Aerosol parameter inversion method and device based on active and passive collaborative detection data

Abstract

The present invention discloses an aerosol parameter inversion method and device based on active and passive collaborative detection data, and relates to the field of aerosol measurement technology. The method is aimed at aerosol information detected by active and passive loads collaboratively, starting from the characteristics of different loads, and based on the constraint relationship of collaborative detection data in radiation / polarization / spatial characteristics, a multi-parameter inversion method such as aerosol optical thickness and extinction coefficient profile based on active and passive load collaborative detection data is proposed. The method increases the total amount of information in aerosol parameter inversion, thereby optimizing the inversion accuracy.

Description

Technical Field

[0001] The present invention belongs to the technical field of aerosol measurement, and in particular relates to an aerosol parameter inversion method and device based on active and passive collaborative detection data. Background Art

[0002] Aerosols, which are widely present in the earth-atmosphere system, play an important role in the changes of the earth's climate and the human living environment. On the one hand, the radiative forcing effect caused by the absorption and scattering of solar radiation and ground long-wave radiation by aerosols is one of the biggest sources of uncertainty in the current climate change assessment; on the other hand, the formation of acid rain, the appearance of haze weather, and the destruction of the ozone layer are closely related to the harmful pollution characteristics of aerosols themselves. The increase in the aerosol content in the atmosphere will lead to reduced atmospheric visibility, aggravated environmental pollution problems, and damage to human health. However, the sources of aerosols are variable, the composition is complex, and the temporal and spatial distribution is uneven, which makes it extremely difficult to quantitatively evaluate their climate and environmental effects.

[0003] Satellite remote sensing is currently the most efficient and promising method for obtaining long-term global atmospheric spatial distribution information. However, due to the continuous vertical changes of the atmosphere and the coupling of surface and atmospheric remote sensing information, the number of unknowns required when inverting atmospheric parameters from satellite data is often greater than the number of equations that can be established based on remote sensing observation information. This pathological nature of the inversion restricts the inversion accuracy of atmospheric products. Different sensors differ in many aspects such as detection mode, band configuration, scanning method, and spatial resolution. Therefore, the use of collaborative detection data from multiple remote sensors can provide more remote sensing observation information, thereby reducing the pathological nature of the inversion and effectively improving the accuracy of atmospheric product inversion. Considering the complexity and variability of aerosol types, it is more difficult to extract aerosol information from remote sensing information.

[0004] Existing aerosol remote sensing detection methods mainly include passive and active types. Passive detection payloads include multi-channel imaging spectrometers, hyperspectral imagers, multi-angle imaging spectrometers, multi-angle polarization detectors, etc., which mainly use the detection information of the visible band in optical sensors to obtain aerosol distribution in a wide range; active detection payloads are mainly atmospheric lidars, which can invert the extinction profile characteristics of aerosols based on the backscattering information of aerosols at different altitudes. At the same time, considering that the polarization characteristics of atmospheric particle scattering include the phase state and radiation characteristics of particles, current aerosol detectors often introduce polarization observations.

[0005] However, the aerosol parameter inversion process based on passive detection payload is highly dependent on the preset aerosol model, which brings large errors to the inversion of parameters such as aerosol optical thickness and fine particle aerosol optical thickness. To obtain high-precision aerosol products, it is necessary to optimize the preset aerosol model.

[0006] Considering the current situation that the atmospheric signal of active detection payload atmospheric lidar is weak and the signal-to-noise ratio is low, the radiation calibration accuracy of atmospheric lidar is low, which affects the accuracy of aerosol extinction coefficient profile inversion. It is necessary to introduce other constraints in the inversion process to weaken the error introduced by payload calibration.

[0007] In the prior art, different methods for detecting atmospheric aerosol parameters using different payloads have their own advantages and focuses, and there is no aerosol inversion algorithm constructed based on the constraint relationship between active and passive collaborative detection data. Summary of the invention

[0008] In order to solve the above technical problems, the present invention provides an aerosol parameter inversion method and device based on active and passive collaborative detection data. The method is aimed at the aerosol information detected by active and passive loads, and based on the constraint relationship of the collaborative detection data in radiation / polarization / spatial characteristics, it realizes high-precision inversion of aerosol optical thickness, extinction coefficient profile and other information.

[0009] In order to achieve the above object, the present invention adopts the following technical scheme:

[0010] An aerosol parameter inversion method based on active and passive collaborative detection data includes the following steps:

[0011] Step S1: spatially matching the active and passive detection payload observation data, extracting pixel point information matching the passive detection payload observation data with the active detection payload observation data, and obtaining spatially matched passive detection payload observation data; the active detection payload includes an atmospheric laser radar payload, and the passive detection payload includes any one of a multi-angle polarization detector, a multi-channel imaging spectrometer, a hyperspectral imager, and a multi-angle imaging spectrometer;

[0012] Step S2: Based on the active detection payload observation data, define the target layer range where the aerosol is located and identify the aerosol type in the target layer;

[0013] Step S3: Determine the aerosol inversion channel of the passive detection payload observation data; when the passive detection payload is a multi-angle polarization detector, based on the matching pixel point information, obtain the simulated apparent reflectivity and apparent polarization reflectivity in the scalar lookup table and polarization lookup table used for aerosol inversion, based on the target layer aerosol type information identified by the active detection payload observation data, using the residuals of the apparent reflectivity and the apparent polarization reflectivity as evaluation indicators, invert and obtain the aerosol model, aerosol optical thickness, fine particle aerosol optical thickness and fine particle ratio; when the passive detection payload is any one of a multi-channel imaging spectrometer, a hyperspectral imager, and a multi-angle imaging spectrometer, based on the matching pixel point information, obtain the simulated apparent reflectivity in the scalar lookup table used for aerosol inversion, based on the matching pixel point information, and based on the target layer aerosol type information identified by the active detection payload observation data, using the residual of the apparent reflectivity as an evaluation indicator, invert and obtain the aerosol model and aerosol optical thickness.

[0014] Step S4: Based on the target layer aerosol type identification results and aerosol model evaluation results of the active detection payload observation data, the lidar ratio corresponding to each target layer aerosol is obtained, and the aerosol extinction coefficient profile is obtained by inverting the aerosol optical thickness inverted from the passive detection payload observation data using the Fernald inversion method.

[0015] On the other hand, the present invention provides an aerosol parameter inversion device based on active and passive collaborative detection data, comprising:

[0016] An observation data matching unit is used to spatially match the observation data of the active and passive detection payloads, extract pixel point information matching the observation data of the passive detection payload and the observation data of the active detection payload, and obtain spatially matched observation data of the passive detection payload; the active detection payload includes an atmospheric laser radar payload, and the passive detection payload includes any one of a multi-angle polarization detector, a multi-channel imaging spectrometer, a hyperspectral imager, and a multi-angle imaging spectrometer;

[0017] The target layer aerosol type identification unit is used to define the target layer range where the aerosol is located and identify the target layer aerosol type based on the active detection payload observation data;

[0018] The aerosol parameter inversion unit is used to determine the aerosol inversion channel of the passive detection payload observation data; when the passive detection payload is a multi-angle polarization detector, based on the matching pixel information, the simulated apparent reflectivity and apparent polarization reflectivity are obtained in the scalar lookup table and polarization lookup table used for aerosol inversion, and based on the target layer aerosol type information identified by the active detection payload observation data, the residuals of the apparent reflectivity and the apparent polarization reflectivity are used as evaluation indicators to invert and obtain the aerosol model, aerosol optical thickness, fine particle aerosol optical thickness and fine particle ratio; when the passive detection payload is any one of a multi-channel imaging spectrometer, a hyperspectral imager and a multi-angle imaging spectrometer, based on the matching pixel information, the simulated apparent reflectivity is obtained in the scalar lookup table used for aerosol inversion, and based on the target layer aerosol type information identified by the active detection payload observation data, the residuals of the apparent reflectivity are used as evaluation indicators to invert and obtain the aerosol model and aerosol optical thickness;

[0019] It is also used to identify the target layer aerosol type and evaluate the aerosol model based on the active detection payload observation data, obtain the lidar ratio corresponding to each target layer aerosol, and use the Fernald inversion method to obtain the aerosol extinction coefficient profile based on the aerosol optical thickness inverted from the passive detection payload observation data as a constraint.

[0020] In a third aspect, the present invention provides an electronic device, comprising: one or more processors; a memory for storing one or more programs; wherein, when the one or more programs are executed by the one or more processors, the one or more processors implement the aforementioned aerosol parameter inversion method based on active and passive collaborative detection data.

[0021] In a fourth aspect, the present invention provides a computer-readable storage medium having executable instructions stored thereon, which, when executed by a processor, enables the processor to implement the aforementioned aerosol parameter inversion method based on active and passive collaborative detection data.

[0022] The beneficial effects of the present invention are:

[0023] The method proposed in the present invention takes into account the constraint relationship between active and passive collaborative detection data in radiation / polarization / spatial characteristics, thereby increasing the total amount of information for aerosol parameter inversion; compared with traditional inversion methods, the method utilizes active and passive collaborative detection data to weaken the load calibration error and the aerosol model preset deviation, thereby improving the parameter inversion accuracy. BRIEF DESCRIPTION OF THE DRAWINGS

[0024] Figure 1 This is a flow chart of the aerosol parameter inversion method based on active and passive collaborative detection data of the present invention;

[0025] Figure 2Active and passive collaborative detection data matching flow chart;

[0026] Figure 3 Flowchart for target layer definition and identification based on active detection payload observation data;

[0027] Figure 4 A flow chart for inverting aerosols from passive detection payload observation data optimized based on active detection information;

[0028] Figure 5 Flowchart for inverting aerosols from active detection payload observation data constrained by passive detection information. DETAILED DESCRIPTION

[0029] The present invention will be further described below with reference to the accompanying drawings and embodiments.

[0030] The main purpose of the present invention is to provide an aerosol parameter inversion method for active and passive collaborative detection data. The method is based on the aerosol information of active and passive collaborative detection, and based on the constraint relationship between the collaborative detection data in radiation / polarization / spatial characteristics, it can achieve high-precision inversion of parameters such as aerosol optical thickness and extinction coefficient profile.

[0031] like Figure 1 As shown, the aerosol parameter inversion method of active and passive collaborative detection data of the present invention includes the following steps:

[0032] Step S1: spatially matching the active and passive detection payload observation data, extracting pixel point information matching the passive detection payload observation data with the active detection payload observation data, and obtaining spatially matched passive detection payload observation data; the active detection payload includes an atmospheric laser radar payload, and the passive detection payload includes any one of a multi-angle polarization detector, a multi-channel imaging spectrometer, a hyperspectral imager, and a multi-angle imaging spectrometer;

[0033] Step S2: Based on the active detection payload observation data, define the target layer range where the aerosol is located and identify the aerosol type in the target layer;

[0034] Step S3: Determine the aerosol inversion channel of the passive detection payload observation data; when the passive detection payload is a multi-angle polarization detector, based on the matching pixel point information, obtain the simulated apparent reflectivity and apparent polarization reflectivity in the scalar lookup table and polarization lookup table used for aerosol inversion, based on the target layer aerosol type information identified by the active detection payload observation data, using the residuals of the apparent reflectivity and the apparent polarization reflectivity as evaluation indicators, invert and obtain the aerosol model, aerosol optical thickness, fine particle aerosol optical thickness and fine particle ratio; when the passive detection payload is any one of a multi-channel imaging spectrometer, a hyperspectral imager, and a multi-angle imaging spectrometer, based on the matching pixel point information, obtain the simulated apparent reflectivity in the scalar lookup table used for aerosol inversion, based on the matching pixel point information, and based on the target layer aerosol type information identified by the active detection payload observation data, using the residual of the apparent reflectivity as an evaluation indicator, invert and obtain the aerosol model and aerosol optical thickness.

[0035] Step S4: Based on the target layer aerosol type identification results and aerosol model evaluation results of the active detection payload observation data, the lidar ratio corresponding to each target layer aerosol is obtained, and the aerosol extinction coefficient profile is obtained by inverting the aerosol optical thickness inverted from the passive detection payload observation data using the Fernald inversion method.

[0036] Specifically, the step S1 includes:

[0037] Since the active detection payload includes an atmospheric lidar payload, and the passive detection payload includes any one of a multi-angle polarization detector, a multi-channel imaging spectrometer, a hyperspectral imager, and a multi-angle imaging spectrometer, and the data acquisition methods of the above-mentioned passive detection payloads are similar, the present invention selects the collaborative detection data based on the multi-angle polarization imager and the atmospheric lidar payload in step S1 (the above-mentioned passive detection payloads are operated according to a similar process and will not be repeated here) for spatial matching, extracts the longitude and latitude coordinates of the atmospheric lidar foot point, matches the pixel point closest to the longitude and latitude coordinates in the observation data of the multi-angle polarization imager, and extracts the observation information of the pixel point. The process is as follows: Figure 2 As shown, the specific steps include:

[0038] Step S1.1, based on the multi-angle polarization imager observation data and the atmospheric lidar payload observation data, respectively extract the longitude and latitude coordinates of the multi-angle polarization imager observation data and the longitude and latitude coordinates of the foot points of the atmospheric lidar payload observation data.

[0039] Step S1.2, based on the latitude and longitude coordinates of the foot point of the atmospheric lidar payload observation data, match the pixel point closest to it in the observation data of the multi-angle polarization imager of collaborative detection.

[0040] Step S1.3, extracting the multi-angle polarization imager observation data corresponding to the matched pixel points described in step S1.2.

[0041] The step S2 comprises:

[0042] Since the active detection payload includes an atmospheric lidar payload, the present invention processes the atmospheric lidar observation data in step S2. First, based on the atmospheric lidar observation data, the atmospheric attenuation backscatter coefficient is calculated, and combined with the atmospheric attenuation backscatter coefficient in the simulated "clear sky" mode, a threshold is established to realize the recognition of the top and bottom position information of the aerosol target layer, thereby realizing the definition of the target layer where the aerosol is located, and further based on the backscatter coefficient, color ratio and surface type information of the target layer, the aerosol type of the target layer is identified. The process is as follows Figure 3 As shown, the specific steps include:

[0043] Step S2.1, definition of aerosol target layer;

[0044] The LiDAR equation can quantitatively describe the propagation process of laser pulse signals in the atmosphere. It is a mathematical expression of the working principle of the atmospheric LiDAR payload. It describes the energy of the atmospheric backscattered echo signals received at different altitudes when the laser beam emitted vertically by the atmospheric LiDAR payload passes through the atmosphere. Applying the Mie scattering theory and combining the working principle of LiDAR, the atmospheric backscattered echo power received by the atmospheric LiDAR payload at the detection distance Z is , which can be expressed by the Mie scattering lidar equation:

[0045] ;

[0046] Where β' is the atmospheric attenuation backscatter coefficient, P is the atmospheric lidar payload observation data, Z is the detection distance, C is the atmospheric lidar payload data calibration coefficient, E is the atmospheric lidar payload emission energy, and subscript indicates parallel polarization channels, the subscript Indicates the vertically polarized channel.

[0047] Based on the above formula, the calculation formula for the atmospheric attenuation backscatter coefficient of the atmospheric lidar payload observation data is as follows:

[0048] ;

[0049] In the formula represents the atmospheric backscattering coefficient, is the atmospheric transmittance, then is the atmospheric bidirectional transmittance, is the extinction coefficient, subscript represents atmospheric aerosol particles, subscript represents atmospheric molecules, subscript Represents the sum of atmospheric aerosol particles and atmospheric molecules.

[0050] In the "clear sky" mode, the atmosphere is assumed to contain only atmospheric molecules. The expected atmospheric molecule attenuation backscattering coefficient in this mode is , calculated and simulated by the following formula:

[0051] ;

[0052] Dividing the above two equations, we can get the attenuation scattering ratio coefficient. The expression is:

[0053] ;

[0054] Expected atmospheric attenuation backscatter coefficients in "clear sky" mode The threshold is established by the sum of the noise information of the atmospheric lidar payload. When the attenuation scattering ratio coefficient of continuous points above 150 meters in height exceeds the threshold, the height range of the continuous points is selected as the target layer.

[0055] Step S2.2, identification of aerosol type;

[0056] Using the three characteristic attributes of depolarization ratio, atmospheric attenuation backscattering coefficient ratio and underlying surface type obtained from atmospheric lidar payload observations, aerosols are classified into six types: dust type, biomass burning type, polluted continental type, ocean type, polluted dust type and continental type.

[0057] The step S3 comprises:

[0058] The present invention targets the passive detection payload including any one of a multi-angle polarization detector, a multi-channel imaging spectrometer, a hyperspectral imager, and a multi-angle imaging spectrometer, and performs aerosol parameter inversion based on passive detection payload observation data corresponding to matching pixel points.

[0059] When the passive detection payload is a multi-angle polarimetric imager, aerosol parameter inversion is carried out for scalar observation data and polarimetric observation data respectively, and the aerosol inversion channel is determined. Based on the geometric observation information of the matching pixel point observation data, the simulated apparent reflectivity and apparent polarized reflectivity are obtained in the scalar lookup table and polarization lookup table used for aerosol inversion, and compared with the apparent reflectivity and apparent polarized reflectivity actually observed by the multi-angle polarimetric imager. The preliminary evaluation results of the aerosol model are obtained in combination with the aerosol type information identified by the atmospheric lidar observation data. The residuals of the apparent reflectivity and apparent polarized reflectivity are used as evaluation indicators to invert the aerosol model, aerosol optical thickness, fine particle aerosol optical thickness and fine particle ratio.

[0060] When the passive detection payload is any one of a multi-channel imaging spectrometer, a hyperspectral imager, or a multi-angle imaging spectrometer, the aerosol parameter inversion of the scalar observation data is carried out, the aerosol inversion channel is determined, and the simulated apparent reflectivity is obtained in the scalar lookup table used for aerosol inversion based on the geometric observation information of the matching pixel point observation data. The simulated apparent reflectivity is compared with the actual observed apparent reflectivity, and the preliminary evaluation results of the aerosol model are obtained by combining the aerosol type information identified by the atmospheric lidar observation data. The residual of the apparent reflectivity is used as the evaluation index to invert the aerosol model and aerosol optical thickness.

[0061] Process such as Figure 4 As shown, the specific steps include:

[0062] Step S3.1, inversion based on scalar observation data;

[0063] The apparent reflectivity in the scalar observation data contains information about both the surface and the atmosphere. Assuming that the surface is a uniform Lambertian surface and ignoring the absorption of gas, the apparent reflectivity received by the load is It can be expressed as:

[0064] ;

[0065] In the formula, , , They are the solar zenith angle, satellite observation zenith angle, and relative azimuth angle. represents the equivalent reflectivity of the atmospheric path radiation term, T is the atmospheric transmittance, is the surface reflectivity, and S is the hemispherical reflectivity of the lower atmospheric boundary.

[0066] Therefore, in the inversion based on scalar observation data, the geometric observation information contained in the observation data is used to interpolate the scalar radiation contribution containing only atmospheric information corresponding to different aerosol types and observation channels in the scalar lookup table. Secondly, the surface reflectivity is calculated, and the simulated on-board apparent reflectivity is obtained in combination with the scalar radiation contribution of the atmosphere. The simulated on-board apparent reflectivity is compared with the apparent reflectivity actually observed by the payload, and the aerosol type information identified by the atmospheric lidar observation data is combined to preliminarily obtain several types of aerosol models that the observation data may correspond to. The residual of the apparent reflectivity is used as the evaluation index. Under the constraints of the aerosol model, the simulated data with the smallest evaluation index value is selected, and the corresponding aerosol type and aerosol optical thickness are the inversion results.

[0067] The construction of the scalar lookup table, the estimation method of the surface reflectivity and the calculation method of the evaluation index are as follows:

[0068] Step S3.11, establishing a scalar lookup table;

[0069] The present invention uses a scalar lookup table method to achieve the inversion of scalar observation data. The vector radiation transfer model 6SV is used to establish the scalar lookup table. By looping and calling 6SV, the scalar radiation information is obtained and saved to construct the scalar lookup table. The variables of the scalar lookup table include: spectral wavelength, aerosol model, aerosol optical thickness, solar zenith angle, observation zenith angle, and relative azimuth. When constructing the scalar lookup table, the contribution of the surface is ignored and the surface reflectivity is taken as Set to 0.

[0070] When constructing a scalar lookup table, the 6SV model parameters are specifically configured as follows:

[0071] a) Spectral wavelength: around 490nm and 670nm.

[0072] b) Aerosol model: Input custom aerosol model data that satisfies the bimodal normal distribution.

[0073] c) Aerosol optical depth: Input the aerosol optical depth at 550nm. The possible values ​​are 0.01, 0.1, 0.2, 0.3, 0.4, 0.5, 0.6, 0.7, 0.8, 0.9, 1.0, 1.25, 1.5, and 2.0 respectively.

[0074] d) Solar zenith angle: The solar zenith angle ranges from 0 to 80 degrees, with an interval of 3 degrees.

[0075] e) Observation zenith angle: The value range of the payload observation zenith angle is 0 to 80 degrees, with an interval of 3 degrees.

[0076] f) Relative azimuth: The relative azimuth between the sun and the payload ranges from 0 to 180 degrees, with an interval of 6 degrees.

[0077] Step S3.12, surface reflectivity estimation method;

[0078] The present invention uses the semi-empirical vegetation and bare soil linear combination model improved by von Hoyningen-Huene et al. to obtain the surface reflectance by scalar estimation. , the calculation formula is:

[0079] ;

[0080] ;

[0081] in, represents the reflectivity in the near-infrared band, represents the reflectivity of red light band, is the empirical weight coefficient, NDVI is the normalized difference vegetation index, is the wavelength. , They represent the spectral reflectance of vegetation and bare soil respectively. The corresponding data can be obtained through statistics of the ground object reflectance spectrum library.

[0082] Step 3.13, calculation method of evaluation index;

[0083] The residual of the apparent reflectivity is used as the evaluation index. The residual corresponding to a specific aerosol model is compared. When the residual is the smallest, the aerosol model and its corresponding aerosol total optical thickness are the inversion results. The evaluation index is published as follows:

[0084] ;

[0085] Where M is the number of bands and N is the number of angles. represents the observed reflectivity, Represents the simulated apparent reflectance.

[0086] Step S3.2, inversion based on polarization observation data;

[0087] Similar to the apparent reflectivity, the apparent polarized reflectivity received by the payload It also contains information about the surface and atmosphere, which can be expressed as:

[0088] ;

[0089] In the formula, is the atmospheric polarization reflectivity, is the surface polarization reflectivity, is the atmospheric mass number, is the molecular optical thickness of the atmosphere, is the total optical thickness of the aerosol, is an empirical coefficient, generally taken as 0.5. and atmospheric molecular optical thickness It can be calculated by the empirical formula:

[0090] ;

[0091] ;

[0092] In the formula, is the surface air pressure value.

[0093] The inversion process based on polarization observation data is basically the same as that of scalar observation data, which is used to obtain the optical thickness of aerosol fine particles. Using the geometric observation information in the polarization observation data, the polarization reflectivity containing only atmospheric information corresponding to different aerosol types and observation channels is interpolated in the polarization lookup table. Secondly, the surface polarization reflectivity is calculated. , combined with the polarization reflectivity of atmospheric information, the simulated on-board apparent polarization reflectivity is obtained. The simulated on-board apparent polarization reflectivity is compared with the apparent polarization reflectivity actually observed by the payload. Combined with the aerosol type information identified by the atmospheric lidar observation data, several types of aerosol models that the observation data may correspond to are preliminarily obtained. The residual of the apparent polarization reflectivity at different observation angles is used as the evaluation index. Under the constraints of the aerosol model, the simulated data with the smallest evaluation index value is selected, and the corresponding aerosol model and aerosol fine particle optical thickness are the inversion results.

[0094] The construction of the polarization lookup table and the estimation method of the surface polarization reflectivity are as follows:

[0095] Step S3.21, establishing a polarization lookup table;

[0096] The present invention uses a polarization lookup table method to realize the inversion of aerosol fine particle optical thickness. The polarization lookup table is constructed in the same way as the scalar lookup table. By calling 6SV in a loop, the radiation information is read and saved to generate a polarization lookup table. The variables of the polarization lookup table include: spectral wavelength, aerosol model, fine particle aerosol optical thickness, solar zenith angle, observation zenith angle, and relative azimuth. At the same time, the polarization contribution of the surface is ignored, and the surface polarization reflectivity is set to 0. The difference is that the aerosol model needs to be set to a fine particle type with a unimodal normal distribution according to the aerosol particle size.

[0097] When building the polarization lookup table, the 6SV model parameters are specifically configured as follows:

[0098] a) Spectral wavelength: around 670nm and 865nm.

[0099] b) Fine particle aerosol model: Input custom fine particle aerosol model data that satisfies a unimodal normal distribution.

[0100] c) Fine particle aerosol optical depth: Input the fine particle aerosol optical depth at 550nm. The possible values ​​are 0.01, 0.1, 0.2, 0.3, 0.4, 0.5, 0.6, 0.7, 0.8, 0.9, 1.0, 1.25, 1.5, and 2.0 respectively.

[0101] d) Solar zenith angle: The solar zenith angle ranges from 0 to 80 degrees, with an interval of 3 degrees.

[0102] e) Observation zenith angle: The value range of the payload observation zenith angle is 0 to 80 degrees, with an interval of 3 degrees.

[0103] f) Relative azimuth: The relative azimuth between the sun and the payload ranges from 0 to 180 degrees, with an interval of 6 degrees.

[0104] Step S3.22, surface polarization reflectivity estimation method;

[0105] The contribution of surface polarization to polarization inversion is small. It is generally believed that the surface polarization reflectivity does not change with wavelength. The semi-empirical model of Nadal and Breon is used to estimate:

[0106] ;

[0107] ;

[0108] in, represents the reflectivity in the near-infrared band, represents the reflectivity of red light band, is the Fresnel scattering coefficient, , are the cosine values ​​of the solar zenith angle and the observation zenith angle, k, It is an empirical coefficient determined by NDVI and land cover type. The specific parameter values ​​are shown in Table 1. Different model parameters are obtained according to the land surface type to calculate the land surface polarization reflectance value.

[0109] ;

[0110] In the formula, Represents the refractive index, which is generally 1.5. Based on the scattering angle , the incident Cosine of an angle and refraction angle The cosine value of , calculated as follows:

[0111] ;

[0112] ;

[0113] ;

[0114] Table 1 Empirical parameter assignment

[0115]

[0116] Step 3.23, calculation method of evaluation index;

[0117] The residual of the apparent polarization reflectivity at different observation angles is used as the evaluation index. The residual corresponding to a specific aerosol model is compared. When the residual is the smallest, the aerosol model and its corresponding aerosol fine particle optical thickness are the inversion results. The evaluation index is publicized as follows:

[0118] ;

[0119] In the formula, represents the observed polarization reflectivity, represents the simulated apparent polarization reflectivity. The step S4 comprises:

[0120] Since the active detection payload includes an atmospheric lidar payload, the present invention processes the atmospheric lidar observation data in step S2. Based on the atmospheric lidar payload observation data, the aerosol target layer definition range, and the aerosol type identification result information, the present invention determines the lidar ratio of the aerosol, uses the Fernald method to invert and obtain the aerosol extinction coefficient profile, and uses the aerosol optical thickness inverted from the passive detection payload observation data matching point as a constraint to optimize the inversion result of the aerosol extinction coefficient profile.

[0121] Step S4.1, determination of aerosol lidar ratio;

[0122] The determination of the aerosol lidar ratio in step S4 is specifically as follows: based on the aerosol type information, the aerosol lidar ratio S corresponding to the 532nm and 1064nm channels of the atmospheric lidar load is selected in combination with Table 2. a .

[0123] Table 2 Lidar ratios corresponding to different types of aerosols

[0124]

[0125] Step S4.2, aerosol inversion based on atmospheric lidar observation data;

[0126] When inverting the atmospheric aerosol extinction coefficient profile based on the lidar equation, the Fernald method considers the scattering of light by atmospheric molecules and aerosol particles respectively, and expresses the lidar equation as follows:

[0127] ;

[0128] Among them, since there are two unknown quantities in the equation: the backscattering coefficient of atmospheric aerosol particles at the detection distance Z and extinction coefficient , the backscattering coefficient of atmospheric molecules and extinction coefficient It can be obtained based on the US Standard Atmosphere Model. In the Fernald method, the aerosol lidar ratio at the detection distance Z is defined as the ratio of the aerosol extinction coefficient to the backscattering coefficient. , and it is assumed to be a constant that does not change with altitude. At the same time, the aerosol backscattering / extinction coefficient at a certain reference altitude, i.e., the boundary value, must be known in advance.

[0129] Therefore, by using the backscattered signals at different heights or detection distances detected by the lidar, based on the determined aerosol lidar ratio value, the atmospheric aerosol extinction characteristic parameters at the corresponding detection distance can be obtained by solving the lidar equation, such as Figure 5 Shown is the flow chart of inverting lidar observation data to obtain the aerosol extinction coefficient.

[0130] Based on the above, the inversion method of the aerosol extinction coefficient in step S4 is specifically: using the Fernald proximal solution to invert and obtain the inverted aerosol extinction coefficient profile of the atmospheric detection lidar data, including: selecting the aerosol backscattering coefficient corresponding to the location of the clean atmosphere near the top of the troposphere as the boundary value of the inversion; based on the aerosol lidar ratio and the boundary value, using the Fernald method to invert and obtain a preliminary aerosol extinction coefficient profile; using the aerosol optical thickness inverted from the matching points of the passive detection payload observation data as a constraint, correcting the atmospheric lidar radiation calibration signal, and optimizing the inversion result of the aerosol extinction coefficient profile.

[0131] Assuming that the air molecule extinction coefficient is proportional to the backscattering ratio Based on the Mie scattering lidar equation, after a series of transformations such as integration, natural logarithm and derivation, if a certain reference height is known in advance Under the premise of the backscattering coefficient of aerosol particles and air molecules, the reference height is The backscatter coefficient of aerosol particles at the following altitudes can be expressed as (forward integral):

[0132] ;

[0133] Similarly, the reference height can be obtained Aerosol extinction coefficient at the following altitudes (Forward integration):

[0134] ;

[0135] in, .

[0136] In summary, the aerosol parameter inversion method based on the collaborative detection data of active and passive detection payloads provided by the present invention can realize the acquisition of aerosol type, aerosol optical thickness, fine particle aerosol optical thickness and fine particle ratio with high precision, providing important data reference for the quantitative evaluation of the climate effect of aerosols and the study of atmospheric environmental changes.

[0137] On the other hand, the present invention provides an aerosol parameter inversion device based on active and passive collaborative detection data, wherein each unit included in the device can implement any step of the aforementioned method. Specifically, the device includes:

[0138] An observation data matching unit is used to spatially match the observation data of the active and passive detection payloads, extract pixel point information matching the observation data of the passive detection payload and the observation data of the active detection payload, and obtain spatially matched observation data of the passive detection payload; the active detection payload includes an atmospheric laser radar payload, and the passive detection payload includes any one of a multi-angle polarization detector, a multi-channel imaging spectrometer, a hyperspectral imager, and a multi-angle imaging spectrometer;

[0139] The target layer aerosol type identification unit is used to define the target layer range where the aerosol is located and identify the target layer aerosol type based on the active detection payload observation data;

[0140] The aerosol parameter inversion unit is used to determine the aerosol inversion channel of the passive detection payload observation data; when the passive detection payload is a multi-angle polarization detector, based on the matching pixel information, the simulated apparent reflectivity and apparent polarization reflectivity are obtained in the scalar lookup table and polarization lookup table used for aerosol inversion, and based on the target layer aerosol type information identified by the active detection payload observation data, the residuals of the apparent reflectivity and the apparent polarization reflectivity are used as evaluation indicators to invert and obtain the aerosol model, aerosol optical thickness, fine particle aerosol optical thickness and fine particle ratio; when the passive detection payload is any one of a multi-channel imaging spectrometer, a hyperspectral imager and a multi-angle imaging spectrometer, based on the matching pixel information, the simulated apparent reflectivity is obtained in the scalar lookup table used for aerosol inversion, and based on the target layer aerosol type information identified by the active detection payload observation data, the residuals of the apparent reflectivity are used as evaluation indicators to invert and obtain the aerosol model and aerosol optical thickness;

[0141] It is also used to identify the target layer aerosol type and evaluate the aerosol model based on the active detection payload observation data, obtain the lidar ratio corresponding to each target layer aerosol, and use the Fernald inversion method to obtain the aerosol extinction coefficient profile based on the aerosol optical thickness inverted from the passive detection payload observation data as a constraint.

[0142] In a third aspect, the present invention provides an electronic device, comprising: one or more processors; a memory for storing one or more programs; wherein, when the one or more programs are executed by the one or more processors, the one or more processors implement the aforementioned aerosol parameter inversion method based on active and passive collaborative detection data.

[0143] In a fourth aspect, the present invention provides a computer-readable storage medium having executable instructions stored thereon, which, when executed by a processor, enables the processor to implement the aforementioned aerosol parameter inversion method based on active and passive collaborative detection data.

[0144] The specific embodiments described above further illustrate the objectives, technical solutions and beneficial effects of the present invention in detail. It should be understood that the above description is only a specific embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principles of the present invention should be included in the scope of protection of the present invention.

Claims

1. An aerosol parameter inversion method based on active and passive collaborative detection data, characterized in that: The method comprises the following steps: Step S1: spatially matching the active and passive detection load observation data, extracting pixel point information matching the passive detection load observation data with the active detection load observation data, and obtaining spatially matched passive detection load observation data; The active detection payload includes an atmospheric laser radar payload, and the passive detection payload includes any one of a multi-angle polarization detector, a multi-channel imaging spectrometer, a hyperspectral imager, and a multi-angle imaging spectrometer; Step S2: Based on the active detection payload observation data, define the target layer range where the aerosol is located and identify the aerosol type in the target layer; Step S3: determining an aerosol inversion channel for passive detection payload observation data; When the passive detection payload is a multi-angle polarization detector, based on the matching pixel information, the simulated apparent reflectivity and apparent polarization reflectivity are obtained in the scalar lookup table and polarization lookup table for aerosol inversion. Based on the target layer aerosol type information identified by the active detection payload observation data, the residuals of the apparent reflectivity and apparent polarization reflectivity are used as evaluation indicators to invert and obtain the aerosol model, aerosol optical thickness, fine particle aerosol optical thickness and fine particle ratio. When the passive detection payload is any one of a multi-channel imaging spectrometer, a hyperspectral imager and a multi-angle imaging spectrometer, based on the matching pixel information, the simulated apparent reflectivity is obtained in the scalar lookup table for aerosol inversion. Based on the target layer aerosol type information identified by the active detection payload observation data, the residuals of the apparent reflectivity are used as evaluation indicators to invert and obtain the aerosol model and aerosol optical thickness. Step S4: Based on the target layer aerosol type identification results and aerosol model evaluation results of the active detection payload observation data, the lidar ratio corresponding to each target layer aerosol is obtained, and the aerosol extinction coefficient profile is obtained by inverting the aerosol optical thickness inverted from the passive detection payload observation data using the Fernald inversion method.

2. The aerosol parameter inversion method based on active and passive collaborative detection data according to claim 1 is characterized in that: The step S1 comprises: Step S1.1, based on the passive detection load observation data and the active detection load observation data, respectively extracting the longitude and latitude coordinates of the passive detection load observation data and the longitude and latitude coordinates of the foot points of the active detection load observation data; Step S1.2, based on the latitude and longitude coordinates of the foot point of the active detection payload observation data, matching the pixel point in the passive detection payload observation data of the collaborative detection that is closest to it; Step S1.3, extracting the passive detection load observation data corresponding to the matched pixel points described in step S1.

2.

3. The aerosol parameter inversion method based on active and passive collaborative detection data according to claim 1 is characterized in that: The method for defining the target layer where the aerosol in S2 is located includes: calculating the actual atmospheric attenuation backscattering coefficient, establishing a threshold value based on the atmospheric attenuation backscattering coefficient and atmospheric lidar noise information under a simulated clear sky mode, and when the attenuation scattering ratio coefficient of continuous points exceeding a preset height exceeds the threshold, the height range of the continuous points is selected as the target layer.

4. The aerosol parameter inversion method based on active and passive collaborative detection data according to claim 2 is characterized in that: Identifying the target layer aerosol type in step S2 includes: identifying the aerosol type based on the depolarization ratio, the atmospheric attenuation backscattering coefficient ratio and the underlying surface type obtained from the active detection payload observation data.

5. The aerosol parameter inversion method based on active and passive collaborative detection data according to claim 1 is characterized in that: In step S3, both the scalar lookup table and the polarization lookup table are constructed using the 6SV vector radiation transfer model, and the variables of the scalar lookup table include spectral wavelength, aerosol model, aerosol optical thickness, solar zenith angle, observation zenith angle, and relative azimuth angle; the variables of the polarization lookup table include spectral wavelength, aerosol model, fine particle aerosol optical thickness, solar zenith angle, observation zenith angle, and relative azimuth angle.

6. The aerosol parameter inversion method based on active and passive collaborative detection data according to claim 5 is characterized in that: In step S3, when the passive detection payload is a multi-angle polarization detector, obtaining the simulated apparent reflectivity and apparent polarization reflectivity in the scalar lookup table and polarization lookup table for aerosol inversion includes: According to the solar zenith angle, observation zenith angle and relative azimuth of the multi-angle polarization imager scalar observation data, the atmospheric scalar contribution corresponding to different aerosol models and aerosol optical thickness is searched in the scalar lookup table and the simulated apparent reflectivity is calculated. The scalar surface reflectivity is calculated to obtain the simulated on-board apparent scalar reflectivity, which is compared with the payload observation value. The aerosol model and aerosol optical thickness corresponding to the simulated value closest to the observed value are the inversion results obtained by the search. According to the solar zenith angle, observation zenith angle and relative azimuth angle of the polarization observation data of the multi-angle polarization imager, the atmospheric polarization contribution corresponding to different fine particle aerosol models and aerosol optical thickness is searched in the polarization lookup table and the simulated apparent reflectivity is calculated. The polarized surface reflectivity is calculated to obtain the simulated on-board apparent polarized reflectivity, which is compared with the payload observation value. The aerosol model and aerosol fine particle optical thickness corresponding to the simulated value closest to the observed value are the inversion results obtained by the search.

7. The aerosol parameter inversion method based on active and passive collaborative detection data according to claim 5 is characterized in that: In step S3, when the passive detection payload is any one of a multi-channel imaging spectrometer, a hyperspectral imager, and a multi-angle imaging spectrometer, obtaining the simulated apparent reflectance in a scalar lookup table for aerosol inversion based on the matched pixel information includes: According to the solar zenith angle, observation zenith angle and relative azimuth of the passive detection payload scalar observation data, the atmospheric scalar contribution corresponding to different aerosol models and aerosol optical thickness is searched in the scalar lookup table and the simulated apparent reflectivity is calculated. The scalar surface reflectivity is calculated to obtain the simulated on-board apparent scalar reflectivity, which is compared with the payload observation value. The aerosol model and aerosol optical thickness corresponding to the simulated value closest to the observed value are the inversion results obtained by the search.

8. The aerosol parameter inversion method based on active and passive collaborative detection data according to claim 5 is characterized in that: The fine particle ratio in step S3 is expressed as the ratio of the aerosol fine particle optical thickness obtained by inversion to the aerosol optical thickness.

9. The aerosol parameter inversion method based on active and passive collaborative detection data according to claim 1, characterized in that: In step S4, the aerosol lidar ratio of each target layer is obtained according to the identified aerosol type information of the target layer.

10. The aerosol parameter inversion method based on active and passive collaborative detection data according to claim 8, characterized in that: The inversion of the aerosol extinction coefficient profile by using the Fernald inversion method in step S4 includes: selecting the aerosol backscattering coefficient corresponding to the clean atmosphere position within a preset range of the troposphere top as the boundary value of the inversion; based on the aerosol lidar ratio and the inversion boundary value, using the Fernald method to perform inversion to obtain a preliminary aerosol extinction coefficient profile and an initial value of aerosol optical thickness; comparing the initial value of aerosol optical thickness obtained by inversion with the aerosol optical thickness obtained based on the pixel points matched by the passive detection payload, and optimizing the obtained preliminary aerosol extinction coefficient profile inversion result.

11. An aerosol parameter inversion device based on active and passive collaborative detection data, characterized in that: include: An observation data matching unit is used to perform spatial matching on the active and passive detection payload observation data, extract pixel point information matching the passive detection payload observation data with the active detection payload observation data, and obtain spatially matched passive detection payload observation data; The active detection payload includes an atmospheric laser radar payload, and the passive detection payload includes any one of a multi-angle polarization detector, a multi-channel imaging spectrometer, a hyperspectral imager, and a multi-angle imaging spectrometer; The target layer aerosol type identification unit is used to define the target layer range where the aerosol is located and identify the target layer aerosol type based on the active detection payload observation data; The aerosol parameter inversion unit is used to determine the aerosol inversion channel of the passive detection payload observation data; When the passive detection payload is a multi-angle polarization detector, based on the matching pixel information, the simulated apparent reflectivity and apparent polarization reflectivity are obtained in the scalar lookup table and polarization lookup table for aerosol inversion. Based on the target layer aerosol type information identified by the active detection payload observation data, the residuals of the apparent reflectivity and apparent polarization reflectivity are used as evaluation indicators to invert and obtain the aerosol model, aerosol optical thickness, fine particle aerosol optical thickness and fine particle ratio. When the passive detection payload is any one of a multi-channel imaging spectrometer, a hyperspectral imager and a multi-angle imaging spectrometer, based on the matching pixel information, the simulated apparent reflectivity is obtained in the scalar lookup table for aerosol inversion. Based on the target layer aerosol type information identified by the active detection payload observation data, the residuals of the apparent reflectivity are used as evaluation indicators to invert and obtain the aerosol model and aerosol optical thickness. It is also used to identify the target layer aerosol type and evaluate the aerosol model based on the active detection payload observation data, obtain the lidar ratio corresponding to each target layer aerosol, and use the Fernald inversion method to obtain the aerosol extinction coefficient profile based on the aerosol optical thickness inverted from the passive detection payload observation data as a constraint.

12. An electronic device, characterized in that: include: one or more processors; A memory for storing one or more programs; Wherein, when one or more programs are executed by the one or more processors, the one or more processors implement the aerosol parameter inversion method based on active and passive collaborative detection data as described in any one of claims 1-10.

13. A computer-readable storage medium, characterized in that: Executable instructions are stored thereon, and when the instructions are executed by the processor, the processor can implement the aerosol parameter inversion method based on active and passive collaborative detection data as described in any one of claims 1 to 10.

Images