On-satellite calibration data processing method and system based on solar cone on-satellite calibrator

Abstract

The invention provides an on-satellite calibration data processing method and system based on a sunlight cone on-satellite calibrator, and the method comprises the steps: S1, collecting data, and constructing an on-satellite calibration database; s2, simulating an incident angle when sunlight enters the sunlight cone, and further calculating and obtaining irradiance of the remote sensor; s3, constructing a BTDF model of the solar cone based on the irradiance, and performing angle interpolation to obtain a multi-angle BTDF model; and S4, calculating and obtaining the radiometric calibration coefficient of the remote sensor, and further evaluating the on-satellite radiometric calibration precision. According to the invention, interpolation is carried out on different incident angles to obtain a multi-angle BTDF statistical model, a bidirectional transmission factor of a solar cone is expanded to a whole solar reflection wave band area through polynomial interpolation, and an angle and wavelength dependent BTDF statistical model is further constructed. Dynamic simulation of the radiation performance of the solar cone is realized by constructing a large number of data unit BTDF models in the life cycle of a remote sensor.

Description

Technical Field

[0001] The present invention belongs to the technical field of on-board calibration of remote sensors, and specifically relates to an on-board calibration data processing method and system based on a sunlight cone on-board calibrator, which is used to uniformly process 10 years of on-board calibration data of remote sensors to obtain on-board absolute radiometric calibration coefficients of remote sensors with uniform radiometric calibration accuracy. Background Art

[0002] After a remote sensor has been operating in orbit, its output characteristics degrade due to factors such as aging of the calibration system and decreased efficiency of optical components, making it difficult to accurately measure. Onboard calibrators can only be used for research purposes to track sensor response degradation, making it difficult to achieve absolute on-orbit radiometric calibration of the sensor. This makes it difficult to guarantee the calibration accuracy of the sensor and the long-term stability and reliability of the remote sensing data products.

[0003] At present, the existing on-board calibration method that combines the on-board calibration source and the remote sensor output DN value to determine the radiation calibration coefficient cannot meet the on-orbit absolute radiation calibration requirements of the remote sensor. It is urgent to develop an independent and effective on-board calibration method to achieve the on-orbit absolute radiation calibration of the remote sensor.

[0004] Furthermore, because each calibration method has different calibration accuracies and error factors, the calibration accuracy obtained using different calibration methods varies. For the same remote sensor, it is difficult to assess its long-term radiation response changes by referencing and combining calibration coefficients obtained using different calibration methods over its lifetime.

[0005] After on-orbit operation, the calibration system's output radiation characteristics decay due to factors such as aging, making them difficult to accurately measure. Consequently, onboard absolute radiometric calibration of remote sensors becomes impossible, impacting the quantitative inversion accuracy of remote sensing data products. Furthermore, both domestic and international radiometric calibration accuracy assessments rely on theoretical derivation, combining the errors of each theoretical error term to derive the total calibration error. Because the calibration accuracy and error factors of each calibration method vary, the calibration accuracy achieved using different calibration methods varies. Therefore, establishing a unified calibration accuracy evaluation method is a pressing challenge in current calibration work.

[0006] Patent document CN113091892A discloses a method and system for on-orbit absolute radiometric calibration of satellite remote sensors. This solution includes: calculating the total radiation flux of sunlight entering the solar cone and then the on-board calibrator using simulation software; using a simulation device to adjust a solar simulator in a laboratory so that the light source in the incident integrating sphere has the same total radiation flux; measuring the radiance of the light beam after homogenization by the integrating sphere and expansion by a beam expansion system and converting it into reflectivity; substituting the reflectivity and the sensor's radiation output count value DN into a radiometric calibration formula to calculate the remote sensor's on-orbit absolute radiometric calibration coefficient. This method uses sunlight entering the on-board calibration system's solar cone and then the on-board calibrator as the radiation calibration light source, resulting in a high calibration frequency and high calibration accuracy, while avoiding the effects of radiation performance degradation of the on-board calibration device itself, changes in atmospheric conditions, and the accuracy of the atmospheric radiation transmission model.

[0007] In other words, this scheme constructs a satellite model that is consistent with the actual satellite in orbit, simulates the satellite's operating status through simulation software, and simulates the calculation of the incident light irradiance E received by the on-board calibrator of the remote sensor at a certain moment. According to the conversion relationship obtained in the laboratory based on the solar simulator, E is converted into the reflectivity ρ. It is necessary to build a laboratory solar simulator simulation system, and according to the count value expressed in DN output by the satellite product at that moment, substitute it into the radiation calibration formula to calculate the on-orbit absolute radiation calibration coefficient A of the remote sensor based on the reflectivity; the radiation calibration formula 1) is: A*DN=ρ.

[0008] However, due to factors such as aging of the remote sensor calibration system and decreased efficiency of optical components, the correspondence between the calibration system input and output decays and becomes difficult to accurately measure. While the sensor's output DN value can be extracted from the sensor data product, simulation calculations based on the calibration input energy of sunlight can only calculate irradiance E, but cannot provide the radiance L or reflectance ρ required for calibration calculations. Furthermore, laboratory measurement costs are high, and measurement accuracy is affected by various factors, including the accuracy of the solar simulator itself, making it unsuitable for high-frequency radiation calibration calculations involving large amounts of data.

[0009] This problem needs to be solved urgently. Summary of the Invention

[0010] In view of the defects in the prior art, the object of the present invention is to provide an on-board calibration data processing method and system based on a solar cone on-board calibrator.

[0011] According to the present invention, a method for processing on-board calibration data based on a solar cone on-board calibrator includes: step S1: collecting data and constructing an on-board calibration database;

[0012] Step S2: Based on the onboard calibration database, simulate the incident angle of sunlight incident on the sunlight cone, and then calculate and obtain the incident irradiance of the sunlight cone of the remote sensor;

[0013] Step S3: constructing a BTDF model of the sunlight cone based on the irradiance, and performing angle interpolation to obtain a multi-angle BTDF model;

[0014] Step S4: Calculate and obtain the calibration coefficient of the remote sensor based on the BTDF model and the radiation output DN value and irradiance of the remote sensor, and then evaluate the on-board radiation calibration accuracy.

[0015] Preferably, in step S1, the data includes: atmospheric parameters, L1-level image data, on-board calibration data and ground-based observation data; the on-board calibration data includes: cold air calibration data and sunlight calibration data;

[0016] In step S1, the sunlight calibration data of the onboard calibration data is extracted, that is, the sunlight calibration radiation output count value, referred to as DN'; the cold air count value of DN' is removed to obtain the change count value caused by sunlight;

[0017] The mathematical expression of the change count value caused by sunlight is:

[0018] DN=DN'-DN0

[0019] Wherein, DN represents the change count value caused by sunlight, DN' represents the sunlight calibration radiation output count value; DN0 represents the background noise.

[0020] Preferably, the step S2 includes:

[0021] Step S2.1: Input the satellite orbit, run the STK software to simulate the satellite's on-orbit operation state, and obtain the simulation results;

[0022] Step S2.2: Based on the simulation results, query the incident angle of sunlight; the incident angle includes the solar zenith angle and the solar azimuth angle;

[0023] Step S2.3: Run TracePro software to calculate and obtain the irradiance at the entrance pupil of the remote sensor based on the incident angle;

[0024] In step S3, the mathematical expression of the multi-angle BTDF model is:

[0025]

[0026] Which is then simplified to:

[0027]

[0028] Among them, BTDF(θ i ,φ i ,θ t ,φ t ) represents the multi-angle BTDF model, θ t With φ t They represent the zenith angle and azimuth angle of the transmitted light of the sun cone, L(θ i ,φ i ,θ t ,φ t ) is the solar radiance received by the remote sensor; A is the calibration coefficient; E(θ i ,φ i ) represents irradiance.

[0029] Preferably, in step S4, the on-board radiation calibration accuracy is evaluated by cross calibration; the cross calibration process includes:

[0030] Step A1: extract the DN value of the image when the reference remote sensor passes through the target site;

[0031] Step A2: Calculate and obtain the normalized apparent reflectance based on the image DN value and the calibration coefficient;

[0032] Step A3: Correcting the normalized apparent reflectivity by the solar zenith angle and the sun-earth distance factor to obtain the top of atmosphere reflectivity;

[0033] Step A4: Based on the top of atmosphere reflectivity, the surface reflectivity in the remote sensor observation direction is calculated; the surface reflectivity to be evaluated is obtained by multi-angle BRDF model correction, and then the top of atmosphere reflectivity to be evaluated is calculated;

[0034] Step A5: Calculating and obtaining a cross-calibration coefficient based on the top-of-atmosphere reflectivity to be evaluated, the solar zenith angle, the sun-earth distance factor, and the average DN value of the target area image;

[0035] Step A6: determining whether the cross calibration coefficient is consistent with the onboard calibration coefficient; if yes, no processing is performed; if no, correcting the calibration coefficient of the remote sensor;

[0036] Step A7: Determine whether the atmospheric parameters extracted from the data product obtained based on the on-board calibration coefficients are consistent with the actually measured atmospheric parameters. If the result is yes, no processing is performed; if the result is no, the calibration coefficients of the remote sensor are corrected.

[0037] Preferably, in step A5, the calculation expression of the cross calibration coefficient is:

[0038]

[0039] Among them, ρ * is the reflectivity of the top of the atmosphere, which comes from the 6S radiation transfer calculation and is directional, θ s is the solar zenith angle, A is the correction factor for the average and actual distance between the sun and the earth; ρ is the cross calibration coefficient;

[0040] The 6S radiation transfer calculation uses the 6S model jointly developed by the French Atmospheric Optics Laboratory and the University of Maryland in the United States to calculate atmospheric scattering and absorption. The model uses the latest approximation and successive scattering algorithms, takes polarization into account, and improves the parameter input of the model.

[0041] In step A6, it is determined whether the ozone product obtained as a data product based on the cross-calibration coefficient is consistent with the ozone content actually measured on the ground. If the result is yes, no processing is performed; if the result is no, the calibration coefficient of the remote sensor is corrected.

[0042] According to the present invention, an on-board calibration data processing system based on a solar cone on-board calibrator is provided, comprising:

[0043] Data preprocessing module: collects data and builds on-board calibration database;

[0044] Simulation module: Based on the onboard calibration database, it simulates the incident angle of sunlight when it enters the sunlight cone, and then calculates and obtains the irradiance of the remote sensor;

[0045] BTDF modeling module: Based on the irradiance, a BTDF model of the sunlight cone is constructed, and angle interpolation is performed to obtain a multi-angle BTDF model;

[0046] Calibration coefficient calculation module: calculates and obtains the calibration coefficient of the remote sensor according to the BTDF model and the radiation output DN value and irradiance of the remote sensor;

[0047] Calibration accuracy verification module: based on the calibration coefficient, evaluates the on-board radiation calibration accuracy and corrects the on-board radiation calibration coefficient.

[0048] Preferably, in the data preprocessing module, the data includes: atmospheric parameters, L1-level image data, on-board calibration data and ground-based observation data; the on-board calibration data includes: cold air calibration data and sunlight calibration data;

[0049] In the data preprocessing module, the sunlight calibration data of the onboard calibration data is extracted, that is, the sunlight calibration radiation output count value, referred to as DN'; the cold air count value of DN' is removed to obtain the change count value caused by sunlight;

[0050] The mathematical expression of the change count value caused by sunlight is:

[0051] DN=DN'-DN0

[0052] Wherein, DN represents the change count value caused by sunlight, DN' represents the sunlight calibration radiation output count value; DN0 represents the background noise.

[0053] Preferably, the simulation module includes:

[0054] Simulation submodule 1: Input the satellite orbit, run the STK software to simulate the satellite's on-orbit operation status, and obtain the simulation results;

[0055] Simulation submodule 2: Based on the simulation results, query the incident angle of sunlight; the incident angle includes the solar zenith angle and the solar azimuth angle;

[0056] Simulation submodule three: running TracePro software to calculate and obtain the irradiance at the entrance pupil of the remote sensor based on the incident angle;

[0057] In the BTDF modeling module, the mathematical expression of the multi-angle BTDF model is:

[0058]

[0059] Which is then simplified to:

[0060]

[0061] Among them, BTDF(θ i ,φ i ,θ t ,φ t ) represents the multi-angle BTDF model, θ t With φ t They represent the zenith angle and azimuth angle of the transmitted light of the sun cone, L(θ i ,φ i ,θ t ,φ t ) is the solar radiance received by the remote sensor; A is the calibration coefficient; E(θ i ,φ i ) represents irradiance.

[0062] Preferably, in the calibration coefficient calculation module, the on-board radiation calibration accuracy is evaluated and the on-board radiation calibration coefficient is corrected by cross-calibration; the modules involved in the cross-calibration include:

[0063] Module A1: Extract the DN value of the image when the remote sensor passes through the target site;

[0064] Module A2: Calculate and obtain the normalized apparent reflectance based on the image DN value and calibration coefficient;

[0065] Module A3: Correcting the normalized apparent reflectivity by the solar zenith angle and the distance factor between the sun and the earth to obtain the reflectivity of the top of the atmosphere;

[0066] Module A4: Based on the top of atmosphere reflectivity, the surface reflectivity in the remote sensor observation direction is calculated; the surface reflectivity to be evaluated is obtained by multi-angle BRDF model correction, and then the top of atmosphere reflectivity to be evaluated is calculated;

[0067] Module A5: Calculating and obtaining a cross-calibration coefficient based on the top-of-atmosphere reflectivity to be evaluated, the solar zenith angle, the sun-earth distance factor, and the average DN value of the target area image;

[0068] In the calibration accuracy verification module, it is determined whether the atmospheric parameters extracted from the data product obtained based on the cross calibration coefficient are consistent with the actually measured atmospheric parameters. If the result is yes, no processing is performed; if the result is no, the radiation calibration accuracy is evaluated and the radiation calibration coefficient of the remote sensor is corrected.

[0069] Preferably, the calculation expression of the cross calibration coefficient in the module A5 is:

[0070]

[0071] Among them, ρ * is the reflectivity of the top of the atmosphere, which comes from the 6S radiation transfer calculation and is directional, θ s is the solar zenith angle, A is the correction factor for the average and actual distance between the sun and the earth; ρ is the cross calibration coefficient; represents the product;

[0072] In the calibration accuracy verification module, it is determined whether the ozone product obtained as a data product based on the cross-calibration coefficient is consistent with the ozone content actually measured on the ground. If the result is yes, no processing is performed; if the result is no, the calibration coefficient of the remote sensor is corrected.

[0073] Compared with the prior art, the present invention has the following beneficial effects:

[0074] 1. The present invention uses stable sunlight as the calibration light source and combines orbit simulation technology to construct the BTDF model of the sunlight cone to achieve the incident irradiance of the sunlight cone, that is, E(θ i ,φ i) and the solar radiance received by the remote sensor to obtain the incident radiance L during the remote sensor sunlight calibration, which solves the problem of the difficulty in measuring the absolute value of the light source radiation at the entrance pupil and can realize the absolute radiation calibration of the remote sensor on the orbiting satellite.

[0075] 2. The present invention obtains a multi-angle BTDF statistical model by interpolating different incident angles, expands the two-way transmission factor of the sunlight cone to the entire solar reflection band through polynomial interpolation, and further constructs an angle- and wavelength-dependent BTDF statistical model. By constructing a large number of data unit BTDF models within the life cycle of the remote sensor, the dynamic simulation of the radiation performance of the sunlight cone is realized. The multi-angle, wavelength-dependent, dynamic BTDF model constructed by the present invention can improve the on-board absolute radiation calibration accuracy of the remote sensor.

[0076] 3. The present invention adopts an on-board calibration method to perform unified radiation calibration processing on the long-term series radiation calibration data of the remote sensor, which can obtain long-term, high-frequency and uniform on-board absolute radiation calibration coefficients with uniform radiation calibration accuracy, solving the problem that the traditional calibration accuracy is difficult to evaluate.

[0077] 4. This paper proposes to construct a multi-angle, wavelength-dependent, dynamic BTDF model to further improve the accuracy of the onboard absolute radiometric calibration of remote sensors. A cross-calibration method with a high-precision remote sensor is used to obtain the cross-calibration coefficients and verify the onboard calibration accuracy.

[0078] 5. The present invention further evaluates the calibration accuracy by combining the difference between the inversion parameters and the actual foundation measurement values. BRIEF DESCRIPTION OF THE DRAWINGS

[0079] Other features, objects and advantages of the present invention will become more apparent upon reading the detailed description of non-limiting embodiments with reference to the following drawings:

[0080] Figure 1 A schematic diagram of the technical process provided by the present invention;

[0081] Figure 2 This is a schematic diagram of the relationship between the on-board calibration data provided by the present invention and time. In the figure, 1 represents the DN' value of the remote sensor when calibrated with sunlight; 2 represents the DN' value when the remote sensor is calibrated with dual calibration lights; 3 represents the DN' value when the remote sensor is calibrated with a single calibration light; 4 represents the DN' value when the remote sensor is calibrated with dual calibration lights and the calibration light source is stable; 5 represents the DN' value when the remote sensor is not calibrated.

[0082] Figure 3 Schematic diagram of the BSDF measurement device provided by the present invention;

[0083] Figure 4Schematic diagram of the ABg model parameter inversion principle provided by the present invention. DETAILED DESCRIPTION

[0084] The present invention will be described in detail below with reference to specific embodiments. The following examples will help those skilled in the art to further understand the present invention, but are not intended to limit the present invention in any form. It should be noted that, for those skilled in the art, several changes and improvements can be made without departing from the scope of the present invention. These all fall within the scope of protection of the present invention.

[0085] The present invention uses an on-orbit calibration data statistical method to establish a BTDF model, replacing the laboratory measurement method using a solar simulation source. It establishes a conversion between the known incident radiation illuminance E and the incident radiance L, obtains the remote sensor radiation energy L, and calculates the remote sensor's on-orbit absolute radiation calibration coefficient in combination with the calibration formula. The incident irradiance E is calculated based on simulation software;

[0086] The advantages of the present invention are: first, the present invention calculates the incident energy L through a statistical method, and can interpolate different incident angles to obtain the radiation energy under different incident angles, thereby improving the calculation accuracy; second, the BTDF model of the simulation calculation can be updated according to the selected time, which is suitable for the calculation of a large amount of radiation calibration data over a long period of time.

[0087] Establishing the conversion relationship between irradiance E and radiance L or between irradiance E and reflectivity ρ is a difficult problem in solving on-orbit absolute radiation calibration. The present invention establishes a BTDF model through a statistical method to realize the conversion between irradiance E and radiance L, calculate the incident radiance L of the remote sensor, and substitute it into the radiation calibration formula to realize the on-orbit absolute radiation calibration of the remote sensor. In addition, the construction of this model facilitates the calculation of a large number of radiation calibrations, and is used to calculate the on-orbit absolute radiation calibration coefficient of the remote sensor for about 10 years. Moreover, the calibration method is unified over a long period of time, the calibration error factor is the same, and the obtained calibration accuracy is unified, which is beneficial to the long-term radiation attenuation assessment of the remote sensor. Among them, the calculation of the radiation calibration coefficient can adopt the reflectivity method or the radiance method.

[0088] The present invention constructs a satellite model consistent with a real on-orbit satellite, simulates the satellite's operating status through simulation software, and simulates the calculation of the incident light irradiance E received by the on-board calibrator of the remote sensor at a certain moment. Specifically, based on about 10 years of on-orbit calibration data, the correspondence between the output count DN value of the remote sensor during radiation calibration and the incident irradiance E is statistically analyzed, and a bidirectional transmittance distribution model, namely the BTDF model, is constructed. E is converted into radiance L, which is substituted into the radiation calibration formula 2): A*DN=L, and then the on-orbit absolute radiation calibration coefficient A of the remote sensor based on the radiance is calculated.

[0089] The incident irradiance E is related to the incident angle of sunlight.

[0090] To address the difficulty in tracing the radiation reference of onboard calibrators, this paper proposes an onboard calibration data processing system and method based on a solar cone onboard calibrator to achieve onboard absolute radiation calibration of remote sensors. The research includes:

[0091] 1. Explore the dependence of sunlight calibration timing on the solar incidence angle, and select qualified sunlight-based on-board calibration data from a large amount of multi-calibration light source on-board calibration data;

[0092] 2. Introduce and improve statistical methods to achieve the conversion between the ratio of DN value and irradiance and the ratio of radiance and irradiance. Develop a bidirectional transmission distribution function (BTDF) model for the sunlight cone depending on the angle of incidence of sunlight. Combined with irradiance simulation calculations, the incident absolute radiance L is obtained, solving the problem of tracing the radiation benchmark on board the satellite.

[0093] 3. Construct a calibration evaluation system based on long-term, high-frequency on-board calibration coefficients. This evaluation system uses a cross-calibration method and combines the difference between the inversion parameters and the ground-based measurements to verify the calibration accuracy, thereby achieving long-term, high-frequency, and consistent calibration accuracy of the remote sensor on-orbit radiation performance evaluation with a unified reference standard.

[0094] The purpose of the present invention is to provide a data processing system and method for on-board calibration based on a solar cone on-board calibrator. The system attempts to establish a relationship between the known output DN value of the remote sensor and the angle-dependent irradiance E by establishing a DN / E(θ i ,φ i ) and L / E(θ i ,φ i ) and construct a new BTDF model to realize the conversion of irradiance to radiance, in order to overcome the current difficulty in measuring the transmitted radiance L that depends on the angle of the sunlight cone.

[0095] like Figure 1 As shown, the present invention provides an on-board calibration data processing method based on a solar cone on-board calibrator, comprising:

[0096] Step 1: Data collection and processing: Collect L1 image data of the remote sensor and reference remote sensor, onboard calibration data, atmospheric parameter reanalysis data, and ground-based atmospheric measurement data to build an onboard calibration basic database;

[0097] Specifically, the calibration basic database includes remote sensor L1 image data and onboard calibration data, reference remote sensor L1 image data, and reference site surface and atmospheric parameter measurement data. The remote sensor onboard calibration data includes cold air calibration data, onboard calibration lamp calibration data, and sunlight calibration information.

[0098] In other words, the collected data include remote sensor L1-level image data, on-board calibration data, atmospheric parameter reanalysis data, and ground-based atmospheric measurement data. The on-board calibration data includes cold air calibration data, on-board calibration light calibration data, and sunlight calibration information.

[0099] The data collection steps include: collecting remote sensor image data, onboard calibration data based on the sun cone, reference remote sensor image data for calibration accuracy assessment, atmospheric parameter reanalysis data, and ground-based atmospheric measurement data;

[0100] Data preprocessing steps: According to the remote sensor onboard calibration file, extract the sunlight calibration radiation output count value, represented by DN', and remove the cold air count value, represented by DN0. Determine the time when sunlight enters the calibrator based on the changing count value;

[0101] Specifically, there are two types of onboard calibration data for remote sensors, both generated every 5 minutes:

[0102] First, the on-board calibration data product file, which reads some count values (DN' values) for a certain period of time, is shown in the figure below. The values close to the blue box 1 in the figure below, that is, position 1, are all sunlight calibration data, and those close to positions 3 and 4 are all calibration data with the on-board calibration light turned on;

[0103] Second, the cold sky observation data file has a count value of similar magnitude to that in position 5 in the figure below, denoted as DN0, which is background noise. The actual change count value DN caused by sunlight is calculated using the following formula:

[0104] DN=DN'-DN0

[0105] Since the DN value received by the detector corresponds to the time of receiving sunlight, the corresponding sunlight incident time t can be determined according to the change of the DN value.

[0106] In other words, the DN' value is extracted based on the remote sensor onboard calibration file. To reduce the impact of the calibration light source on the accuracy of the radiation calibration based on sunlight, the radiation calibration data during the light-on period and other light source calibration periods should be avoided. In addition, the cold air count value DN0 needs to be removed from the extracted DN' value during data processing.

[0107] Specifically, to improve calibration accuracy, radiation calibration data from non-sunlight calibration sources should be avoided. The sensor radiation output DN' value is extracted from the remote sensing dataset, and the cold air count value is removed. The time when sunlight enters the solar cone is determined based on the changing DN value.

[0108] Step 2: Simulation Calculation: Based on the satellite orbit report, STK simulation software was used to simulate the satellite's actual on-orbit operation and the incident angle of sunlight entering the solar cone. The incident angle of the sunlight source was set in TracePro simulation software, and the satellite mechanical model was imported to simulate the irradiance at the focal plane of the remote sensor.

[0109] In other words, the simulation calculation steps are: based on the sunlight incident time and combined with the latest remote sensor orbit report, simulate the geometric angle of sunlight incident on the sunlight cone at that moment, and use TracePro simulation software to simulate the incident irradiance based on the incident angle;

[0110] Specifically, only sunlight incident on the sunlight cone is used as the onboard calibration light source. To reduce the impact of the calibration light source on the accuracy of the radiation calibration based on sunlight, the radiation calibration data during the light-on period and other light source calibration data are not used.

[0111] In other words, the simulation calculation steps are: according to the actual on-orbit operation status of the satellite, simulate the incident angle of sunlight entering the sunlight cone, perform irradiance simulation calculation based on the incident angle, and obtain the absolute irradiance value at the entrance pupil;

[0112] Specifically, STK simulation software is used to simulate the actual on-orbit operation state of the satellite. The simulation software inputs the time t to calculate the incident angle of sunlight corresponding to this time; the satellite mechanical model is imported into the TracePro simulation software, and the incident angle of the sunlight source is set. The simulation software can calculate the irradiance E at the entrance pupil of the remote sensor; the incident angle includes: the solar zenith angle θ i and the solar azimuth φ i ;

[0113] Step 3: BTDF construction: Construct the sunlight cone BTDF model and perform angle interpolation to obtain a multi-angle BTDF model;

[0114] Specifically, in this embodiment, the on-board calibration data of the remote sensor for many years are sorted into a cycle of 2 weeks. The data in one cycle is called a data unit. The DN value and E(θ i ,φ i There are about 196 on-board calibration data products in one cycle.

[0115] Assuming that the radiation attenuation of the remote sensor is very small within 2 weeks, the calibration coefficient A in the following formula is considered a constant. i ,φ i ) with the incident angle represents the trend of BTDF with the incident angle; specifically, due to the effect of the calibrator beam expander system, the transmission angle θ t ,φ t The BTDF in the direction perpendicular to the entrance of the sun's cone is set to 1, and the BTDF values in other directions are divided by the BTDF value in that direction for normalization. This results in a two-dimensional surface that changes with the angle, which is the BTDF function. The two-dimensional horizontal coordinates are the solar zenith angle θ i and the solar azimuth φ i The vertical axis is the normalized BTDF value, and its maximum value is 1. Since there are only about 196 calibration data in practice, multi-angle interpolation is performed on the surface to obtain a relatively smooth surface, that is, a multi-angle BTDF model is obtained.

[0116]

[0117] In other words, given the fitting function and fitting coefficient, BTDF is the BTDF distribution function; interpolation is performed for different incident angles to obtain the multi-angle BTDF statistical model;

[0118] In other words, by interpolating different incident angles, a multi-angle BTDF statistical model is obtained. The bidirectional transmission factor of the sunlight cone is extended to the entire solar reflection band through polynomial interpolation, and an angle- and wavelength-dependent BTDF statistical model is constructed. The data sets of all data units in the life cycle of the remote sensor are processed in the same way to achieve a long-term, high-frequency, and uniformly accurate on-board absolute radiometric calibration coefficient for the remote sensor.

[0119] In short, BTDF modeling: Starting from the correspondence between the known output DN value of the remote sensor and the angle-dependent irradiance E, a sunlight cone transmittance distribution model BTDF is constructed. By inputting the incident angle of sunlight, combining the simulated irradiance E and the constructed BTDF model, the transmitted radiance L can be calculated, that is, the radiance of the entrance pupil of the calibration light source. The irradiance E is a function of the incident angle, and E(θ i ,φ i )express.

[0120] The sun's cone has no shielding, so sunlight enters the cone every time a satellite passes over the South Pole. This paper proposes organizing the long-term onboard calibration data of remote sensors into two-week cycles, dividing the large amount of onboard calibration data into several data intervals.

[0121] For each data interval, according to a large number of DN values and the corresponding irradiance E(θ i ,φ i ) to establish a statistical model, that is, the comparison value DN / E(θ i ,φ i ) After normalization, the least squares method is used for fitting, and the fitting function and fitting coefficients are given to construct a dynamic BTDF model of the sunlight cone. Combining the BTDF model with the incident angle-dependent incident irradiance E obtained by simulation, the radiance L of the calibrated light source can be obtained;

[0122] Specifically, the radiance E is obtained by simulation using TracePro simulation software, which corresponds one-to-one to the incident angle of sunlight. The BTDF surface value of the corresponding angle is found and substituted into the following formula to obtain the radiance L value.

[0123]

[0124] Among them, BTDF(θ i ,φ i ,θ t ,φ t ) represents the multi-angle BTDF model, θ t With φ t They represent the zenith angle and azimuth angle of the transmitted light of the sun cone, L(θ i ,φ i ,θ t ,φ t ) is the solar radiance received by the remote sensor; A is the calibration coefficient; E(θ i ,φ i ) represents irradiance.

[0125] Step 4: Calibration coefficient calculation: Substitute the BTDF model, the sensor's radiation output DN value, and the simulated irradiance into the calibration formula to calculate the sensor's on-orbit absolute radiation calibration coefficient.

[0126] In other words, the calibration coefficient calculation steps are: Based on the obtained transmitted radiance L and the extracted DN value at the corresponding time, combined with the radiometric calibration formula, the radiometric calibration coefficient of the remote sensor is calculated. The long-term series of onboard calibration data of the remote sensor is uniformly radiometrically processed to obtain the radiometric calibration coefficient of the remote sensor with long-term, high-frequency and uniform radiometric calibration accuracy.

[0127] Specifically, combined with the radiation calibration formula: A*DN=L, the radiation calibration coefficient A of the remote sensor is calculated, where * represents the product; specifically, according to the satellite orbit report, STK simulation software is used to simulate the actual on-orbit operation status of the satellite, and the incident angle of sunlight is simulated according to the time when sunlight enters the sunlight cone; TracePro simulation software is used to simulate the irradiance and establish a three-dimensional mechanical model of the satellite to ensure that the profile of the satellite platform and optical payload is consistent with the actual product status of the satellite, and the surface reflectivity characteristics of the satellite model are established based on the actual measurement results of the satellite surface material.

[0128] In other words, the imported model is consistent with the real product state of the satellite, and the surface reflectivity characteristics of the satellite model are established based on the actual measurement results of the satellite surface material;

[0129] Specifically, when performing irradiance simulation calculations in TracePro simulation software, a satellite model must be imported and the satellite's surface reflectance characteristics, such as reflectivity and scattering, must be configured. This is because sunlight striking the satellite surface undergoes multiple scattering and reflections before entering the instrument and being received by the detector. Valid information for sunlight calibration should be sunlight directly incident on the calibrator, while light reflected from other surfaces and received is considered stray light. Therefore, during in-orbit operation, simulation accuracy is affected not only by the satellite's external shape but also by its surface reflectivity, specular reflection, diffuse reflection, and other reflective or scattering properties. Therefore, the imported satellite model's external shape should be consistent with the actual satellite, and its surface reflectivity should approximate the satellite's actual state.

[0130] Specifically, a simplified version of the satellite's real-world model was directly imported, while retaining its outline. The inventor's organization, responsible for the satellite's overall development, had access to simplified versions of various real-world satellite models for simulation calculations. This simplified version of the real-world model removed internal optical components while retaining its outline.

[0131] Specifically, in this embodiment, the satellite surface is coated with polyimide material, and the heat dissipation surface is S781 thermal control coating, which is a kind of white paint, or other materials. These surface materials are common surface materials for satellites, and their surface scattering characteristics (BSDF) properties are generally measured in the laboratory. Different materials determine the scattering characteristics of the light scattered by the satellite surface. When sunlight hits the satellite surface, the light energy entering the instrument is different due to different scattering characteristics. Therefore, the irradiance simulation needs to take into account the characteristics of the satellite surface. Its measurement device is as follows: Figure 3 shown.

[0132] The scattering characteristics are generally simulated by the ABg model. The ABg model parameters are inverted based on the BSDF values measured in the laboratory. Here, A, B, and g are the three parameters that make the formula consistent with the measured results. Their values are related to the material of the star, the polishing method, and the polishing angle. The expression for the BSDF value is as follows:

[0133]

[0134] in, is the unit vector of the scattering direction projection on the surface, is the unit vector of the mirror direction Projection on the surface, see Figure 4 . BSDF, that is, BSDF value.

[0135] Substituting the known BTDF model, DN value, and calculated irradiance into the following formula, the corresponding calibration coefficient A can be calculated. In other words, the BTDF distribution function is described by the following formula:

[0136]

[0137] Among them, θ i and φ i are the zenith angle and azimuth angle of the incident light cone of the sun, θ t and φ t are the zenith angle and azimuth angle of the transmitted light of the sunlight cone, L(θ i ,φ i ,θ t ,φ t ) is the solar radiance received by the remote sensor, A is the calibration coefficient; the BTDF in the direction perpendicular to the entrance of the sunlight cone is set to 1. BTDF is the BTDF distribution function.

[0138] The datasets of all data units in the life cycle of the remote sensor are processed in the same way to achieve long-term, high-frequency on-board absolute radiometric calibration of the remote sensor.

[0139] Step 5: Calibration accuracy verification: The present invention uses a cross-calibration method to evaluate the calibration accuracy of the long-time series on-board calibration data, and further evaluates the calibration accuracy by combining the difference between the inversion parameters and the actual measurement values.

[0140] In other words, based on the long-term, high-frequency on-board calibration coefficients, a remote sensor with higher calibration accuracy is used to perform cross-calibration to verify the on-board calibration accuracy. A high-precision reference remote sensor and a similar remote sensor to be calibrated observe the same target. The observation data of the remote sensor to be calibrated is inferred based on the observation data of the high-precision reference remote sensor. Combined with the output DN value of the remote sensor to be calibrated, the cross-calibration coefficient is calculated, thereby achieving the radiometric calibration accuracy evaluation of the remote sensor to be calibrated. The steps of cross-calibration include:

[0141] Step 1: Extract the DN value of the image when the high-precision reference remote sensor passes through the target site;

[0142] Step 2: Calculate the normalized apparent reflectance based on the image DN value and the extracted apparent reflectance calibration coefficient and bias;

[0143] Specifically, the apparent reflectance calibration coefficients and biases are stored in the remote sensing data products and only need to be extracted, just as the DN' value is directly extracted. The apparent reflectance calibration coefficients and biases here refer to the apparent reflectance calibration coefficients and biases stored in the high-precision remote sensor data product used for reference, not the remote sensor to be calibrated.

[0144] The reference sensor's apparent reflectance calibration coefficient and bias, along with its DN value, can be used to calculate the reference sensor's top-of-atmosphere reflectance. Theoretically, the sensor to be calibrated should measure the same value. This means the sensor to be calibrated uses this top-of-atmosphere reflectance as its true measurement value and substitutes the measured DN value into the calibration formula to determine the calibration coefficient. This is known as the cross-calibration coefficient for the sensor to be calibrated, obtained using the cross-calibration method. The top-of-atmosphere reflectance value is corrected for geometric and spectral differences.

[0145] Step 3: Correct the normalized apparent reflectance by the solar zenith angle and the distance factor between the sun and the earth to obtain the reflectance of the top of the atmosphere;

[0146] Step 4: Perform radiative transfer calculation on the top of the atmosphere reflectivity and make atmospheric corrections to obtain the surface reflectivity in the reference remote sensor observation direction. Perform BRDF correction on the reflectivity to obtain the surface reflectivity in the remote sensor observation direction to be evaluated.

[0147] Specifically, BRDF correction does not refer to the "multi-angle BTDF model" correction mentioned above, but both are used for geometric correction. BRDF correction is used for cross-calibration, and is used for geometric correction of the ground reflectivity between the reference satellite observation direction and the observation direction of the remote sensor to be calibrated. The multi-angle BTDF model is used for the geometric correction of sunlight incidence on the calibrator on the remote sensor to be calibrated.

[0148] The BRDF of the site under the remote sensor observation geometry can be simulated using the kernel-driven model, which is expressed by the following equation:

[0149] E(θ i ,θ v ,φ,λ)=f iso (λ)+f vol (λ)k vol (θ i ,θ v ,φ,λ)+f geo (λ)k geo (θ i ,θ v ,φ,λ)

[0150] The model uses a linear combination of kernels to fit the bidirectional reflection characteristics of the surface. These kernels have certain physical meanings, namely, isotropic scattering kernel, geometric optics scattering kernel, i.e., k geo (θ i ,θ v ,φ,λ) and the volume scattering kernel, i.e., k vol (θ i ,θ v ,φ,λ). λ represents wavelength; θ v represents the zenith angle of the reflection direction; θ i represents the zenith angle of the incident direction; φ represents the relative azimuth; R(θ i ,θ v ,φ,λ) represents the BRDF model.

[0151] f iso 、f vol With f geo The constant coefficients are: one constant coefficient, another constant coefficient, and another constant coefficient; they represent the proportional factors for isotropic scattering, geometric optics scattering, and volume scattering, respectively. The solar zenith angle, solar azimuth angle, satellite zenith angle, and satellite azimuth angle corresponding to the image product of the uncalibrated remote sensor passing over the site are substituted into the kernel model calculation formula to calculate the geometric optics scattering kernel and volume scattering kernel. Using the constant coefficients and scattering kernels, the surface reflectance of the reference remote sensor at the corresponding channel wavelength corresponding to the observation direction of the uncalibrated remote sensor is calculated.

[0152] Step 5: The surface reflectance after spectral matching is calculated through radiation transfer to obtain the top of atmosphere reflectance;

[0153] Step 6: Combining the solar zenith angle, the sun-earth distance correction factor, and the average DN value of the extracted target area image, the cross-calibration coefficient of the remote sensor to be evaluated can be calculated.

[0154] The obtained cross-calibration coefficients are applied to the remote sensor to be evaluated. The quantitative data products generated by the sensor are then used to extract the measured atmospheric parameters. For example, the ozone content can be extracted from the ozone product of the remote sensor to be evaluated and compared with the actual ozone content measured on the ground. If the ozone content is consistent, the remote sensor has a high calibration accuracy. Specifically, different satellites measure different atmospheric parameters.

[0155] In other words, the calibration accuracy verification steps are: using the cross-calibration method to obtain the cross-calibration coefficient of the remote sensor, and verifying the on-board calibration coefficient; in addition, by comparing the inverted atmospheric parameters with the actual ground-based measurement results, the on-board radiation calibration accuracy of the remote sensor is further evaluated; in other words, the present invention uses a calibration formula to calculate the on-board radiation calibration coefficient.

[0156] Furthermore, the calibration time described in step 1 is determined based on the DN value of the sunlight calibration data to avoid the impact of time simulation deviation. In other words, the DN' value of the remote sensor radiation output is extracted according to the onboard calibration file, and the cold air count value is removed. The time when sunlight enters the sunlight cone is determined based on the changing DN value.

[0157] The satellite 3D mechanical model in step 2 has the same profile of the satellite platform and optical payload as the actual product state of the satellite. The model is imported into the TracePro simulation software, and the reflective characteristics of the model surface are defined based on the actual ground measurements of the satellite surface characteristics.

[0158] In step 3, the long-term onboard calibration data of the remote sensor is organized into a two-week cycle. The data within a cycle is called a data unit. After obtaining data from multiple solar reflection channels of the remote sensing instrument, the bidirectional transmission factor of the solar cone is expanded to the entire solar reflection band through polynomial interpolation to construct an angle- and wavelength-dependent BTDF statistical model. Multiple BTDF models constructed from a large number of data units over the remote sensor's life cycle enable dynamic simulation of the radiation performance of the solar cone.

[0159] In other words, this method allows for uniform radiometric processing of long-term sensor onboard calibration data, yielding long-term, high-frequency, and uniform radiometric calibration coefficients. Calibration accuracy is assessed using a cross-calibration method, and the difference between the inverted parameters and ground-based measurements is used to further evaluate onboard radiometric calibration accuracy.

[0160] When selecting cross-calibration image pairs in step 5, clear, cloud-free image pairs with the same observation time and geometry as much as possible should be selected. The ground-based atmospheric measurement parameters in the constructed on-board calibration basic database should be the atmospheric parameters that can be inverted from the remote sensing data products for comparison;

[0161] This embodiment is further described below with reference to the accompanying drawings. This embodiment meets the requirements for absolute radiation calibration accuracy on remote sensor satellites.

[0162] See also Figure 1 , see the following description for details.

[0163] Due to the long-term effects of the harsh environment of outer space, the radiation performance of onboard calibration systems has gradually degraded, making it difficult to trace the onboard radiometric calibration benchmark. Currently, onboard calibration data is only used to track changes in remote sensor response for research purposes, making it difficult to achieve absolute onboard radiometric calibration of remote sensors.

[0164] This paper proposes combining the simulation software STK and TracePro to simulate and calculate the solar irradiance E incident on the solar cone as a satellite passes over the South Pole, i.e., during solar calibration. Furthermore, the radiance L is calculated by establishing a BTDF model. This radiance L represents the absolute radiation value, and combined with the radiation calibration formula, the onboard absolute radiation calibration coefficient can be calculated. By performing unified radiation calibration on long-term series onboard calibration data from remote sensors, the present invention obtains long-term, high-frequency, and uniformly accurate onboard radiation calibration coefficients.

[0165] In view of the above method, an on-board calibration data processing system based on a solar cone on-board calibrator is proposed, which includes the following processing modules:

[0166] Data preprocessing module: Extract the DN' value at the time of sunlight calibration, remove the cold air count value DN0, and determine the time when sunlight enters the sunlight cone based on the changing DN value; organize the long-term on-board calibration data into a cycle of 2 weeks, and define the data within one cycle as a data unit;

[0167] In other words, the data preprocessing module includes extracting the DN' value of remote sensing data products, eliminating data interfered by non-calibrated light sources, sorting long-term on-board calibration products into a two-week cycle, and dividing a large amount of on-board calibration data into several data intervals so that BTDF modeling can be performed for each interval.

[0168] Simulation module: STK simulation software is used to calculate the incident angle of sunlight incident on the sunlight cone when the satellite passes over the South Pole. STK simulation simulates the actual operation status of the satellite in orbit based on the satellite orbit report, and TracePro simulation software is used to simulate the incident irradiance at the corresponding angle;

[0169] In other words, the simulation module uses STK simulation software to calculate the angle of incidence of sunlight incident on the sunlight cone when the satellite passes over the South Pole, and uses TracePro simulation software to simulate the incident irradiance at the corresponding angle. STK simulation uses satellite orbit reports to simulate the actual operating status of the satellite in orbit. The satellite model is consistent with the actual product status, and the surface reflectivity characteristics of the satellite model are established based on the actual measurement results of the satellite surface material.

[0170] BTDF modeling module: For each data interval, the corresponding relationship between the solar zenith angle, azimuth angle, incident irradiance E and the remote sensor output DN value is analyzed to construct the BTDF model of the sunlight cone; different incident angles are interpolated to obtain a multi-angle BTDF statistical model.

[0171] In other words, the BTDF modeling module: for each data interval, analyzes the sunlight incident angle, including the solar zenith angle and azimuth angle, the zenith angle and azimuth angle of the sunlight cone transmission light, the incident irradiance E(θ i ,φ i ) and the corresponding relationship between the remote sensor output DN value, through DN / E(θ i ,φ i ) and L / E(θ i ,φ i ) to construct the BTDF model of the sunlight cone; and interpolate different incident angles to obtain a multi-angle BTDF statistical model.

[0172] Calibration coefficient calculation module: According to the BTDF model, the calibration coefficient is calculated by combining the DN value and irradiance at the time of sunlight calibration;

[0173] In other words, the calibration coefficient calculation module calculates the on-board calibration coefficient according to the BTDF model and the DN value at that moment using the following calibration formula:

[0174]

[0175] Among them, θ i and φ i are the zenith angle and azimuth angle of the incident light of the sunlight cone; θ t and φ t are the zenith angle and azimuth angle of the transmitted light of the sunlight cone, L(θ i ,φ i ,θ t ,φ t ) is the solar radiance received by the remote sensor, and A is the calibration coefficient.

[0176] Calibration Accuracy Verification Module: Using a high-precision remote sensor as a reference, the sensor is cross-calibrated to obtain the cross-calibration coefficient. By comparing the inverted atmospheric parameters, such as temperature and water vapor, with ground-based measurement results, the onboard radiometric calibration accuracy of the remote sensor is further verified.

[0177] In other words, the calibration accuracy verification module cross-calibrates the sensor using a high-precision reference, obtains the cross-calibration coefficient, and evaluates the onboard calibration accuracy. After onboard absolute radiometric calibration, atmospheric parameters such as temperature and water vapor are inverted from the sensor's remote sensing data products. These parameters are then compared with ground-based measurements to further evaluate the sensor's onboard radiometric calibration accuracy.

[0178] The present invention also provides an on-board calibration data processing system based on a solar light cone on-board calibrator. The on-board calibration data processing system based on a solar light cone on-board calibrator can be realized by executing the process steps of the on-board calibration data processing method based on the solar light cone on-board calibrator, that is, those skilled in the art can understand the on-board calibration data processing method based on the solar light cone on-board calibrator as a preferred implementation of the on-board calibration data processing system based on the solar light cone on-board calibrator.

[0179] According to the present invention, an on-board calibration data processing system based on a solar cone on-board calibrator is provided, comprising:

[0180] Data preprocessing module: collects data and builds on-board calibration database;

[0181] Simulation module: Based on the onboard calibration database, it simulates the incident angle of sunlight when it enters the sunlight cone, and then calculates and obtains the incident irradiance of the remote sensor;

[0182] BTDF modeling module: Based on the irradiance, a BTDF model of the sunlight cone is constructed, and angle interpolation is performed to obtain a multi-angle BTDF model;

[0183] Calibration coefficient calculation module: calculates and obtains the calibration coefficient of the remote sensor according to the BTDF model and the radiation output DN value and irradiance of the remote sensor;

[0184] Calibration accuracy verification module: based on the cross calibration coefficient, evaluates the on-board radiation calibration accuracy.

[0185] Those skilled in the art will appreciate that, in addition to implementing the system and its various devices, modules, and units provided by the present invention in purely computer-readable program code, it is entirely possible to implement the same functions of the system and its various devices, modules, and units provided by the present invention in the form of logic gates, switches, application-specific integrated circuits, programmable logic controllers, and embedded microcontrollers by logically programming the method steps. Therefore, the system and its various devices, modules, and units provided by the present invention can be considered a hardware component, and the devices, modules, and units included therein for implementing various functions can also be considered as structures within the hardware component; the devices, modules, and units for implementing various functions can also be considered as both software modules implementing the method and structures within the hardware component.

[0186] The above describes specific embodiments of the present invention. It should be understood that the present invention is not limited to the specific embodiments described above, and those skilled in the art may make various changes or modifications within the scope of the claims, which do not affect the essence of the present invention. The embodiments of this application and the features in the embodiments may be combined with each other in any manner unless there is a conflict.

Claims

1. A method for processing on-board calibration data based on a solar cone on-board calibrator, characterized in that: include: Step S1: Collect data and build an on-board calibration database; Step S2: Based on the onboard calibration database, simulate the incident angle of sunlight incident on the sunlight cone, and then calculate and obtain the incident irradiance of the sunlight cone of the remote sensor; Step S3: constructing a BTDF model of the sunlight cone based on the irradiance, and performing angle interpolation to obtain a multi-angle BTDF model; Step S4: Calculate and obtain the calibration coefficient of the remote sensor based on the BTDF model and the radiation output DN value and irradiance of the remote sensor, and then evaluate the on-board radiation calibration accuracy.

2. The on-board calibration data processing method based on the sunlight cone on-board calibrator according to claim 1 is characterized in that: In step S1, the data includes: atmospheric parameters, L1-level image data, on-board calibration data and ground-based observation data; the on-board calibration data includes: cold air calibration data and sunlight calibration data; In step S1, the sunlight calibration data of the onboard calibration data is extracted, that is, the sunlight calibration radiation output count value, referred to as DN'; the cold air count value of DN' is removed to obtain the change count value caused by sunlight; The mathematical expression of the change count value caused by sunlight is: DN=DN'-DN0 Wherein, DN represents the change count value caused by sunlight, DN' represents the sunlight calibration radiation output count value; DN0 represents the background noise.

3. The on-board calibration data processing method based on the sunlight cone on-board calibrator according to claim 2, characterized in that: In the step S2, it includes: Step S2.1: Input the satellite orbit, run the STK software to simulate the satellite's on-orbit operation state, and obtain the simulation results; Step S2.2: Based on the simulation results, query the incident angle of sunlight; the incident angle includes the solar zenith angle and the solar azimuth angle; Step S2.3: Run TracePro software to calculate and obtain the irradiance at the entrance pupil of the remote sensor based on the incident angle; In step S3, the mathematical expression of the multi-angle BTDF model is: Which is then simplified to: Among them, BTDF(θ i ,φ i ,θ t ,φ t ) represents the multi-angle BTDF model, θ t With φ t They represent the zenith angle and azimuth angle of the transmitted light of the sun cone, L(θ i ,φ i ,θ t ,φ t ) is the solar radiance received by the remote sensor; A is the calibration coefficient; E(θ i ,φ i ) represents irradiance.

4. The on-board calibration data processing method based on the sunlight cone on-board calibrator according to claim 3 is characterized in that: In step S4, the on-board radiation calibration accuracy is evaluated by cross-calibration. The cross calibration process includes: Step A1: extract the image DN value when the remote sensor passes through the target site; Step A2: Calculate and obtain the normalized apparent reflectance based on the image DN value and the calibration coefficient; Step A3: Correcting the normalized apparent reflectivity by the solar zenith angle and the sun-earth distance factor to obtain the top of atmosphere reflectivity; Step A4: Based on the top of atmosphere reflectivity, the surface reflectivity in the remote sensor observation direction is calculated; the surface reflectivity to be evaluated is obtained by multi-angle BRDF model correction, and then the top of atmosphere reflectivity to be evaluated is calculated; Step A5: Calculating and obtaining a cross-calibration coefficient based on the top-of-atmosphere reflectivity to be evaluated, the solar zenith angle, the sun-earth distance factor, and the average DN value of the target area image; Step A6: determining whether the cross calibration coefficient is consistent with the onboard calibration coefficient; if yes, no processing is performed; if no, correcting the calibration coefficient of the remote sensor; Step A7: Determine whether the atmospheric parameters extracted from the data product obtained based on the on-board calibration coefficients are consistent with the actually measured atmospheric parameters. If the result is yes, no processing is performed; if the result is no, the calibration coefficients of the remote sensor are corrected.

5. The on-board calibration data processing method based on the sunlight cone on-board calibrator according to claim 4 is characterized in that: In step A5, the calculation expression of the cross calibration coefficient is: Among them, ρ * is the top of atmosphere reflectivity, θ s is the solar zenith angle, A is the correction factor for the average and actual distance between the sun and the earth; ρ is the cross calibration coefficient; In step A7, it is determined whether the ozone product obtained as a data product based on the cross-calibration coefficient is consistent with the ozone content actually measured on the ground. If the result is yes, no processing is performed; if the result is no, the calibration coefficient of the remote sensor is corrected.

6. An on-board calibration data processing system based on a solar cone on-board calibrator, characterized in that: include: Data preprocessing module: collects data and builds on-board calibration database; Simulation module: Based on the onboard calibration database, it simulates the incident angle of sunlight when it enters the sunlight cone, and then calculates and obtains the irradiance of the remote sensor; BTDF modeling module: Based on the irradiance, a BTDF model of the sunlight cone is constructed, and angle interpolation is performed to obtain a multi-angle BTDF model; Calibration coefficient calculation module: calculates and obtains the calibration coefficient of the remote sensor according to the BTDF model and the radiation output DN value and irradiance of the remote sensor; Calibration accuracy verification module: based on the calibration coefficient, evaluates the on-board radiation calibration accuracy.

7. The on-board calibration data processing system based on the sunlight cone on-board calibrator according to claim 6, characterized in that: In the data preprocessing module, the data includes: atmospheric parameters, L1-level image data, on-board calibration data and ground-based observation data; the on-board calibration data includes: cold air calibration data and sunlight calibration data; In the data preprocessing module, the sunlight calibration data of the onboard calibration data is extracted, that is, the sunlight calibration radiation output count value, referred to as DN'; the cold air count value of DN' is removed to obtain the change count value caused by sunlight; The mathematical expression of the change count value caused by sunlight is: DN=DN'-DN0 Wherein, DN represents the change count value caused by sunlight, DN' represents the sunlight calibration radiation output count value; DN0 represents the background noise.

8. The on-board calibration data processing system based on the sunlight cone on-board calibrator according to claim 7, characterized in that: The simulation module includes: Simulation submodule 1: Input the satellite orbit, run the STK software to simulate the satellite's on-orbit operation status, and obtain the simulation results; Simulation submodule 2: Based on the simulation results, query the incident angle of sunlight; the incident angle includes the solar zenith angle and the solar azimuth angle; Simulation submodule three: running TracePro software to calculate and obtain the irradiance at the entrance pupil of the remote sensor based on the incident angle; In the BTDF modeling module, the mathematical expression of the multi-angle BTDF model is: Which is then simplified to: Among them, BTDF(θ i ,φ i ,θ t ,φ t ) represents the multi-angle BTDF model, θ t With φ t They represent the zenith angle and azimuth angle of the transmitted light of the sun cone, L(θ i ,φ i ,θ t ,φ t ) is the solar radiance received by the remote sensor; A is the calibration coefficient; E(θ i ,φ i ) represents irradiance.

9. The on-board calibration data processing system based on the sunlight cone on-board calibrator according to claim 8, characterized in that: In the calibration coefficient calculation module, the on-board radiation calibration accuracy is evaluated by cross-calibration. The modules involved in the cross-calibration include: Module A1: Extract the DN value of the image when the remote sensor passes through the target site; Module A2: Calculate and obtain the normalized apparent reflectance based on the image DN value and calibration coefficient; Module A3: Correcting the normalized apparent reflectivity by the solar zenith angle and the distance factor between the sun and the earth to obtain the reflectivity of the top of the atmosphere; Module A4: Based on the top of atmosphere reflectivity, the surface reflectivity in the remote sensor observation direction is calculated; the surface reflectivity to be evaluated is obtained by multi-angle BRDF model correction, and then the top of atmosphere reflectivity to be evaluated is calculated; Module A5: Calculating and obtaining a cross-calibration coefficient based on the top-of-atmosphere reflectivity to be evaluated, the solar zenith angle, the sun-earth distance factor, and the average DN value of the target area image; Module A6: judging whether the cross calibration coefficient is consistent with the on-board calibration coefficient, if the result is yes, no processing is performed; if the result is no, correcting the calibration coefficient of the remote sensor; In the calibration accuracy verification module, it is determined whether the atmospheric parameters extracted from the data product obtained based on the cross calibration coefficient are consistent with the actually measured atmospheric parameters. If the result is yes, no processing is performed; if the result is no, the calibration coefficient of the remote sensor is corrected.

10. The on-board calibration data processing system based on the sunlight cone on-board calibrator according to claim 9, characterized in that: In the module A5, the calculation expression of the cross calibration coefficient is: Among them, ρ * is the top of atmosphere reflectivity, θ s is the solar zenith angle, A is the correction factor for the average and actual distance between the sun and the earth; ρ is the cross calibration coefficient; In the calibration accuracy verification module, it is determined whether the ozone product obtained as a data product based on the cross-calibration coefficient is consistent with the ozone content actually measured on the ground. If the result is yes, no processing is performed; if the result is no, the calibration coefficient of the remote sensor is corrected.

Images