Methods, systems, media, and apparatuses based on simplified aerosol optical depth inversion

Abstract

The application discloses a method, system, medium and equipment based on simplified aerosol optical depth inversion, and relates to the technical field of aerosol satellite remote sensing, and the method comprises the following steps: acquiring ground surface multi-spectral remote sensing data and ground-based measured AOD data with different spatial resolutions; calculating the spatial distribution heterogeneity of aerosols under different spatial resolutions according to the ground surface multi-spectral remote sensing data, and taking the ground surface multi-spectral remote sensing data corresponding to the maximum value of the spatial distribution heterogeneity as target data; based on the processed target data, using a Newton iteration algorithm to perform AOD inversion on the basis of a SARA algorithm; limiting the ranges of single scattering albedo SSA and an asymmetry parameter G, testing SSA and G in all limited ranges according to a set step length based on a two-layer loop of the SARA algorithm, finding SSA and G that make the inverted AOD closest to the ground-based measured AOD, and obtaining the AOD inversion result of a research area. The application can reduce the iteration number and time consumption.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of aerosol satellite remote sensing, and particularly relates to a method, system, medium and equipment based on simplified aerosol optical depth inversion. BACKGROUND

[0002] Aerosol optical depth (AOD) refers to the degree of influence of aerosol particulate matter on solar radiation, and is a key parameter for calculating aerosol characteristics and atmospheric correction. In recent years, with the development of satellite remote sensing technology, the development of aerosol inversion methods has also been driven. However, the complex surface environment, uncertainty of cloud masking method and different calibration methods all have an impact on satellite inversion of AOD. Aerosol inversion methods can be mainly divided into polar orbit and geostationary orbit. Polar orbit satellite inversion methods include single-view method, multi-view method, polarization feature inversion method, radar inversion method and multi-sensor collaborative inversion method. In the geostationary satellite inversion method, the main satellites include GOES-W, GOES-E, Himawari-8 and Meteosat-8, which can produce multi-temporal AOD products. Since they can simultaneously monitor dynamic aerosols and surface reflection information and have a constant view, many methods have been developed, the most classic of which is the GOES Aerosol / Smoke Product (GASP). So far, in the aerosol satellite remote sensing inversion method, the single-view method is the most common. The early development is the contrast reduction method, which is based on the contrast modulation function, the geometric position of the sun sensor, the radiation brightness values of two different targets at different albedos, the average surface reflectivity and a specific aerosol model for retrieving AOD. However, the more mature method developed in the single-view method is the MODIS operational retrieval method: from the earliest dark dense vegetation (DDV) method to the current dark target (DT), although this method works well in areas with dark vegetation cover, its accuracy is not good on bright ground. In view of this problem, in 2004, scholars proposed the deep blue (DB) method, which can effectively improve the accuracy of the dark target (DT) method on bright ground.

[0003] Scholars developed a simplified aerosol retrieval algorithm (SARA), which uses MODIS primary and secondary data to retrieve AOD at a spatial resolution of 500 m, without the need for lookup tables. Compared with the dark target method and the deep blue method, this method has higher accuracy and spatial coverage, and can improve the retrieval ability of urban AOD. However, the key parameters of this method, the single albedo and the geometric symmetry factor, are obtained from AERONET sites, so the implementation of this method is hindered in areas where AERONET sites are not available. Therefore, researchers have studied and optimized this defect of the SARA method. Among them, researchers proposed to analyze the sensitivity of the single albedo and the geometric symmetry factor, and calculate the empirical value of the single albedo and the symmetry factor, which has a high correlation with the MOD04 AOD product. Although this method solves the problem of AOD retrieval in areas without AERONET parameters, the retrieval accuracy of this method is lower than that of the SARA method. Researchers proposed to simulate the single scattering albedo and the asymmetry factor using the optical properties of aerosols and clouds (OPAC), and to test the inversion of the region using the improved method, obtaining the 500 m AOD product of the region from 2012 to 2014. Compared with the 3 km AOD and 10 km AOD products, the accuracy is significantly improved, increasing the applicability of the method. The SARA method uses fixed-point iteration to retrieve AOD in the iteration process, but this iteration method is not suitable for the nonlinear radiation transfer equation of aerosol inversion, which requires multiple iterations to converge, and the inversion of high-resolution data is also relatively time-consuming. In addition, the input data in the atmospheric inversion process is not fine enough, and these problems need to be solved. Therefore, the SARA method needs to be improved. SUMMARY

[0004] The purpose of the present application is to solve the problem of time-consuming iteration method of SARA, and to propose a method based on simplified aerosol optical thickness inversion, comprising the following steps:

[0005] S1, obtaining ground-based multi-spectral remote sensing data of different spatial resolutions, and ground-based measured aerosol optical thickness data;

[0006] S2, calculating the aerosol spatial distribution heterogeneity at different spatial resolutions according to the ground-based multi-spectral remote sensing data, and selecting the ground-based multi-spectral remote sensing data corresponding to the maximum aerosol spatial distribution heterogeneity as the target data;

[0007] S3, processing the target data, and based on the processed target data, using Newton iteration algorithm to retrieve aerosol optical thickness based on the simplified aerosol retrieval algorithm;

[0008] S4. Limit the range of single scattering albedo SSA and asymmetry parameter G. Based on the two-layer loop of the simplified aerosol inversion algorithm, test SSA and G within all restricted ranges according to the set step size, find the SSA and G that make the inverted aerosol optical thickness closest to the ground-based measured aerosol optical thickness, and use the closest SSA and G to obtain the aerosol optical thickness inversion result of the study area.

[0009] Furthermore, surface multispectral remote sensing data of different spatial resolutions are obtained from Sentinel-2 data.

[0010] Furthermore, processing the target data includes:

[0011] Band selection is performed on the target data, and the target data is divided into two levels: L1C and L2A. The L1C level data includes: top atmosphere reflectivity, solar zenith angle, satellite azimuth, and observation zenith angle; the L2A level data includes: bottom atmosphere reflectivity;

[0012] The target data after band selection is projected into the same coordinate system and resampled using bilinear interpolation; the atmospheric bottom reflectance mask data is grid-subdivided and clipped; and the processed target data is obtained.

[0013] Furthermore, when the ground-based aerosol optical depth data are missing, the average value of the data within one hour before and after the ground-based measured data is used to fill the gap.

[0014] Furthermore, the bands of the ground-based measured aerosol optical depth data need to be interpolated, which can be expressed as:

[0015] τ a =βλ -α

[0016]

[0017] Among them, τ a represents the known aerosol optical depth value of the band, β represents Turbidity coefficient, λ represents the band that needs to be interpolated, α represents The index, τ(λ), represents the aerosol optical depth value of the requested band, λ a Indicates a known band.

[0018] Furthermore, based on the simplified aerosol inversion algorithm, the Newton iteration algorithm is used to invert the aerosol optical depth, specifically:

[0019]

[0020] in, represents the aerosol optical depth under Newton iteration, τa represents the aerosol optical thickness of a known waveband, f(τ a ) represents a function constructed with respect to τ a a prime represents the derivative of f(τ a ) with respect to τ a s represents the cosine of the solar zenith angle, μ v represents the cosine of the view zenith angle, w0 represents the single scattering albedo, g represents an asymmetry parameter, represents the scattering phase angle, represents the top of atmospheric reflectance, represents the Rayleigh reflectance, represents the downward transmittance, represents the upward transmittance, represents the surface reflectance, S (λ) represents the atmospheric backscatter ratio, represents the derivative of f(τ ) with respect to τ a represents the derivative of f(τ ) with respect to τ a R represents the Rayleigh optical depth, represents the derivative of S(τ (λ) ) with respect to τ a s represents the central value of the corresponding wavelength, θ v represents the view zenith angle, φ represents the relative azimuth angle.

[0021] The application further provides a system based on simplified aerosol optical thickness inversion, comprising a data acquisition module, a data selection module, a data processing module and a result obtaining module.

[0022] The data acquisition module is used for acquiring ground multi-spectral remote sensing data of different spatial resolutions and ground-based measured aerosol optical thickness data.

[0023] The data selection module is used for calculating aerosol spatial distribution heterogeneity under different spatial resolutions according to the ground multi-spectral remote sensing data, and selecting ground multi-spectral remote sensing data corresponding to the maximum aerosol spatial distribution heterogeneity as target data.

[0024] The data processing module is used for processing the target data, and based on the processed target data, performing aerosol optical thickness inversion by using a Newton iteration algorithm on the basis of a simplified aerosol inversion algorithm.

[0025] ​​​​​​The result obtaining module is configured to limit the range of single scattering albedo (SSA) and asymmetry parameter (G), test SSA and G in all limited ranges according to a set step based on a two-layer loop of the simplified aerosol inversion algorithm, find SSA and G that make the inversion aerosol optical thickness closest to the ground-based measured aerosol optical thickness, and obtain the aerosol optical thickness inversion result of the study area by using the closest SSA and G.

[0026] The application further provides a computer readable storage medium, which stores a computer program, and the computer program is executed by a processor to implement the method for aerosol optical thickness inversion based on simplification.

[0027] The application further provides an electronic device, which comprises a processor and a memory, and the processor and the memory are connected to each other, wherein the memory is configured to store a computer program, the computer program comprises computer readable instructions, and the processor is configured to invoke the computer readable instructions to implement the method for aerosol optical thickness inversion based on simplification.

[0028] The application provides the technical scheme and brings the beneficial effects:

[0029] The application modifies the SARA method to adopt fixed-point iteration for AOD inversion in the iteration process on the basis of SARA, changes the nonlinear radiation transfer equation that is not applicable to aerosol inversion into the Newton iteration method applicable to nonlinear equations, and reduces the iteration number and time consumption. BRIEF DESCRIPTION OF DRAWINGS

[0030] Figure 1 is a flowchart of the method for aerosol optical thickness inversion based on simplification according to an embodiment of the application;

[0031] Figure 2 is a block diagram of an electronic device in an example embodiment according to the first embodiment of the application;

[0032] Figure 3 is a diagram for verifying the accuracy of SARA and XSARA (the method of the application) by using AOD data of A, B and C sites according to the embodiment of the application, wherein, Figure 3 (a) is the performance of SARA verified by using AOD data of the A site, (b) is the performance of SARA verified by using AOD data of the B site, (c) is the performance of SARA verified by using AOD data of the C site, (d) is the performance of XSARA verified by using AOD data of the A site, (e) is the performance of XSARA verified by using AOD data of the B site, and (f) is the performance of XSARA verified by using AOD data of the C site. DETAILED DESCRIPTION

[0033] In order to make the objects, technical solutions and advantages of the present application clearer, the embodiments of the present application will be further described below with reference to the drawings.

[0034] The flowchart of the method based on the simplified aerosol optical depth inversion of the embodiments of the present application is as shown in Figure 1 , and specifically includes the following steps:

[0035] S1, obtaining ground multi-spectral remote sensing data of different spatial resolutions and ground-based measured aerosol optical depth data.

[0036] Specifically, in the preferred embodiments of the present application, the ground multi-spectral remote sensing data of different spatial resolutions adopts Sentinel-2 series data. Sentinel-2 is a polar orbit multi-spectral high-resolution imaging mission for land monitoring to provide images of, for example, vegetation, soil and water cover, inland waterways and coastal areas.

[0037] The Sentinel-2 data products are mainly divided into Level-1B, Level-1C and Level-2A.

[0038] Level-1B is the original data, and the Level-1B product needs professional correction technology;

[0039] Level-1C is the product after the application of radiation and geometric correction (including orthorectification and spatial registration), and the 1C-level image is a set of 100 square kilometers of magnetic stickers, each of which is about 500 MB.

[0040] Level-2A is the product after the atmospheric correction and orthorectification by Sen2Cor.

[0041] The Sentinel-2 satellite carries a multi-spectral instrument (MSI) that can cover 13 spectral bands with ground resolutions of 10m, 20m and 60m.

[0042] The ground-based measured aerosol optical depth data (AOD) adopts AERONET AOD data, and the average of the AOD inversion results within 3km x 3km of the ground station and the average of the AOD results within half an hour before and after the ground station data are used for accuracy verification. In addition, when the ground station data is missing, the average of the data within one hour before and after the ground station data can be used.

[0043] The AERONET AOD data product does not have the Sentinel-2 band, so the satellite-inverted AOD and the ground-based measured AOD cannot be directly compared. Therefore, in order to obtain the AOD in the same band as Sentinel-2, the band of the AERONET AOD data must be interpolated. It is assumed that the atmosphere is quasi-stationary on the time scale of observation, and that τa and λ satisfy the following relationship:

[0044] τ a = βλ -α

[0045]

[0046] wherein τ a represents a known band aerosol optical depth value, β represents a turbidity coefficient, λ represents a band requiring interpolation, α represents an index, τ(λ) represents a requested band aerosol optical depth value, and λ a represents a known band.

[0047] S2, calculating the aerosol spatial distribution heterogeneity under different spatial resolutions according to the ground surface multi-spectral remote sensing data, and selecting ground surface multi-spectral remote sensing data corresponding to a maximum average local variance of the aerosol spatial distribution heterogeneity as target data.

[0048] Specifically, in the preferred embodiment of the present application, the Sentinel-2 data with resolutions of 10 m, 20 m and 60 m are respectively used to calculate the aerosol spatial distribution heterogeneity, the Sentinel-2 data with strong aerosol spatial distribution heterogeneity is the target data, and the corresponding resolution is the optimal resolution. Wherein, the local variance method is used to calculate the aerosol spatial distribution heterogeneity.

[0049] S3, processing the target data, selecting bands of the target data, and dividing the target data into L1C and L2A two levels, wherein the L1C level data includes: atmospheric top-of-atmosphere reflectance (TOA), solar zenith angle (SZA), satellite azimuth angle (SAA), and observation zenith angle (VZA); and the L2A level data includes: atmospheric bottom-of-atmosphere reflectance (BOA).

[0050] The target data after band selection is projected to the same coordinate system and resampled using bilinear interpolation; the BOA mask data is mesh subdivided and cropped; and the processed target data is obtained.

[0051] In the preferred embodiment of the present application, the band calculation expression used is "if (scl_cloud_high_proba>0 (high cloud probability greater than 0) or scl_thin_cirrus>0 (thin cloud probability greater than 0) or scl_cloud_medium_proba>0 (medium cloud probability greater than 0) or scl_snow_ice>0 (snow ice probability greater than 0) or ((B3-B8) / (B3+B8))>0.0) then set pixel value to 0, otherwise set pixel value to 1", and the BOA mask data (ice, snow, cloud, water) is meshed and cropped. These bands are stacked into a multi-band file. In the above expression, "scl_cloud_high_proba" represents the presence of high-level clouds; "scl_thin_cirrus" represents cirrus clouds; "scl_cloud_medium_proba" represents medium cloud probability, and "scl_snow_ice" represents ice and snow.

[0052] (B3-B8) / (B3+B8))>0.0" is used to detect water bodies. When this condition is met, i.e. (B3-B8) / (B3+B8) is greater than 0.0, it indicates that the pixel is a water body, and at this time, the pixel value is set to 0, which is masked. When this condition is not met, i.e. (B3-B8) / (B3+B8) is less than or equal to 0.0, it indicates that the pixel is not a water body. At this time, the pixel value is set to 1, which is retained for subsequent aerosol optical depth (AOD) retrieval. This condition is used to identify and mask water body pixels to ensure that only non-water body pixels participate in the AOD retrieval calculation. B3 and B8 represent different bands of the remote sensing image. B3 represents the blue band, and B8 represents the near-infrared band.

[0053] The above steps mask the cloud, snow, ice and water body, and only the effective pixels are retained for subsequent aerosol optical depth (AOD) retrieval.

[0054] Based on the processed target data, the Newton iteration algorithm is used for aerosol optical thickness retrieval on the basis of a simplified aerosol retrieval algorithm. Fixed-point iteration is generally only convergent for linear equations, while the Newton iteration algorithm is suitable for solving nonlinear equations and complex surface AOD retrieval.

[0055] In the preferred embodiment of the present application, the Newton iteration algorithm is specifically:

[0056]

[0057]

[0058] wherein, represents the aerosol optical thickness under Newton iteration, τ arepresents the aerosol optical depth of the known band, f(τ a ) represents the constructed a The function of f(τ a ) prime represents f(τ a ) for τ a The derivative of μ s represents the cosine of the solar zenith angle, μ v represents the cosine of the apparent zenith angle, w0 represents the single scattering albedo, g represents the asymmetry parameter, represents the scattering phase angle, represents the top of the atmospheric reflectivity, represents the Rayleigh reflectivity, represents the downward transmittance, represents the upper transmittance, represents the surface reflectivity, S (λ) represents the atmospheric backscatter ratio, express For τ a The derivative of express For τ a The derivative of τ R represents the Rayleigh optical depth, Indicates S (λ) For τ a The derivative of ,λ represents the center value of the corresponding wavelength, θ s represents the sun's zenith angle, θ v represents the vertical angle, and φ represents the relative azimuth.

[0059] Scattering phase angle The calculation is as follows:

[0060]

[0061] Atmospheric pressure is calculated as follows:

[0062] P Z =P0*(1-Gr*Z*10 3 / (Cp*T)) Cp*M0 / R ,

[0063] Among them, P Z represents the surface atmospheric pressure, P0 represents the atmospheric pressure at sea level, Gr represents the Earth's gravitational acceleration, and its value is 9.807 (m / s^2). Z represents the elevation obtained from the SRTM DEM data. Cp represents the specific heat at constant pressure. T represents the standard temperature. M0 represents the molar mass of dry air. R represents the atmospheric constant.

[0064] S4, limiting the range of single scattering albedo SSA and asymmetry parameter G, testing SSA and G in all limited ranges according to a set step length based on a two-layer loop of the simplified aerosol inversion algorithm, finding SSA and G that make the inverted aerosol optical thickness closest to the ground-based measured aerosol optical thickness, and obtaining the aerosol optical thickness inversion result of the study area by using the closest SSA and G.

[0065] In the preferred embodiment of the present application, the range of single scattering albedo SSA and asymmetry parameter G is limited to [0.5-1.0], and the step length is set to 0.001.

[0066] The present application also provides a system based on simplified aerosol optical thickness inversion, comprising a data acquisition module, a data selection module, a data processing module, and a result obtaining module.

[0067] The data acquisition module is configured to acquire ground-based multi-spectral remote sensing data of different spatial resolutions and ground-based measured aerosol optical thickness data.

[0068] The data selection module is configured to calculate aerosol spatial distribution heterogeneity at different spatial resolutions according to the ground-based multi-spectral remote sensing data, and select ground-based multi-spectral remote sensing data corresponding to the resolution of the maximum aerosol spatial distribution heterogeneity as target data.

[0069] The data processing module is configured to process the target data, and perform aerosol optical thickness inversion based on the simplified aerosol inversion algorithm and using the Newton iteration algorithm based on the processed target data.

[0070] The result obtaining module is configured to limit the range of single scattering albedo SSA and asymmetry parameter G, test SSA and G in all limited ranges according to a set step length based on a two-layer loop of the simplified aerosol inversion algorithm, find SSA and G that make the inverted aerosol optical thickness closest to the ground-based measured aerosol optical thickness, and obtain the aerosol optical thickness inversion result of the study area by using the closest SSA and G.

[0071] In an exemplary embodiment, the present application provides a computer readable storage medium storing a computer program, wherein the computer program is executed by a processor to implement the above-mentioned method based on simplified aerosol optical thickness inversion.

[0072] Please refer to Figure 2 In an exemplary embodiment, the present application also provides an electronic device comprising at least one processor, at least one memory, and at least one communication bus.

[0073] The memory stores a computer program, and the computer program includes computer readable instructions.

[0074] Compared with the simplified aerosol retrieval method (SARA), the AOD retrieval (XSARA) of the application has higher precision and better consistency with AERONET AOD, and in the final result, the AOD data of A, B and C are used to verify the accuracy of the method (XSARA) of the application. Figure 3 , Figure 3 The figure for verifying the accuracy of SARA and XSARA (the method of the application) of the embodiments of the application uses the AOD data of A, B and C sites, wherein, Figure 3 (a) is the performance of SARA verified by using the AOD data of A site, (b) is the performance of SARA verified by using the AOD data of B site, (c) is the performance of SARA verified by using the AOD data of C site, (d) is the performance of XSARA verified by using the AOD data of A site, (e) is the performance of XSARA verified by using the AOD data of B site, and (f) is the performance of XSARA verified by using the AOD data of C site. The inversion results of SARA and XSARA have higher correlation with the observation data of the ground station, but the RMSE and MAE of SARA are larger than those of XSARA. In addition, the results of SARA exceed the EE interval more, while XSARA reaches the EE interval. At the same time, the correlation of the inversion results of XSARA and AERONET AOD data is more than 0.98. In addition, the fitting straight line slope of the AOD inversion results of XSARA and AERONET AOD data at A, B and C sites is 0.9662, 1.0022 and 1.0002 respectively, indicating that the linear fitting of the two is good.

[0075] The above description of disclosed embodiments enables a person skilled in the art to implement or use the present application. Various modifications to these embodiments will be apparent to those skilled in the art, and the general principles defined herein can be implemented in other embodiments without departing from the spirit or scope of the present application. Therefore, the present application will not be limited to these embodiments shown herein, but will conform to the widest scope consistent with the principles and novel features disclosed herein.

Claims

1. A method for simplified aerosol optical thickness inversion, characterized in that, The method comprises the following steps: S1, obtaining ground multi-spectral remote sensing data of different spatial resolutions and ground-based measured aerosol optical thickness data; S2, calculating aerosol spatial distribution heterogeneity under different spatial resolutions according to the ground multi-spectral remote sensing data, and selecting ground multi-spectral remote sensing data corresponding to the resolution of the maximum aerosol spatial distribution heterogeneity as target data; S3, processing the target data, and based on the processed target data, using Newton iteration algorithm to perform aerosol optical thickness inversion on the basis of a simplified aerosol inversion algorithm; S4, limiting the range of single scattering albedo SSA and asymmetric parameter G, testing SSA and G in all limited ranges according to a set step length based on two-layer circulation of the simplified aerosol inversion algorithm, finding SSA and G closest to the inversion of the aerosol optical thickness and the ground-based measured aerosol optical thickness, and obtaining the aerosol optical thickness inversion result of the research area by using the closest SSA and G.

2. The method of claim 1, wherein, The ground multi-spectral remote sensing data of different spatial resolutions are obtained from Sentinel-2 data.

3. The method of claim 2, wherein, The processing of the target data comprises: performing band selection on the target data, and dividing the target data into L1C and L2A two levels, wherein the L1C level data comprises: atmospheric top layer reflectivity, solar zenith angle, satellite azimuth angle, and observation zenith angle; and the L2A level data comprises: atmospheric bottom reflectivity; projecting the target data after the band selection to the same coordinate system and resampling by using bilinear interpolation; and performing mesh subdivision and cutting on the atmospheric bottom reflectivity mask data; and obtaining the processed target data.

4. The method of claim 1, wherein, When the ground-based measured aerosol optical thickness data is missing, the average value of the data within one hour before and after the ground-based measured data is used for filling.

5. The method of claim 1, wherein, The band of the ground-based measured aerosol optical thickness data needs to be interpolated and is expressed as: τ a = βλ -α where τ a represents the known band aerosol optical thickness value, β represents the turbidity coefficient, λ represents the band for which interpolation is required, α represents the exponent, τ(λ) represents the requested band aerosol optical thickness value, λ a represents the known band.

6. The method of inverting based on a simplified aerosol optical thickness according to claim 1, wherein, On the basis of the simplified aerosol inversion algorithm, the Newton iteration algorithm is used to perform aerosol optical thickness inversion, and the specific process is as follows: in, represents the aerosol optical depth under Newton iteration, τ a represents the aerosol optical depth of the known band, f(τ a ) represents the constructed a Function, f(τ a ) prime represents f(τ a ) for τ a The derivative of μ s represents the cosine of the solar zenith angle, μ v represents the cosine of the apparent zenith angle, w0 represents the single scattering albedo, g represents the asymmetry parameter, represents the scattering phase angle, ρ TOA (λ,θ s ,θ v , φ) represents the top of the atmospheric reflectivity, represents the Rayleigh reflectivity, represents the downward transmittance, represents the upper transmittance, represents the surface reflectivity, S (λ) represents the atmospheric backscatter ratio, express For τ a The derivative of express For τ a The derivative of τ R represents the Rayleigh optical depth, Indicates S (λ) For τ a The derivative of λ represents the center value of the corresponding wavelength, θ s represents the sun's zenith angle, θ v represents the vertical angle, and φ represents the relative azimuth.

7. A system for aerosol optical depth retrieval based on simplification, characterized in that, It comprises: a data acquisition module, a data selection module, a data processing module, and a result obtaining module; the data acquisition module is used to obtain ground multi-spectral remote sensing data of different spatial resolutions and ground-based measured aerosol optical thickness data; the data selection module is used to calculate aerosol spatial distribution heterogeneity under different spatial resolutions according to the ground multi-spectral remote sensing data, and select ground multi-spectral remote sensing data corresponding to the resolution of the maximum aerosol spatial distribution heterogeneity as target data; the data processing module is used to process the target data, and based on the processed target data, using Newton iteration algorithm to perform aerosol optical thickness inversion on the basis of a simplified aerosol inversion algorithm; the result obtaining module is used to limit the range of single scattering albedo SSA and asymmetric parameter G, test SSA and G in all limited ranges according to a set step length based on two-layer circulation of the simplified aerosol inversion algorithm, find SSA and G closest to the inversion of the aerosol optical thickness and the ground-based measured aerosol optical thickness, and obtain the aerosol optical thickness inversion result of the research area by using the closest SSA and G.

8. A computer-readable storage medium, the computer-readable storage medium storing a computer program, characterized in that: The computer program, which is executed by a processor, implements the method according to any one of claims 1-6.

9. An electronic device, comprising: A computer program product, comprising a processor and a memory, which are interconnected, wherein the memory is configured to store a computer program, the computer program comprising computer readable instructions, and the processor is configured to invoke the computer readable instructions to execute the method according to any one of claims 1-6.

Images