particle radius of 1.0 μm as the boundary between fine and coarse modes. After statistical analysis of a large amount of sample data, theoretical reference values ​​for the volumetric spectral distribution parameters (peak radius, standard deviation, peak concentration) of the fine and coarse modes most commonly encountered in nature are provided. (See [link to relevant documentation]). Figure 6 The above, Figure 6 This diagram illustrates a theoretical reference value for a combination of fine-mode and coarse-mode volume-scale spectral distribution models provided in Embodiment 1 of the present invention, wherein... Figure 6The left side shows the volumetric spectral distribution curve, with the horizontal axis labeled "radius (μm)" and the vertical axis labeled "volumetric spectral distribution dV(r) / dln(r) (μm)". 3 / μm 2 The curve exhibits a double-peak characteristic, reflecting the volume distribution patterns of fine-mode and coarse-mode particles; Figure 6 The right side is a bar chart with the horizontal axis labeled "AOD-F, AOD-C, VolCon-F, VolCon-C" and the vertical axis labeled "concentration value". The four groups of bars visually present the concentration differences of different categories of parameters.

[0092] In the verification scenario of vertical aerosol distribution, see... Figure 7 The above, Figure 7 This diagram illustrates the vertical concentration distribution of an aerosol provided in Embodiment 1 of the present invention, where the vertical axis is labeled "altitude (km)" and the horizontal axis is labeled "extinction coefficient (km)". -1 The figure contains three curves labeled "a, upper extinction contribution 29%", "b, upper extinction contribution 34%", and "c, upper extinction contribution 40%", along with the "boundary layer height 0.9 km" and the "boundary layer" range. This figure assumes the concentration distribution of aerosols along the vertical height (the three cases correspond to "significant transport layer", "some transport material layer", and "virtually no transport material layer" above the boundary layer, respectively), with the upper-boundary particulate matter contributing 40%, 34%, and 29% to the total optical distribution of the entire atmospheric column, respectively. This provides a relatively comprehensive view of the actual vertical distribution of the atmosphere.

[0093] Given the scarcity of equipment capable of directly measuring aerosol volumetric spectral distribution and complex refractive index under open atmospheric paths, direct verification of the inversion results for these two microphysical parameters is challenging. Therefore, this application employs an indirect comparative verification method: assuming the aerosol volumetric spectral distribution is as follows... Figure 6 As shown, inversion was performed on the three vertical distribution scenarios described above, and the results show that the inverted values ​​are consistent with... Figure 6 The deviation from the theoretical reference value is approximately 12.7%. This result indicates that the inversion results of the coarse and fine modal aerosol volume-scale spectral distribution and complex refractive index profiles obtained in this application have high accuracy, demonstrating that the inversion method can be effectively used to obtain aerosol microphysical parameter profiles.

[0094] In summary, this application employs a data-driven approach combining satellite and ground-based data to invert the profiles of coarse and fine-mode aerosol microphysical properties. Specifically, the process involves using a ground-based observation network to acquire the column-averaged aerosol volumetric spectral distribution (PVSD) across the entire atmospheric layer, as well as the real part (n) of the complex refractive index at multiple wavelengths. λ ) and imaginary part (k) λThis method involves statistical analysis of long-term data series to construct coarse and fine-mode aerosol volumetric spectral distribution databases. Data fitting is then used to form six-parameter distribution databases for coarse and fine modes of aerosols. Cluster analysis is performed on these six-parameter databases to construct coarse and fine-mode aerosol volumetric spectral distribution models. These models are then jointly inverted with observation data from a spaceborne lidar system to obtain the aerosol microphysical property profiles. This method has significant advantages in terms of spatial coverage, inversion accuracy, observation periodicity, and result comparability.

[0095] Meanwhile, existing technologies often overlook the intrinsic physical relationships between the microphysical parameters of coarse and fine aerosol modes. In reality, the combination of coarse and fine aerosol modes is not only related to season and geographical location but also affected by extreme weather, necessitating a combined approach of satellite observation and ground-based measurements. This invention not only inverts the volumetric spectral distribution profiles of coarse and fine modes applicable to different regions, seasons, and types of aerosols globally, but also possesses the ability to invert complex refractive indexes and provides layered visualization of aerosol vertical profiles, making abstract aerosol properties more intuitive.

[0096] Example 2 See Figure 8 As shown, Figure 8 A schematic diagram of an aerosol microphysical property profile determination device provided in Embodiment 2 of the present invention is shown, wherein the device includes: The monitoring data acquisition module 801 is used to acquire monitoring data of the atmospheric whole column average aerosol volume scale spectral distribution, real part and imaginary part of multi-wavelength complex refractive index collected from several ground stations at different time periods. Database construction module 802 is used to construct a six-parameter distribution database of coarse and fine mode aerosols based on various monitoring data; The distribution model construction module 803 is used to cluster the six-parameter distribution database to generate several coarse-mode aerosol scale distribution models and several fine-mode aerosol scale distribution models. The calculation value determination module 804 is used to calculate the backscattering coefficient value at a preset wavelength based on Mie scattering theory for each of the aerosol scale spectral distribution models. The microphysical property profile determination module 805 is used to use a preset cost function to iteratively compare the calculated backscattering coefficient with the observed backscattering coefficient at the corresponding wavelength measured by the spaceborne lidar at different altitudes, and invert to obtain the aerosol microphysical property profile.

[0097] In one feasible implementation, the construction of the six-parameter distribution database for coarse-mode and fine-mode aerosols includes: Using the preset aerosol particle radius as the boundary, the aerosol volume scale distribution in each monitoring data is separated into coarse mode volume distribution data and fine mode volume distribution data; A dual-modal log-normal distribution function is used to fit the coarse-mode volume distribution data and the fine-mode volume distribution data respectively to obtain six-parameter distribution data; The six-parameter distribution data include peak radius, peak height, standard deviation, particle size distribution, and the real and imaginary parts of the complex refractive index.

[0098] In a feasible implementation, the bimodal log-normal distribution function is:

[0099] in, This represents the particle volume-scale spectral distribution. Indicates the first Modal volume peak height, Indicates the first Peak radius of the mode, Indicates the first The standard deviation of the modes =1 corresponds to the fine mode. =2 corresponds to the coarse mode.

[0100] In a feasible implementation, the clustering of the six-parameter distribution database to generate several coarse-modal aerosol scale spectral distribution models and several fine-modal aerosol scale spectral distribution models includes: The peak radius, peak height, and standard deviation in the six-parameter distribution data are normalized. Based on the normalized parameters, a predetermined number of aerosol-scale spectral distribution models are generated by using a clustering algorithm that aims to minimize the total dissimilarity J to calculate the dissimilarity between the parameters and the cluster centers using Euclidean distance. Wherein, the total dissimilarity J is expressed as: ; in, The preset number of models, These are the normalized parameter values. For the first Cluster centers, For belonging to the first The parameter set of each cluster.

[0101] In one feasible implementation, the calculation of the backscattering coefficient at a preset wavelength based on Mie scattering theory for each aerosol-scale spectral distribution model includes: The backscattering coefficient for the preset wavelength is calculated using the following formula. ;

[0102] in, This represents the particle volume-scale spectral distribution. The size parameter representing the particle. Indicates particle radius, The backscattering efficiency factor of particulate matter. Indicates the preset wavelength Preset wavelength The real part of the complex refractive index below, Preset wavelength The imaginary part of the complex refractive index, superscript This represents a calculated value.

[0103] In one feasible implementation, the preset wavelength includes 355nm, 532nm and 1064nm; The cost function is:

[0104] in, This represents the calculated value of the cost function. , , These are the observed backscattering coefficients measured by the spaceborne lidar at wavelengths of 355nm, 532nm, and 1064nm, respectively. , , These are the calculated values ​​of the backscattering coefficient at the corresponding wavelengths. , , The uncertainty of the observed backscattering coefficients at the corresponding wavelengths is expressed as the standard deviation of the observed values ​​per hour, indicated by the superscript. This represents the observed value.

[0105] In one feasible implementation, the calculated backscattering coefficient is iteratively compared layer by layer with the observed backscattering coefficient values ​​at corresponding wavelengths measured by a spaceborne lidar at different altitudes using a preset cost function to invert and obtain the aerosol microphysical property profile, including: The iterative comparison is performed layer by layer from low to high altitude. At each altitude level, the coarse-mode aerosol scale distribution model, the fine-mode aerosol scale distribution model, and the corresponding complex refractive index that minimize the value of the cost function are selected as the inversion results for that layer. The aerosol microphysical profile is constructed from the inversion results of all altitude layers.

[0106] Example 3 Based on the same application concept, see [link / reference] Figure 9 As shown, Figure 9 A schematic diagram of the structure of a computer device provided in Embodiment 3 of the present invention is shown, wherein, as Figure 9 As shown, the computer device 900 provided in Embodiment 3 of this application includes: The computer device 900 includes a processor 901, a memory 902, and a bus 903. The memory 902 stores machine-readable instructions that can be executed by the processor 901. When the computer device 900 is running, the processor 901 communicates with the memory 902 through the bus 903. When the machine-readable instructions are executed by the processor 901, they perform the steps of the aerosol microphysical property profile determination method shown in Embodiment 1 above.

[0107] Example 4 Based on the same concept, this application also provides a computer-readable storage medium storing a computer program, which, when executed by a processor, performs the steps of the aerosol microphysical property profile determination method described in any of the above embodiments.

[0108] Those skilled in the art will clearly understand that, for the sake of convenience and brevity, the specific working process of the system and apparatus described above can be referred to the corresponding process in the foregoing method embodiments, and will not be repeated here.

[0109] The aerosol microphysical property profile determination device provided in this embodiment of the invention can be specific hardware on a device or software or firmware installed on the device. The implementation principle and technical effects of the device provided in this embodiment of the invention are the same as those in the foregoing method embodiments. For the sake of brevity, any parts not mentioned in the device embodiments can be referred to the corresponding content in the foregoing method embodiments. Those skilled in the art will clearly understand that, for the sake of convenience and brevity, the specific working processes of the systems, devices, and units described above can all be referred to the corresponding processes in the above method embodiments, and will not be repeated here.

[0110] In the embodiments provided by this invention, it should be understood that the disclosed apparatus and method can be implemented in other ways. The apparatus embodiments described above are merely illustrative. For example, the division of units is only a logical functional division, and in actual implementation, there may be other division methods. Furthermore, multiple units or components may be combined or integrated into another system, or some features may be ignored or not executed. Additionally, the displayed or discussed mutual couplings, direct couplings, or communication connections may be through some communication interfaces; indirect couplings or communication connections between devices or units may be electrical, mechanical, or other forms.

[0111] The units described as separate components may or may not be physically separate. The components shown as units may or may not be physical units; that is, they may be located in one place or distributed across multiple network units. Some or all of the units can be selected to achieve the purpose of this embodiment according to actual needs.

[0112] In addition, the functional units in the embodiments provided by the present invention can be integrated into one processing unit, or each unit can exist physically separately, or two or more units can be integrated into one unit.

[0113] If the aforementioned functions are implemented as software functional units and sold or used as independent products, they can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of this invention, or the part that contributes to the prior art, or a portion of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods described in the various embodiments of this invention. The aforementioned storage medium includes various media capable of storing program code, such as USB flash drives, portable hard drives, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical disks.

[0114] It should be noted that similar labels and letters in the following figures indicate similar items. Therefore, once an item is defined in one figure, it does not need to be further defined and explained in subsequent figures. In addition, the terms "first", "second", "third", etc. are used only to distinguish descriptions and should not be construed as indicating or implying relative importance.

[0115] Finally, it should be noted that the above-described embodiments are merely specific implementations of the present invention, used to illustrate the technical solutions of the present invention, and not to limit it. The scope of protection of the present invention is not limited thereto. Although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that any person skilled in the art can still modify or easily conceive of changes to the technical solutions described in the foregoing embodiments within the technical scope disclosed in the present invention, or make equivalent substitutions for some of the technical features; and these modifications, changes, or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of the present invention. All should be covered within the scope of protection of the present invention. Therefore, the scope of protection of the present invention should be determined by the scope of the claims.

Claims

1. A method for determining the microphysical property profile of aerosols, characterized in that, The method comprises: obtaining monitoring data of atmospheric whole-column average aerosol volume size spectrum distribution, real part and imaginary part of multi-wavelength complex refractive index collected by a plurality of ground sites at different time periods; constructing a six-parameter distribution database of coarse-mode and fine-mode aerosols according to the monitoring data; clustering the six-parameter distribution database to generate a plurality of coarse-mode aerosol size spectrum distribution models and a plurality of fine-mode aerosol size spectrum distribution models; for each of the aerosol size spectrum distribution models, calculating a backscattering coefficient calculation value at a preset wavelength based on Mie scattering theory; using a preset cost function, performing layer-by-layer iterative comparison between the backscattering coefficient calculation value and a backscattering coefficient observation value at a corresponding wavelength measured by a spaceborne lidar at different altitudes, and inversely deriving an aerosol microphysical property profile.

2. The method of claim 1, wherein, The method comprises: separating the aerosol volume size spectrum distribution in each monitoring data into coarse-mode volume distribution data and fine-mode volume distribution data by taking a preset aerosol particle radius as a boundary; using a bimodal lognormal distribution function, fitting the coarse-mode volume distribution data and the fine-mode volume distribution data respectively to obtain six-parameter distribution data; wherein the six-parameter distribution data comprises a peak radius, a peak height, a standard deviation, a particle size distribution, a real part of a complex refractive index, and an imaginary part of the complex refractive index.

3. The method of claim 2, wherein, The bimodal lognormal distribution function is: wherein, represents the volume size distribution of the particles, represents the volume peak height of the first mode, represents the particle radius, represents the peak radius of the first mode, represents the standard deviation of the first mode, = 1 corresponds to a fine mode, = 2 corresponds to a coarse mode.

4. The method of claim 1, wherein, The method comprises: normalizing the peak radius, the peak height, and the standard deviation in the six-parameter distribution data; based on the normalized parameters, using a clustering algorithm with the objective of minimizing the total dissimilarity J to calculate the dissimilarity between the parameters and the cluster centers using Euclidean distance for grouping, and generating a predetermined number of aerosol size spectrum distribution models; wherein the total dissimilarity J is expressed as: wherein, is a preset number of models, is a normalized parameter value, is a th cluster center, is a parameter set belonging to a th cluster.

5. The method of claim 1, wherein, The method comprises: The backscattering coefficient calculation value of the preset wavelength is obtained by the following formula ; wherein, represents a particle volume size distribution, represents a size parameter of a particle, represents a particle radius, represents a backscattering efficiency factor of a particle, represents a predetermined wavelength is a real part of a complex refractive index at a predetermined wavelength is an imaginary part of a complex refractive index at a predetermined wavelength represents a calculated value.​​ 6. The method of claim 1, wherein, the preset wavelength includes 355 nm, 532 nm, and 1064 nm; the cost function is: wherein, represents the cost function calculation value, , , are the backscatter coefficient observations measured by the spaceborne lidar at 355 nm, 532 nm, 1064 nm wavelengths respectively, , , are the backscatter coefficient calculations at the corresponding wavelengths respectively, , , are the uncertainties of the backscatter coefficient observations at the corresponding wavelengths respectively and are expressed in terms of the hourly standard deviation of the observations, the upper index represents the observation.

7. The method of claim 1, wherein, The method comprises: performing the iterative comparison layer by layer from low to high in altitude; at each altitude layer, selecting a coarse-mode aerosol size spectrum distribution model, a fine-mode aerosol size spectrum distribution model, and a corresponding complex refractive index that minimize the value of the cost function as the inverse result of the layer; the aerosol microphysical property profile is composed of the inverse results of all altitude layers.

8. An aerosol microphysical property profile determination apparatus, characterized by, The device comprises: a monitoring data acquisition module configured to obtain monitoring data of atmospheric whole-column average aerosol volume size spectrum distribution, real part and imaginary part of multi-wavelength complex refractive index collected by a plurality of ground sites at different time periods; A database construction module is configured to construct a six-parameter distribution database of coarse-mode and fine-mode aerosols according to the monitoring data; A distribution model construction module is configured to cluster the six-parameter distribution database to generate a plurality of coarse-mode aerosol size spectrum distribution models and a plurality of fine-mode aerosol size spectrum distribution models; A calculation value determination module is configured to calculate, for each of the aerosol size spectrum distribution models, a backscattering coefficient calculation value at a preset wavelength based on Mie scattering theory; A microphysical property profile determination module is configured to iteratively compare, layer by layer, the backscattering coefficient calculation value and a corresponding backscattering coefficient observation value at the preset wavelength measured by a spaceborne lidar at different altitudes by using a preset cost function, and inversely derive an aerosol microphysical property profile.

9. A computer device, comprising: The computer readable storage medium stores a computer program, and the computer program is executed by the processor to perform the steps of the aerosol microphysical property profile determination method according to any one of claims 1 to 7. The computer readable storage medium stores a computer program, and the computer program is executed by the processor to perform the steps of the aerosol microphysical property profile determination method according to any one of claims 1 to 7.

10. A computer-readable storage medium, characterized in that, ​

Images