Calculation method and device for earth-gas light scattering characteristics

Abstract

The invention provides a method and a device for calculating earth-gas light scattering characteristics. The method comprises the following steps: dispersing detection light rays in a detection view field range of a detector to obtain a plurality of detection light rays; determining a detection scene corresponding to each detection light ray according to the position of the detector and the direction of each detection light ray, and calculating a geometrical characteristic parameter value under the detection scene; according to the detection scene of each piece of detection light, extracting an environment characteristic parameter value of a scattering surface geographic position corresponding to each piece of detection light in the detection scene; and respectively inputting each group of environment characteristic parameter values and geometric characteristic parameter values into the limb light scattering characteristic prediction model or the non-limb light scattering characteristic prediction model to obtain the geogas light scattering radiation brightness in the detection view field range. According to the scheme, the calculation efficiency is further improved while the actual requirement for high precision of space-time dynamic evolution of the geogas system is met.

Description

Technical Field

[0001] The embodiments of the present invention relate to the technical field of ground-air background light scattering radiation, and in particular to a method and device for calculating ground-air light scattering characteristics. Background Art

[0002] The scattering of ground-air light is a key factor affecting the optical imaging of the detector by the Earth's atmosphere. The calculation of scattering characteristics needs to consider the spatial position relationship between the sun, the Earth and the detector, as well as the optical characteristics of the ground-air background system. At the same time, the dynamics of the atmosphere, changes in the position and thickness of clouds, and changes in aerosol concentration and distribution all increase the complexity of ground-air light calculation.

[0003] However, existing radiation transfer algorithms mostly average the optical properties of surface features, clouds, aerosols and the atmosphere, and have poor adaptability to scenes with large differences in the ground and atmosphere, and cannot meet the actual needs of high-precision under dynamic evolution conditions of time and space. At the same time, traditional radiation transfer algorithms use complex differential equations to characterize the multiple scattering and absorption processes of photons, and the calculation process is relatively time-consuming.

[0004] Therefore, there is an urgent need to provide a method and device for calculating the light scattering characteristics of the ground atmosphere. Summary of the invention

[0005] In order to solve the problem that the traditional radiation transmission algorithm cannot meet the actual requirements of high precision of spatiotemporal dynamic evolution conditions and has low calculation efficiency, the embodiment of the present invention provides a method and device for calculating the ground-air light scattering characteristics.

[0006] In a first aspect, an embodiment of the present invention provides a method for calculating ground-air light scattering characteristics, the method comprising:

[0007] Discrete the detection light within the detection field of view of the detector to obtain a plurality of detection lights;

[0008] According to the position of the detector and the direction of each detection light, the detection scene corresponding to each detection light is determined, and the geometric characteristic parameter value and the geographical location of the scattering surface under the detection scene are calculated; wherein the detection scene includes an edge detection scene and a non-edge detection scene, and the geometric characteristic parameters corresponding to different detection scenes are different;

[0009] Extracting the environmental characteristic parameter values ​​of each detection light at the geographical location of the scattering surface in the corresponding detection scene;

[0010] Input each set of environmental characteristic parameter values ​​and geometric characteristic parameter values ​​into the light scattering characteristic prediction model of the corresponding detection scene respectively, and obtain the ground-air light scattering radiation brightness within the detection field of view under the detection scene;

[0011] Among them, the light scattering characteristic prediction model includes an edge light scattering characteristic prediction model and a non-edge light scattering characteristic prediction model; the edge light scattering characteristic prediction model and the non-edge light scattering characteristic prediction model are both trained using known geometric characteristic parameter values, known environmental characteristic parameter values ​​and known ground light scattering radiation brightness values ​​as training sample sets.

[0012] In a second aspect, an embodiment of the present invention further provides a device for calculating the light scattering characteristics of the ground atmosphere, the device comprising:

[0013] A discrete unit, used for discretizing the detection light within the detection field of the detector to obtain a plurality of detection lights;

[0014] A determination unit, used to determine the detection scene corresponding to each detection light according to the position of the detector and the direction of each detection light, and calculate the geometric characteristic parameter value and the geographical location of the scattering surface under the detection scene; wherein the detection scene includes an edge detection scene and a non-edge detection scene, and the geometric characteristic parameters corresponding to different detection scenes are different;

[0015] An extraction unit, used to respectively extract the environmental characteristic parameter value of each detection light at the geographical location of the scattering surface in the corresponding detection scene;

[0016] A prediction unit is used to input each set of environmental characteristic parameter values ​​and geometric characteristic parameter values ​​into the light scattering characteristic prediction model of the corresponding detection scene, so as to obtain the ground-air light scattering radiation brightness within the detection field of view under the detection scene; wherein the light scattering characteristic prediction model includes an edge light scattering characteristic prediction model and a non-edge light scattering characteristic prediction model; both the edge light scattering characteristic prediction model and the non-edge light scattering characteristic prediction model are obtained by training with known geometric characteristic parameter values, known environmental characteristic parameter values ​​and known ground-air light scattering radiation brightness values ​​as training sample sets.

[0017] In a third aspect, an embodiment of the present invention further provides a computing device, including a memory and a processor, wherein the memory stores a computer program, and when the processor executes the computer program, the method described in any embodiment of this specification is implemented.

[0018] In a fourth aspect, an embodiment of the present invention further provides a computer-readable storage medium having a computer program stored thereon, which, when executed in a computer, enables the computer to execute the method described in any embodiment of this specification.

[0019] On the other hand, an embodiment of the present application further provides a computer program product, which includes a computer program. A processor of a computer device reads the computer program from a computer-readable storage medium, and the processor executes the computer program, so that the computer device executes any of the methods described in the above embodiments.

[0020] The embodiment of the present invention provides a method and device for calculating the ground-air light scattering characteristics. First, an equivalent ground-air light bidirectional reflection prediction model based on machine learning is established, and known environmental characteristic parameters, known geometric characteristic parameters and known ground-air light scattering radiation brightness values ​​in edge detection scenarios and non-edge detection scenarios are selected as training sample sets to train the two prediction models respectively, thereby obtaining edge and non-edge equivalent ground-air light bidirectional reflection (EBRDF) prediction models; then, for the real scene of the detector, the detector field of view is evenly discretized into a large number of detection lights, and by performing reverse ray tracing on each detection light, the geometric characteristic parameter calculation in any detection scene is realized, and the environmental parameters of the geographical location of the scattering surface are further extracted, and the ground-air light scattering characteristics are calculated through the EBRDF prediction model; in this way, while meeting the actual high-precision actual needs of the spatiotemporal dynamic evolution of the ground-air system, the calculation efficiency is further improved. BRIEF DESCRIPTION OF THE DRAWINGS

[0021] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0022] Figure 1 is a flow chart of a method for calculating ground-air light scattering characteristics provided by one embodiment of the present invention;

[0023] Figure 2 is a schematic diagram of a neural network structure of a light scattering characteristic prediction model provided by an embodiment of the present invention;

[0024] Figure 3 It is a flow chart of speed test of a method for calculating ground-air light scattering characteristics provided by one embodiment of the present invention;

[0025] Figure 4 is a hardware architecture diagram of a computing device provided by an embodiment of the present invention;

[0026] Figure 5 It is a structural diagram of a device for calculating the light scattering characteristics of the ground atmosphere provided by one embodiment of the present invention. DETAILED DESCRIPTION

[0027] In order to make the purpose, technical solutions and advantages of the embodiments of the present invention clearer, the technical solutions in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are part of the embodiments of the present invention, rather than all the embodiments. Based on the embodiments in the present invention, all other embodiments obtained by ordinary technicians in this field without making creative work are within the scope of protection of the present invention.

[0028] As mentioned above, in the spatiotemporal dynamic ground-atmosphere background system, the ground-atmosphere light scattering characteristics are affected by the surface, atmosphere, aerosol, and cloud environment background. At the same time, there are dynamic processes such as the earth's rotation, the movement of the sun, the movement of the detector, and the spatiotemporal changes of clouds and the atmosphere. However, the existing radiation transfer algorithm only considers static scenes and cannot meet the requirements of spatiotemporal dynamic evolution conditions. It cannot take into account both the accuracy and efficiency of the calculation when processing dynamic scenes.

[0029] To address this problem, the inventors proposed that a rapid prediction model for light scattering characteristics can be established through machine learning methods, and further combined with a reverse ray tracing algorithm to calculate the spatial geometry of the detection scene. Based on this prediction model, it is possible to achieve rapid calculation of light scattering characteristics while meeting the actual high-precision needs of the spatiotemporal dynamic evolution of the earth-atmosphere system.

[0030] The specific implementation of the above concept is described below.

[0031] Please refer to Figure 1 The embodiment of the present invention provides a method for calculating the light scattering characteristics of the ground atmosphere, the method comprising:

[0032] Step 100, discretizing the detection light within the detection field of view of the detector to obtain a plurality of detection lights;

[0033] Step 102, determining the detection scene corresponding to each detection light according to the position of the detector and the direction of each detection light, and calculating the geometric characteristic parameter value and the geographical location of the scattering surface under the detection scene; wherein the detection scene includes an edge detection scene and a non-edge detection scene, and different detection scenes have different corresponding geometric characteristic parameters;

[0034] Step 104, respectively extracting the environmental characteristic parameter value of each detection light at the geographical location of the scattering surface in the corresponding detection scene;

[0035] Step 106, inputting each set of environmental characteristic parameter values ​​and geometric characteristic parameter values ​​into the light scattering characteristic prediction model of the corresponding detection scene, and obtaining the ground-air light scattering radiation brightness within the detection field of view under the detection scene;

[0036] Among them, the light scattering characteristic prediction model includes an edge light scattering characteristic prediction model and a non-edge light scattering characteristic prediction model; the edge light scattering characteristic prediction model and the non-edge light scattering characteristic prediction model are both trained using known geometric characteristic parameter values, known environmental characteristic parameter values ​​and known ground light scattering radiation brightness values ​​as training sample sets.

[0037] In an embodiment of the present invention, first, an equivalent air-ground bidirectional reflectance prediction model based on machine learning is established, and known environmental characteristic parameters, known geometric characteristic parameters and known air-ground light scattering radiation brightness values ​​in edge detection scenarios and non-edge detection scenarios are selected as training sample sets to train the two prediction models respectively, thereby obtaining edge and non-edge equivalent air-ground bidirectional reflectance (EBRDF) prediction models; then, for the real scene of the detector, the detector field of view is evenly discretized into a large number of detection lights, and by performing reverse ray tracing on each detection light, the calculation of geometric characteristic parameters in any detection scene is realized, and the environmental parameters of the geographical location of the scattering surface are further extracted, and the air-ground light scattering characteristics are calculated through the EBRDF prediction model; in this way, while meeting the actual needs of high-precision actual spatiotemporal dynamic evolution of the air-ground system, the computing efficiency is further improved.

[0038] In some embodiments, the environmental characteristic parameters include surface environmental parameters, aerosol environmental parameters and cloud environmental parameters; wherein the surface environmental parameters include surface reflectivity, the aerosol environmental parameters include aerosol type and aerosol optical thickness, and the cloud environmental parameters include cloud top height and cloud optical thickness.

[0039] The surface albedo characterizes the ratio of the surface reflected energy to the total incident energy. Different types of surfaces have different albedo, which affects the intensity and spatial distribution of scattered light. Therefore, in the embodiment of the present invention, according to the statistical distribution information of the surface type, the surface types such as forests, snow, lakes, seas, and deserts are extracted, and the surface albedo is extracted as a characteristic physical quantity to distinguish different surfaces. In addition, different types of aerosols have different optical properties and different scattering and absorption capabilities for light. Aerosol optical thickness reflects the extinction effect of the entire layer of aerosols, and highly summarizes the effects of aerosol concentration, particle size distribution, etc. on light scattering. Therefore, in the embodiment of the present invention, according to the statistical distribution information of aerosols, aerosol types such as villages, towns, and oceans are extracted, and aerosol types and aerosol optical thickness are extracted as characteristic physical quantities to distinguish different aerosols. At the same time, the cloud top height determines the path length between the incident light passing through the atmosphere and reaching the scattering surface. The longer the optical path, the more obvious the scattering effect. Cloud optical thickness describes the absorption and scattering ability of the cloud layer to light, and highly summarizes the effects of factors such as cloud particle size and structure on light scattering. Therefore, in the embodiment of the present invention, cloud types such as stratus clouds, cumulus clouds, cirrus clouds, and nimbostratus clouds are extracted based on cloud statistical distribution information, and cloud top height and cloud optical thickness are extracted as characteristic physical quantities to distinguish different cloud clusters.

[0040] For steps 100 to 102:

[0041] In some embodiments, the geometric characteristic parameters in the non-edge detection scenario include the solar zenith angle, the relative azimuth angle and the observation zenith angle, and the geometric characteristic parameters in the edge detection scenario include the solar zenith angle, the relative azimuth angle and the cut height.

[0042] Considering that the directions of scattered light generated in edge detection scenarios and non-edge detection scenarios are different, the corresponding geometric characteristic parameters are different. Specifically, in non-edge detection scenarios, there is an intersection between the detection light and the earth's surface, and the scattered light is distributed in the hemispherical space, while in edge detection scenarios, there is no intersection between the detection light and the earth's surface, and the scattered light is distributed in a plane at a certain height from the earth's surface. Therefore, in the embodiment of the present invention, the detection scene corresponding to each detection light is determined based on the intersection of the detection light of the detector and the earth's surface during the transmission calculation of the ground atmosphere background light radiation. The intersection is a non-edge detection scene, and the non-intersection is an edge detection scene.

[0043] In some embodiments, the geographical location of the scattering surface includes the location of the scattering surface at the intersection of the detection light and the earth and the location of the scattering surface at the tangent point of the detection light and the earth;

[0044] Step 102 includes:

[0045] According to the position of the detector and the direction of each detection light, the intersection discriminant of each detection light and the earth is constructed respectively;

[0046] Based on each of the intersection discriminants, respectively determining the detection scene corresponding to each detection light;

[0047] If the intersection discriminant is greater than zero, then the detection light has an intersection with the earth, and the detection scene corresponding to the current detection light is determined as a non-edge detection scene, and the intersection position and geometric characteristic parameter values ​​under the detection scene are calculated respectively;

[0048] If the intersection discriminant is less than zero, the detection light has no intersection with the earth, and the detection scene corresponding to the current detection light is determined as an edge detection scene, and the tangent point position and geometric characteristic parameter values ​​under the detection scene are calculated.

[0049] In the embodiment of the present invention, in the earth-fixed coordinate system, the origin of the coordinate system is taken as the center of the earth, and according to the relative positions of the sun, the earth and the detector, a discriminant of the intersection of each detection light ray and the earth is constructed:

[0050] Δ=(p·d) 2 -(p·pR E 2 )

[0051] Where p is the coordinate position of the detector, d is the direction of the detection light, R E is the radius of the Earth.

[0052] When the intersection discriminant is greater than 0, the detection ray intersects with the earth. The current detection ray is a non-edge detection scene. According to the actual detection process, the near intersection point where the detection ray intersects with the earth surface is determined as the actual intersection point r int Position vector:

[0053]

[0054] r int =p+t int d

[0055] Among them, t int is the distance between the detector and the intersection point;

[0056] Thus, the longitude and latitude of the intersection position can be obtained:

[0057] φ int =arctan 2 (y int ,x int )

[0058] θ int =arccos(z int / R E )

[0059] In the formula, φint is the longitude of the intersection point, θ int is the latitude of the intersection point, x int ,y int and z int is the spatial position coordinate of the intersection point.

[0060] Furthermore, the geometric characteristic parameter values ​​(solar zenith angle, relative azimuth angle and observation zenith angle) in the non-edge detection scenario can be obtained by the following formula:

[0061] θ i =arccos(s·n)

[0062]

[0063] s n =s-(s·n)n

[0064] e n =e-(e·n)n

[0065] θ r =arccos(e·n)

[0066] In the formula, θ i The solar zenith angle, s is the unit vector from the intersection position to the sun position, n is the unit vector from the center of the earth to the intersection position, is the relative azimuth, e is the unit vector pointing from the intersection point to the detector position, s n is the projection vector of vector s on the plane at the observation point, e n is the projection vector of vector e on the plane at the observation point, θ r is the observation zenith angle, and e is the unit vector pointing from the intersection position to the detector position.

[0067] When the discriminant is less than 0, the detection light has no intersection with the earth's surface, and the current detection light is a non-edge detection scene. According to the actual detection process, the tangent point r where the detection light is tangent to the earth's surface is determined. int Position vector:

[0068] t int =-p·d

[0069] r int =p+t int d

[0070] Among them, t int is the distance between the detector and the tangent point;

[0071] The scattered light is distributed in a plane at a certain height from the earth's surface. By calculating the tangential height (tangent height) between the current detection light and the earth, the geometric characteristic parameter values ​​in the edge detection scene and the scattering surface position at the tangent point are obtained:

[0072] According to the positional relationship between the detection light and the earth's surface, the tangential height (tangential height) of the detection light can be obtained:

[0073]

[0074] Wherein, H is the cutting height, and OP is the vector pointing from the detector position p to the center of the earth.

[0075] The calculation method of the solar zenith angle and relative azimuth is the same as that of the non-limb detection scene, except that there is no intersection between the limb detection scene and the spherical surface.

[0076] θ i =arccos(s·n)

[0077]

[0078] Therefore, s is the unit vector from the tangent point to the sun, and n is the unit vector from the center of the earth to the tangent point. is the relative azimuth, e is the unit vector from the tangent point to the detector position, s n is the projection vector of vector s on the plane at the observation point, e n is the projection vector of vector e on the plane at the observation point.

[0079] Regarding step 104:

[0080] In an embodiment of the present invention, for a real scene of the detector, the field of view of the detector is firstly discretized uniformly into a large number of detection rays, and then the detection rays are traced backward to determine whether there is an intersection with the sphere. When there is an intersection, it is a non-edge detection scene. The position of the intersection and the geometric characteristic parameter values ​​(observation zenith angle, solar zenith angle and relative azimuth) are solved, and the environmental characteristic parameter values ​​at the intersection position are further extracted. When there is no intersection, it is an edge detection scene. The tangential height of the detection rays, the solar zenith angle and the relative azimuth are calculated, and the environmental characteristic parameters at the tangent position at the tangential height are further extracted. Finally, they are respectively input into the constructed equivalent ground-air bidirectional reflectance (EBRDF) prediction model to obtain the light scattering characteristic values ​​(ground-air light scattering radiation brightness) of all detection rays within the detector detection field of view.

[0081] Regarding step 106:

[0082] In some preferred embodiments, the training sample sets are obtained by simulating three-dimensional radiation transmission based on data sets consisting of environmental characteristic parameter values ​​and geometric characteristic parameter values ​​in edge detection scenarios and non-edge detection scenarios.

[0083] Considering that the distribution range of scattered light is different in different detection scenarios, therefore, in an embodiment of the present invention, firstly, the distribution range of geometric characteristic parameter values ​​and environmental characteristic parameter values ​​in two different detection scenarios (edge ​​detection scenario and non-edge detection scenario) is observed through satellite observation data, and characteristic values ​​are extracted at intervals within the distribution range to achieve full coverage of the detection scene conditions of the complex ground-atmosphere system; then, according to the characteristic parameter values ​​(including geometric characteristic parameter values ​​and environmental characteristic parameter values) extracted respectively in the two detection scenarios, all array sets of characteristic parameter values ​​are traversed respectively, and three-dimensional radiation transmission simulation (such as Monte Carlo transmission simulation method) is performed on each obtained array set, and finally, a surface atmospheric cloud database containing geometric characteristic parameter values, environmental characteristic parameter values ​​and light scattering radiation brightness values ​​in different detection scenarios can be obtained respectively, and the database is cleaned of erroneous, repeated or incomplete data, and classified and hierarchically managed according to importance and characteristic data to improve accuracy and reliability. Sample sampling is performed from the database to provide a statistical sample set for training and verification of the prediction model.

[0084] In some embodiments, the edge light scattering property prediction model and the non-edge light scattering property prediction model are both trained in the following manner:

[0085] Build a preset neural network learning model;

[0086] Setting the hyperparameters and loss function of the neural network learning model; wherein the loss function is a mean relative error loss function;

[0087] Using known environmental characteristic parameter values ​​and known geometric characteristic parameter values ​​in edge or non-edge detection scenarios as input training sample sets for the neural network learning model, using known ground-air light scattering radiation brightness values ​​in edge or non-edge detection scenarios as output training sample sets, and using the stochastic gradient descent method to train the neural network learning model; wherein the geometric characteristic parameters in the edge detection scenario include the solar zenith angle, relative azimuth angle, and observation zenith angle, and the geometric characteristic parameters in the non-edge detection scenario include the solar zenith angle, relative azimuth angle, and cut height;

[0088] During the training process, the loss and accuracy of the neural network model are calculated, and the network parameters of the neural network learning model are iterated until the model converges or reaches a preset number of iterations to obtain a prediction model for edge light scattering characteristics or a prediction model for non-edge light scattering characteristics.

[0089] like Figure 2 As shown, the prediction model in the embodiment of the present invention includes an input layer, a hidden layer and an output layer. The input layer is set with 8 sample features, which characterize the environmental characteristic parameters and detection geometric characteristic parameters of the surface, atmosphere and cloud characteristics. The ReLU activation function is adopted, and a fully connected network relationship is established between the input layer and the hidden layer. The output is the light scattering radiance result under given detection conditions. Since it is a regression problem with output continuous values, the number of neurons in the output layer is set to 1.

[0090] At the same time, since the output results vary greatly under different inputs, that is, the target values ​​of the training data set have large orders of magnitude differences, the commonly used root mean square error cannot be used as the loss function. Instead, the average relative error is defined as the loss function, and the expression is as follows:

[0091]

[0092] In the formula, η is the relative error, y is the target value of the sample data set, and y i is the predicted value after the i-th iteration.

[0093] Network training is the process of continuously adjusting parameters to reduce the loss function. The gradient descent algorithm is used to converge the prediction accuracy by adjusting the learning rate and iterative step size, and a prediction model for the non-edge and edge light scattering characteristics under the ground-atmosphere background is obtained.

[0094] In summary, in the embodiments of the present invention, a light scattering characteristic tracking algorithm based on a neural network learning model is adopted, and the computational efficiency is much higher than that of a traditional three-dimensional radiation transmission algorithm. In view of the real scene of the detector, the detector field of view is evenly discretized into a large number of detection lights, and multi-threaded parallel technology is used to improve the calculation speed. At the same time, the EBRDF prediction model and the meteorological support database are preloaded in the algorithm, and the prediction model is called for all detection lights as a whole, which significantly improves the efficiency of the algorithm.

[0095] Finally, the inventors verified the accuracy and speed of the method for calculating the ground-air light scattering characteristics in this embodiment.

[0096] In the non-edge detection scenario, the measured data of Modis satellite is used as the benchmark to compare the relative deviation between the characterization model results of eight characteristic parameters and the measured results. The observation scenarios are selected in three mountainous areas of Hefei, Zhuozhou and Lijiang, which have different surface and climate characteristics. Selecting multiple detection locations to test the accuracy of the model can avoid the randomness of the test conditions. Therefore, the selected observation scenarios are representative and can test the coverage of the calculation model on the functional indicators.

[0097] (1) Hefei

[0098] Hefei is a mixed urban area in the central region, mainly including water and forest cities. Its geographical coordinates range from 31°45′ to 31°55′ north latitude and 117°15′ to 117°35′ east longitude. The observed values ​​of characteristic physical quantities are shown in Table 1.

[0099] Table 1 Statistics of characteristic physical quantities of Hefei detection scene

[0100]

[0101]

[0102] Using the calculation method proposed in the embodiment of the present invention, the light scattering radiation brightness under this condition is 70.50W / (m 2 *sr), the satellite observation value is 77.55W / (m 2 *sr), the relative deviation is about 9.09%.

[0103] (2) Zhuozhou

[0104] Zhuozhou area is a mixed urban landform in the northern region, mainly including water forest cities. The geographical coordinate range is approximately between 115°44′-116°15′ east longitude and 39°21′-39°36′ north latitude. The observed values ​​of characteristic physical quantities are shown in Table 2.

[0105] Table 2 Statistics of characteristic physical quantities of Zhuozhou detection scene

[0106]

[0107] Using the calculation method proposed in the embodiment of the present invention, the radiance result under this condition is 52.79W / (m 2 *sr), the satellite observation value is 48.24W / (m 2 *sr), the relative deviation is about 8.60%.

[0108] (3) Lijiang

[0109] The Lijiang mountainous area is a mixed urban feature in the southwest region, mainly including water forest cities. The geographical coordinate range is approximately between 26°05′ and 27°50′ north latitude and 99°40′ and 101°20′ east longitude. The observed values ​​of characteristic physical quantities are shown in Table 3.

[0110] Table 3 Statistics of characteristic physical quantities of Lijiang detection scene

[0111]

[0112]

[0113] Using the calculation method proposed in the embodiment of the present invention, the radiance result under this condition is 43.99W / (m 2 *sr), the satellite observation value is 41.90W / (m 2 *sr), the relative deviation is about 4.99%.

[0114] In summary, in the non-edge detection scenario, different ground backgrounds, cloud and aerosol models and detection geometries are selected, the calculation method of the ground-atmosphere light scattering characteristics proposed in the present invention is adopted, and the Modis measured value is used as the benchmark. The relative errors in the three groups of detection scenarios are 9.09%, 8.60% and 4.99% respectively, and the root mean square error is about 7.78%.

[0115] (4) Speed

[0116] Test environment: The hardware environment is CPU: Intel core i9-14900KF, memory 128GB; the software environment is an executable file compiled and linked by Fortran language.

[0117] Figure 3 This is a schematic diagram of the speed test process. Under given hardware and software conditions, the sequence calculation conditions are set for different calculation conditions. The running time of the test algorithm is from reading the detector input to completing all the ground-air light scattering calculations. Since a large amount of meteorological support data (clouds, aerosols, and surface) needs to be read from the hard disk to the memory when calculating the ground-air light radiance sequence, it only needs to be read from the hard disk once. When evaluating the speed of the ground-air light scattering characteristic algorithm, the preloading time is not included.

[0118] For the calculation of the ground-air radiance in a single detection direction, a sequence calculation of eight sets of detection conditions is set, with an average running time of about 0.75ms, which can be regarded as the inherent time of a single detection direction. Combined with speed-up technologies such as multi-threaded parallelism, 16384 (128×128) discrete detection directions are used, and the average running time to complete the calculation of all detection directions is about 7ms.

[0119] In the limb detection scenario, the SCIATRAN atmospheric radiation transfer model, a foreign commercial computing model, is used as a benchmark. Since the cut height is the main influencing factor of the light scattering characteristics, the variation trend of the scattered radiance with the cut height under different characteristic physical quantity values ​​is compared.

[0120] The characteristic physical quantities of the selected edge comparison group 1 are: surface albedo 0.2, rural aerosol, aerosol optical thickness 0.05, cloud top height 2.5km, cloud optical thickness 4, solar zenith angle 48.68°, relative azimuth 50.54°. The comparison between the parameterized characterization model proposed in the present invention and the SCIATRAN results is shown in Table 4.

[0121] Table 4 Statistical table of marginal comparison group 1 results

[0122]

[0123] It can be seen from Table 4 that, using the calculation method proposed in the present invention, the root mean square error under this set of conditions is approximately 4.59%.

[0124] The characteristic physical quantities of the selected edge comparison group 2 are: surface albedo 0.3, rural aerosol, aerosol optical thickness 0.1, cloud top height 3km, cloud optical thickness 3, solar zenith angle 48.68°, relative azimuth 50.54°. The comparison between the parameterized characterization model proposed in the present invention and the SCIATRAN results is shown in Table 5.

[0125] Table 5 Statistical table of marginal comparison group 2 results

[0126]

[0127] It can be seen from Table 5 that, using the calculation method proposed in the present invention, the root mean square error under this set of conditions is about 5.98%.

[0128] In summary, in the edge detection scenario, different ground backgrounds, cloud and aerosol models and detection geometries are selected, the parameterized characterization method proposed in the present invention is adopted, and the commercial computing model SCIATRAN is used as a benchmark. The root mean square errors in the two groups of detection scenarios are approximately 4.59% and 5.98%, respectively.

[0129] Using the same hardware and software configuration as the non-edge velocity test, for the calculation of the ground-air radiance in a single detection direction, a sequence calculation of eight sets of detection conditions was set, with an average running time of about 0.80ms, which can be regarded as the inherent time of a single detection direction. Combined with speed-up technologies such as multi-threaded parallelism, 16384 (128×128) discrete detection directions were used, and the average running time to complete the calculation of all detection directions was about 8.5ms.

[0130] like Figure 4 , Figure 5 As shown, an embodiment of the present invention provides a device for calculating the light scattering characteristics of the ground atmosphere. The device embodiment can be implemented by software, or by hardware or a combination of software and hardware. From the hardware level, Figure 4 As shown, it is a hardware architecture diagram of a computing device where a computing device for calculating the ground-air light scattering characteristics provided by an embodiment of the present invention is located, except Figure 4 In addition to the processor, memory, network interface, and non-volatile memory shown in the figure, the computing device in which the device is located in the embodiment may also generally include other hardware, such as a forwarding chip responsible for processing messages, etc. Taking software implementation as an example, Figure 5As shown, as a device in a logical sense, the CPU of the computing device in which it is located reads the corresponding computer program in the non-volatile memory into the memory and runs it. This embodiment provides a computing device for the light scattering characteristics of the ground atmosphere, and the device includes:

[0131] The discrete unit 601 is used to discretize the detection light within the detection field of the detector to obtain a plurality of detection lights;

[0132] A determination unit 602 is used to determine the detection scene corresponding to each detection light according to the position of the detector and the direction of each detection light, and calculate the geometric characteristic parameter value and the geographical location of the scattering surface in the detection scene; wherein the detection scene includes an edge detection scene and a non-edge detection scene, and different detection scenes correspond to different geometric characteristic parameters;

[0133] An extraction unit 603 is used to extract the environmental characteristic parameter value of each detection light at the geographical location of the scattering surface in the corresponding detection scene;

[0134] The prediction unit 604 is used to input each set of environmental characteristic parameter values ​​and geometric characteristic parameter values ​​into the light scattering characteristic prediction model of the corresponding detection scene, and obtain the ground-air light scattering radiation brightness within the detection field of view under the detection scene;

[0135] Among them, the light scattering characteristic prediction model includes an edge light scattering characteristic prediction model and a non-edge light scattering characteristic prediction model; the edge light scattering characteristic prediction model and the non-edge light scattering characteristic prediction model are both trained using known geometric characteristic parameter values, known environmental characteristic parameter values ​​and known ground light scattering radiation brightness values ​​as training sample sets.

[0136] In an embodiment of the present invention, the discrete unit 601 can be used to execute step 100 in the above method embodiment, the determination unit 602 can be used to execute step 102 in the above method embodiment, the extraction unit 603 can be used to execute step 104 in the above method embodiment, and the prediction unit 604 can be used to execute step 106 in the above method embodiment.

[0137] In one embodiment of the present invention, the environmental characteristic parameters include surface environmental parameters, aerosol environmental parameters and cloud environmental parameters; wherein the surface environmental parameters include surface reflectivity, the aerosol environmental parameters include aerosol type and aerosol optical thickness, and the cloud environmental parameters include cloud top height and cloud optical thickness.

[0138] In one embodiment of the present invention, the geographical location of the scattering surface includes the scattering surface location at the intersection of the detection light and the earth and the scattering surface location at the tangent point of the detection light and the earth;

[0139] The determining unit 602 determines the detection scene corresponding to each detection light and calculates the geometric feature parameter value of the detection scene according to the position of the detector and the direction of each detection light, including:

[0140] According to the position of the detector and the direction of each detection light, the intersection discriminant of each detection light and the earth is constructed respectively;

[0141] Based on each of the intersection discriminants, respectively determining the detection scene corresponding to each detection light;

[0142] If the intersection discriminant is greater than zero, then the detection light has an intersection with the earth, and the detection scene corresponding to the current detection light is determined as a non-edge detection scene, and the intersection position and geometric characteristic parameter values ​​under the detection scene are calculated respectively;

[0143] If the intersection discriminant is less than zero, the detection light has no intersection with the earth, and the detection scene corresponding to the current detection light is determined as an edge detection scene, and the tangent point position and geometric characteristic parameter values ​​under the detection scene are calculated.

[0144] In one embodiment of the present invention, the intersection discriminant of the detection light and the earth is as follows:

[0145] Δ=(p·d) 2 -(p·pR E 2 )

[0146] Where p is the coordinate position of the detector, d is the direction of the detection light, R E is the radius of the Earth.

[0147] In one embodiment of the present invention, the vector of the intersection point position and the vector of the tangent point position are both calculated by the following formula:

[0148] r int =p+t int d

[0149] In the formula, r int is the vector of the intersection / tangent point, t int is the distance between the detector and the intersection / tangent point. In non-edge detection scenarios, In the edge detection scenario, t int =-p·d.

[0150] In one embodiment of the present invention, the geometric characteristic parameters in the non-edge detection scenario include the solar zenith angle, the relative azimuth angle and the observation zenith angle, and the geometric characteristic parameters in the edge detection scenario include the solar zenith angle, the relative azimuth angle and the cut height;

[0151] Among them, the solar zenith angle and relative azimuth angle in the non-limb detection scenario and the limb detection scenario are calculated by the following formula:

[0152] θ i =arccos(s·n)

[0153]

[0154] In the formula, θ i The solar zenith angle, s is the unit vector pointing from the intersection / tangent point to the sun, n is the unit vector pointing from the center of the earth to the intersection / tangent point, is the relative azimuth, s n For n is the projection vector of vector s on the plane at the observation point, e n is the projection vector of vector e on the plane at the observation point;

[0155] The cutting height is calculated by the following formula:

[0156]

[0157] Wherein, H is the cutting height, OP is the vector pointing from the detector position p to the center of the earth;

[0158] The observation zenith angle is calculated by the following formula:

[0159] θ r =arccos(e·n)

[0160] In the formula, θ r is the observation zenith angle, and e is the unit vector pointing from the intersection position to the detector position.

[0161] In one embodiment of the present invention, in the prediction unit 604, the training sample sets are respectively obtained by simulating three-dimensional radiation transmission based on data sets consisting of environmental characteristic parameter values ​​and geometric characteristic parameter values ​​in edge detection scenarios and non-edge detection scenarios.

[0162] It is to be understood that the structure illustrated in the embodiment of the present invention does not constitute a specific limitation on a device for calculating the light scattering characteristics of the ground atmosphere. In other embodiments of the present invention, a device for calculating the light scattering characteristics of the ground atmosphere may include more or fewer components than those illustrated, or combine certain components, or separate certain components, or arrange the components differently. The illustrated components may be implemented in hardware, software, or a combination of software and hardware.

[0163] The information interaction, execution process and other contents between the modules in the above-mentioned device are based on the same concept as the embodiment of the method of the present invention. For the specific contents, please refer to the description in the embodiment of the method of the present invention, and no further description is given here.

[0164] An embodiment of the present invention further provides a computing device, including a memory and a processor, wherein the memory stores a computer program, and when the processor executes the computer program, a method for calculating the light scattering characteristics of the ground atmosphere in any embodiment of the present invention is implemented.

[0165] An embodiment of the present invention further provides a computer-readable storage medium, on which a computer program is stored. When the computer program is executed by a processor, the processor executes a method for calculating the light scattering characteristics of the ground atmosphere in any embodiment of the present invention.

[0166] Specifically, a system or device equipped with a storage medium can be provided, on which software program code that implements the functions of any of the above-mentioned embodiments is stored, and a computer (or CPU or MPU) of the system or device can be enabled to read and execute the program code stored in the storage medium.

[0167] In this case, the program code itself read from the storage medium can realize the function of any one of the above-mentioned embodiments, and thus the program code and the storage medium storing the program code constitute a part of the present invention.

[0168] The storage medium embodiments for providing the program code include a floppy disk, a hard disk, a magneto-optical disk, an optical disk (such as CD-ROM, CD-R, CD-RW, DVD-ROM, DVD-RAM, DVD-RW, DVD+RW), a magnetic tape, a non-volatile memory card, and a ROM. Alternatively, the program code can be downloaded from a server computer by a communication network.

[0169] In addition, it should be clear that the functions of any of the above embodiments can be implemented not only by executing the program code read by the computer, but also by enabling an operating system operating on the computer to complete part or all of the actual operations based on instructions from the program code.

[0170] In addition, it can be understood that the program code read from the storage medium is written to a memory provided in an expansion board inserted into the computer or to a memory provided in an expansion module connected to the computer, and then based on the instructions of the program code, a CPU installed on the expansion board or expansion module is enabled to perform part or all of the actual operations, thereby realizing the functions of any of the above-mentioned embodiments.

[0171] An embodiment of the present application also provides a computer-readable storage medium, on which is stored at least one instruction, at least one program, code set or instruction set, and the at least one instruction, at least one program, code set or instruction set is loaded and executed by a processor to implement a method for calculating the ground-air light scattering characteristics provided by the above-mentioned method embodiments.

[0172] It should be noted that, in this article, relational terms such as first and second are only used to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Moreover, the terms "include", "comprise" or any other variants thereof are intended to cover non-exclusive inclusion, so that a process, method, article or device including a series of elements includes not only those elements, but also other elements not explicitly listed, or also includes elements inherent to such process, method, article or device.

[0173] A person of ordinary skill in the art can understand that all or part of the steps of implementing the above method embodiments can be completed by hardware related to program instructions, and the aforementioned program can be stored in a computer-readable storage medium. When the program is executed, it executes the steps of the above method embodiments; and the aforementioned storage medium includes: ROM, RAM, magnetic disk or optical disk, etc., various media that can store program codes.

[0174] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit it. Although the present invention has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the aforementioned embodiments, or make equivalent replacements for some of the technical features therein. However, these modifications or replacements do not deviate the essence of the corresponding technical solutions from the spirit and scope of the technical solutions of the embodiments of the present invention.

Claims

1. A method for calculating the light scattering characteristics of the ground atmosphere, characterized in that: include: The detection light is discretized within the detection field of view of the detector to obtain a plurality of detection lights; According to the position of the detector and the direction of each detection light, the detection scene corresponding to each detection light is determined, and the geometric characteristic parameter value and the geographical location of the scattering surface under the detection scene are calculated; wherein the detection scene includes an edge detection scene and a non-edge detection scene, and the geometric characteristic parameters corresponding to different detection scenes are different; Extracting the environmental characteristic parameter values ​​of each detection light at the geographical location of the scattering surface in the corresponding detection scene; Input each set of environmental characteristic parameter values ​​and geometric characteristic parameter values ​​into the light scattering characteristic prediction model of the corresponding detection scene respectively, and obtain the ground-air light scattering radiation brightness within the detection field of view under the detection scene; Among them, the light scattering characteristic prediction model includes an edge light scattering characteristic prediction model and a non-edge light scattering characteristic prediction model; the edge light scattering characteristic prediction model and the non-edge light scattering characteristic prediction model are both trained using known geometric characteristic parameter values, known environmental characteristic parameter values ​​and known ground light scattering radiation brightness values ​​as training sample sets.

2. The method according to claim 1, characterized in that The environmental characteristic parameters include surface environmental parameters, aerosol environmental parameters and cloud environmental parameters; wherein the surface environmental parameters include surface reflectivity, the aerosol environmental parameters include aerosol type and aerosol optical thickness, and the cloud environmental parameters include cloud top height and cloud optical thickness.

3. The method according to claim 1, characterized in that The geographical location of the scattering surface includes the scattering surface location at the intersection of the detection light and the earth and the scattering surface location at the tangent point of the detection light and the earth; The step of determining the detection scene corresponding to each detection light and calculating the geometric characteristic parameter value under the detection scene according to the position of the detector and the direction of each detection light includes: According to the position of the detector and the direction of each detection light, the intersection discriminant of each detection light and the earth is constructed respectively; Based on each of the intersection discriminants, respectively determining the detection scene corresponding to each detection light; If the intersection discriminant is greater than zero, then the detection light has an intersection with the earth, and the detection scene corresponding to the current detection light is determined as a non-edge detection scene, and the intersection position and geometric characteristic parameter values ​​under the detection scene are calculated respectively; If the intersection discriminant is less than zero, the detection light has no intersection with the earth, and the detection scene corresponding to the current detection light is determined as an edge detection scene, and the tangent point position and geometric characteristic parameter values ​​under the detection scene are calculated.

4. The method according to claim 3, characterized in that The intersection discriminant of the detection light and the earth is as follows: Δ=(p·d) 2 -(p·pR E 2 ) Where p is the coordinate position of the detector, d is the direction of the detection light, R E is the radius of the Earth; and / or The vector of the intersection point and the vector of the tangent point are calculated by the following formula: r int =p+t int d In the formula, r int is the vector of the intersection / tangent point, t int is the distance between the detector and the intersection / tangent point. In non-edge detection scenarios, In the edge detection scenario, t int =-p·d.

5. The method according to claim 4, characterized in that The geometric characteristic parameters in the non-edge detection scenario include the solar zenith angle, the relative azimuth angle and the observation zenith angle, and the geometric characteristic parameters in the edge detection scenario include the solar zenith angle, the relative azimuth angle and the cutting height; Among them, the solar zenith angle and relative azimuth angle in the non-limb detection scenario and the limb detection scenario are calculated by the following formula: i i =arccos(s·n) In the formula, θ i The solar zenith angle, s is the unit vector pointing from the intersection / tangent point to the sun, n is the unit vector pointing from the center of the earth to the intersection / tangent point, is the relative azimuth, s n is the projection vector of vector s on the plane at the observation point, e n is the projection vector of vector e on the plane at the observation point; The cutting height in the edge detection scenario is calculated by the following formula: Wherein, H is the cutting height, OP is the vector pointing from the detector position p to the center of the earth; The observation zenith angle in the non-edge detection scenario is calculated by the following formula: θ r =arcs(e·n) In the formula, θ r is the observation zenith angle, and e is the unit vector pointing from the intersection position to the detector position.

6. The method according to claim 1, characterized in that The training sample sets are respectively obtained by simulating three-dimensional radiation transmission based on data sets consisting of environmental characteristic parameter values ​​and geometric characteristic parameter values ​​in edge detection scenarios and non-edge detection scenarios.

7. A device for calculating the light scattering characteristics of the ground atmosphere, characterized in that: include: A discrete unit, used for discretizing the detection light within the detection field of the detector to obtain a plurality of detection lights; A determination unit, used to determine the detection scene corresponding to each detection light according to the position of the detector and the direction of each detection light, and calculate the geometric characteristic parameter value and the geographical location of the scattering surface under the detection scene; wherein the detection scene includes an edge detection scene and a non-edge detection scene, and the geometric characteristic parameters corresponding to different detection scenes are different; An extraction unit, used to respectively extract the environmental characteristic parameter value of each detection light at the geographical location of the scattering surface in the corresponding detection scene; The prediction unit is used to input each set of environmental characteristic parameter values ​​and geometric characteristic parameter values ​​into the light scattering characteristic prediction model of the corresponding detection scene, and obtain the ground-air light scattering radiation brightness within the detection field of view under the detection scene; Among them, the light scattering characteristic prediction model includes an edge light scattering characteristic prediction model and a non-edge light scattering characteristic prediction model; the edge light scattering characteristic prediction model and the non-edge light scattering characteristic prediction model are both trained using known geometric characteristic parameter values, known environmental characteristic parameter values ​​and known ground light scattering radiation brightness values ​​as training sample sets.

8. A computing device, comprising a memory and a processor, wherein the memory stores a computer program, and when the processor executes the computer program, the method according to any one of claims 1 to 6 is implemented.

9. A computer-readable storage medium having a computer program stored thereon, which, when executed in a computer, causes the computer to execute the method according to any one of claims 1 to 6.

10. A computer program product, characterized in that The method comprises a computer program, wherein when the computer program is executed by a processor, the steps of the method according to any one of claims 1 to 6 are implemented.

Images