Infrared payload cross-calibration method and related devices

Abstract

The application relates to the technical field of thermal infrared remote sensing, and in particular to an infrared load cross-calibration method and related equipment. In the background of lacking effective calibration references at the low-temperature end of thermal infrared, a thermal infrared cross-calibration technical scheme considering the low-temperature end and the normal-temperature end is established. The DCC is selected to provide calibration references in the low-temperature interval, the sea is selected to provide calibration references in the normal-temperature interval, a cross-calibration system considering the low-temperature end and the normal-temperature end is constructed, and the applicability of calibration coefficients in different temperature intervals is improved.

Description

Technical Field

[0001] This application relates to the field of thermal infrared remote sensing technology, and in particular to an infrared load cross-calibration method and related equipment. Background Technology

[0002] Radiometric calibration is a prerequisite for the quantitative application of remote sensing data. It is the process of determining the quantitative relationship between the payload's output signal and the incident radiation. A continuous link connects the remote sensing payload to a reference radiation source, ensuring the quality of data and products throughout the payload's lifecycle. Existing thermal infrared payloads generally employ schemes such as "on-board blackbody + deep space observation" for on-orbit calibration. For example, the MODIS Thermal Emission Band (TEB) uses an on-board blackbody as the primary calibration source, with an operating temperature range of 270-315K and a temperature uniformity of 0.03-0.08K. However, satellite observations include numerous low-temperature scenarios below the on-board blackbody's operating temperature. Due to insufficient coverage of the blackbody's temperature range, the uncertainty of the calibration offset term increases significantly as the target temperature decreases.

[0003] Currently, on-board calibration of infrared payloads mainly adopts the "blackbody + cold space" scheme. Cold space observations can provide a stable low-temperature reference of approximately 2.73K, but this temperature is far below the common low-temperature target range of most infrared detectors, leading to significant nonlinear errors and making it difficult to verify the calibration accuracy in common low-temperature ranges. Some studies have proposed an on-orbit calibration scheme of "room-temperature blackbody + low-temperature blackbody," in which the low-temperature blackbody provides a traceable temperature reference covering common low-temperature ranges, compensating for the excessively low temperature of cold space observations and reducing nonlinear errors. However, this increases payload mass, power consumption, and system complexity, and its engineering application and performance optimization are still in the research exploration stage. Summary of the Invention

[0004] This application provides an infrared load cross-calibration method and related equipment, which uses DCC as a low-temperature reference target and the ocean as a normal-temperature reference target for cross-calibration. This can effectively make up for the lack of low-temperature reference targets in thermal infrared cross-calibration and reduce the nonlinear error caused by extrapolation of calibration relationships in the low-temperature region.

[0005] In a first aspect, an infrared payload cross-calibration method is provided, the method comprising: acquiring first thermal infrared data observed by the payload to be calibrated, and acquiring second thermal infrared data observed by a reference payload; filtering deep convection cloud DCC pixels from the first thermal infrared data and the second thermal infrared data respectively, obtaining a first DCC pixel set and a second DCC pixel set respectively; filtering ocean pixels from the first thermal infrared data and the second thermal infrared data respectively, obtaining a first ocean pixel set and a second ocean pixel set respectively; determining the atmospheric parameters corresponding to the first DCC pixel set, the second DCC pixel set, the first ocean pixel set, and the second ocean pixel set, and determining the DCC emissivity spectrum and Ocean emissivity spectrum; based on the DCC emissivity spectrum, the ocean emissivity spectrum, the atmospheric parameters, the spectral response function of the payload to be calibrated, and the spectral response function of the reference payload, determine the spectral matching factor between the payload to be calibrated and the reference payload; according to the spectral matching factor, correct the DCC observed radiance and the ocean observed radiance of the reference payload respectively to obtain the DCC equivalent reference radiance and the ocean equivalent reference radiance; based on the DCC equivalent reference radiance, the ocean equivalent reference radiance, the DN value of the DCC pixels in the first DCC pixel set, and the DN value of the ocean pixels in the first ocean pixel set, determine the calibration coefficient of the payload to be calibrated.

[0006] In one embodiment, the step of filtering deep convection cloud DCC pixels from the first thermal infrared data and the second thermal infrared data to obtain a first DCC pixel set and a second DCC pixel set respectively includes: filtering DCC pixels from the first thermal infrared data and the second thermal infrared data based on brightness temperature thresholds and spatial uniformity; wherein, the brightness temperature thresholds include: brightness temperature of the 11μm channel in thermal infrared less than 205K; the spatial uniformity includes: standard deviation of brightness temperature of the 3×3 pixel thermal infrared channel less than 1K, and standard deviation of reflectance of the visible light channel less than 3%.

[0007] In one embodiment, DCC pixels in the first DCC pixel set and DCC pixels in the second DCC pixel set are matched, and ocean pixels in the first ocean pixel set and ocean pixels in the second ocean pixel set are matched; wherein the matched pixels correspond to the same observation area, and the time difference between the observation times of the matched pixels is less than or equal to 10 minutes.

[0008] In one example of this embodiment, the observed zenith angles of the DCC pixels in the first DCC pixel set and the DCC pixels in the second DCC pixel set are both less than 40°; the first DCC pixel in the first DCC pixel set and the second DCC pixel in the second DCC pixel set are matched; wherein, the relationship between the observed zenith angles of the first DCC pixel and the observed zenith angles of the second DCC pixel satisfies ;in, The observed zenith angle representing the first DCC pixel, This represents the observed zenith angle of the second DCC pixel.

[0009] In one embodiment, determining the atmospheric parameters corresponding to the first DCC pixel set, the second DCC pixel set, the first ocean pixel set, and the second ocean pixel set includes: obtaining corresponding atmospheric profile information based on the observation time and geographical location of each pixel in the first DCC pixel set, the second DCC pixel set, the first ocean pixel set, and the second ocean pixel set; and calculating the atmospheric parameters of each pixel using a radiative transfer model based on the atmospheric profile information; wherein the atmospheric parameters include atmospheric transmittance, upward atmospheric radiation, and downward atmospheric radiation.

[0010] In one embodiment, determining the DCC emissivity spectrum includes: acquiring hyperspectral thermal infrared data corresponding to the DCC pixels in the first DCC pixel set and the DCC pixels in the second DCC pixel set based on their observation time and geographical location; performing atmospheric correction on the hyperspectral thermal infrared data based on the atmospheric parameters of the DCC pixels; and retrieving the DCC emissivity spectrum using a temperature-emissivity separation method based on the atmospherically corrected hyperspectral thermal infrared data.

[0011] In one embodiment, determining the spectral matching factor between the payload to be calibrated and the reference payload includes: simulating the DCC atmospheric top radiance and the ocean atmospheric top radiance based on the DCC emissivity spectrum, the ocean emissivity spectrum, and the atmospheric parameters; convolving the DCC atmospheric top radiance and the ocean atmospheric top radiance with the spectral response function of the payload to be calibrated and the spectral response function of the reference payload, respectively, to obtain the corresponding simulated channel radiance; and determining the spectral matching factor based on the simulated channel radiance.

[0012] Secondly, an infrared payload cross-calibration device is provided, comprising: an acquisition unit for acquiring first thermal infrared data observed by the payload to be calibrated and second thermal infrared data observed by a reference payload; a first filtering unit for filtering deep convection cloud DCC pixels from the first and second thermal infrared data respectively, obtaining a first DCC pixel set and a second DCC pixel set; a second filtering unit for filtering ocean pixels from the first and second thermal infrared data respectively, obtaining a first ocean pixel set and a second ocean pixel set; and a first determining unit for determining atmospheric parameters corresponding to the first DCC pixel set, the second DCC pixel set, the first ocean pixel set, and the second ocean pixel set, and determining DCC emission. The system comprises: a first determination unit, which determines the spectral matching factor between the DCC emissivity spectrum and the ocean emissivity spectrum; a second determination unit, which determines the spectral matching factor between the DCC emissivity spectrum, the ocean emissivity spectrum, the atmospheric parameters, the spectral response function of the payload to be calibrated, and the spectral response function of the reference payload; a correction unit, which corrects the DCC observation radiance and the ocean observation radiance of the reference payload according to the spectral matching factor, respectively, to obtain the DCC equivalent reference radiance and the ocean equivalent reference radiance; and a third determination unit, which determines the calibration coefficient of the payload to be calibrated according to the DCC equivalent reference radiance, the ocean equivalent reference radiance, the DN value of the DCC pixels in the first DCC pixel set, and the DN value of the ocean pixels in the first ocean pixel set.

[0013] Thirdly, a computing device is provided, including a memory and a processor; the memory stores computing instructions; the processor executes the computing instructions, causing the computing device to perform the method provided in the first aspect.

[0014] Fourthly, a computer-readable storage medium is provided, including computer program instructions that, when executed by a computing device, perform the method provided in the first aspect.

[0015] Fifthly, a computer program product is provided, including a computer program that, when executed by a processor, implements the method provided in the first aspect.

[0016] The infrared payload cross-calibration method and related equipment provided in this application construct a cross-calibration system that takes into account both low-temperature and normal-temperature ends. By using DCC as a low-temperature reference target and the ocean as a normal-temperature reference target in combination for cross-calibration, it can effectively compensate for the lack of low-temperature reference targets in thermal infrared cross-calibration and reduce the nonlinear errors caused by extrapolation of calibration relationships in the low-temperature region. Compared with schemes that rely on only a single type of target, this application can improve the reliability and applicability of calibration results under different scenarios. Attached Figure Description

[0017] Figure 1 A flowchart of an infrared load cross-calibration method provided in this application;

[0018] Figure 2 A flowchart for DCC emissivity spectrum inversion is provided in this application;

[0019] Figure 3 A schematic diagram of an infrared load cross-calibration device provided in this application;

[0020] Figure 4 A schematic diagram of a computing device provided in this application. Detailed Implementation

[0021] The solution provided in this application will now be described with reference to the accompanying drawings. In this application, "multiple" refers to two or more objects, and "various kinds" refers to two or more types. Terms such as "first," "second," etc., are used only to distinguish similar objects and are not necessarily used to describe a specific order or number of objects.

[0022] To facilitate understanding of the solutions provided in this application, the technical terms that may be involved in this application will be introduced first.

[0023] Existing research has shown that deep convective clouds (DCCs) are characterized by low brightness temperatures, widespread distribution in tropical regions, and relative short-term stability. They have been widely used in radiometric calibration and long-term attenuation assessment of visible light channels. Related studies often use radiometric reference loads such as MODIS as a reference, constructing DCC samples and conducting cross-load comparisons to monitor load stability and correct calibration coefficients. Visible light calibration studies typically need to address the differences in spectral response functions (SRFs) between different load channels. This is often achieved by combining hyperspectral observations or radiative transfer simulation data with the SRF convolution of two loads to calculate the spectral band adjustment factor (SBAF) to correct for spectral differences between channels.

[0024] Existing research has proposed a thermal infrared band assessment method based on quasi-deep convective clouds (qDCC). This scheme selects nighttime data, filters qDCC pixels in tropical regions, identifies target pixels using an 11μm brightness temperature threshold (MODIS band 31 brightness temperature less than 205K) as the criterion, removes mixed pixels at cloud edges based on uniformity (standard deviation of infrared brightness temperature of 3×3 neighboring pixels less than 1K), normalizes the samples to reduce the impact of DCC distribution changes over time, and then conducts stability assessments of the MODIS thermal infrared emission band based on the samples.

[0025] With the increasing number of multi-source remote sensing payloads, relying solely on single-payload stability assessments is insufficient to meet operational needs for cross-satellite data consistency maintenance, long-term attenuation monitoring, and historical data recalibration. This application draws upon the application of DCC in visible light cross-calibration to establish a cross-calibration method for thermal infrared payloads. First, a sample library is constructed by selecting DCC and ocean pixels from the payload to be calibrated and the reference payload data. The observed radiance values ​​of the two payloads for corresponding pixels are extracted. Then, the reference payload is spectrally matched using DCC emissivity spectra retrieved from thermal infrared hyperspectral data and the temperature-emissivity separation algorithm, ocean emissivity spectra extracted from the ASTER spectral library, and atmospheric parameters calculated using the radiative transfer model. Finally, the calibration coefficients of the payloads are calculated by combining all samples, providing support for long-term attenuation monitoring and historical data recalibration of infrared remote sensing payloads. This application selects DCC and ocean as calibration targets. Compared to calibration methods relying solely on a single target type, this application can accommodate both low-temperature and normal-temperature targets, reducing temperature bias in the calibration results and improving the applicability of the calibration coefficients.

[0026] During cross-calibration, differences in the SRF of different payload channels can lead to systematic errors. Therefore, it is necessary to calculate the SBAF (Spotlight-Based Emissivity Spectrum) to perform spectral matching between the two payloads before calibration. Calculating the SBAF requires the emissivity spectrum of the calibration target and a series of atmospheric parameters (including atmospheric transmittance, upward atmospheric radiation, and downward atmospheric radiation). Atmospheric parameters can be calculated using the Moderate Resolution Atmospheric Transmission Model (MODTRAN) combined with atmospheric profile data. The emissivity spectra of most ground features (including oceans) can be obtained from spectral libraries measured in laboratories, while DCCs (Digital Carrier Cells) located at high altitudes are difficult to measure directly, making their emissivity an unknown quantity in the calculation. The technical problems that this application can solve include: in infrared payload cross-calibration, how to construct calibration samples that take into account both low-temperature and normal-temperature ends, and how to complete spectral matching under the condition that DCC emissivity cannot be directly obtained, thereby achieving on-orbit cross-calibration of infrared payloads.

[0027] To address the issues of insufficient low-temperature reference and difficulty in directly obtaining DCC emissivity in thermal infrared payload cross-calibration, this application provides an infrared payload cross-calibration method. This method is based on deep convection clouds and marine targets for infrared payload cross-calibration. The method may include... Figure 1 The steps are shown.

[0028] Step 101: Obtain the first thermal infrared data observed by the load to be calibrated, and obtain the second thermal infrared data observed by the reference load.

[0029] The payload to be calibrated is located on one satellite, and the reference payload is located on another satellite.

[0030] In step 101, thermal infrared data of the load to be calibrated and the reference load are acquired, and data that match in time and space (observation time difference not exceeding 10 minutes and observation area being the same) are selected. That is, the first thermal infrared data and the second thermal infrared data are observed by the load to be calibrated and the reference load, respectively, for the same observation area, and the time difference between the observation time of the first thermal infrared data and the observation time of the second thermal infrared data is less than or equal to 10 minutes.

[0031] Step 102: Select deep convection cloud DCC pixels from the first thermal infrared data and the second thermal infrared data respectively to obtain the first DCC pixel set and the second DCC pixel set respectively.

[0032] In this case, the DCC pixels in the first DCC pixel set and the DCC pixels in the second DCC pixel set are matched one-to-one.

[0033] Step 103: Select ocean pixels from the first thermal infrared data and the second thermal infrared data respectively to obtain a first ocean pixel set and a second ocean pixel set. The ocean pixels in the first ocean pixel set and the ocean pixels in the second ocean pixel set are matched one-to-one.

[0034] In steps 102 and 103, DCC pixels and marine pixels are screened from the data (i.e., the first thermal infrared data and the second thermal infrared data) of the two payloads (i.e., the payload to be calibrated and the reference payload), respectively, to construct a cross-calibration sample set of low-temperature and normal-temperature ends, and obtain the DN value and observed radiance of each sample. DCC pixels are extracted based on the brightness temperature threshold (brightness temperature of the 11μm channel of thermal infrared less than 205K) and spatial uniformity (standard deviation of brightness temperature of 3×3 pixel thermal infrared channel less than 1K, standard deviation of reflectance of visible light channel less than 3%), and the mean and standard deviation of brightness temperature of pixels are statistically analyzed. Non-uniform samples with brightness temperatures higher than the mean plus 3 times the standard deviation are removed to obtain the DCC sample set; marine pixels can be obtained by joint screening based on the geolocation product and cloud product of the reference payload. When using MODIS as a reference payload, the latitude and longitude information and land surface type identifiers in the MOD03 geolocation product can be used to filter pixels whose underlying surface is ocean. The standard deviation of the brightness temperature in the 3×3 neighborhood of the 11μm thermal infrared channel must be less than 1K. Cloud mask information from MOD06 is used to avoid cloud edge interference. Finally, non-uniform samples with brightness temperatures higher than the mean plus three times the standard deviation are removed to obtain the ocean sample set. The DCC of the corresponding locations of the two payloads and the observed zenith angle of the ocean samples are compared. Samples with a zenith angle less than 40° and satisfying the following formula are used as the final matching samples:

[0035]

[0036] Specifically, DCC pixels selected from the first thermal infrared data form the first DCC pixel set, and DCC pixels selected from the second thermal infrared data form the second DCC pixel set. Similarly, ocean pixels selected from the first thermal infrared data form the first ocean pixel set, and ocean pixels selected from the second thermal infrared data form the second ocean pixel set.

[0037] The observed zenith angle of the DCC pixels in the first DCC pixel set and the DCC pixels in the second DCC pixel set is less than 40°.

[0038] The first DCC pixel in the first DCC pixel set and the second DCC pixel in the second DCC pixel set are matched; wherein, the relationship between the observed zenith angle of the first DCC pixel and the observed zenith angle of the second DCC pixel satisfies ; In this paragraph The observed zenith angle representing the first DCC pixel, in this paragraph The observed zenith angle represents the second DCC cell. The first DCC cell is any DCC cell in the first DCC cell set.

[0039] The first ocean pixel in the first ocean pixel set and the second ocean pixel in the second ocean pixel set are matched; wherein, the relationship between the observed zenith angle of the first ocean pixel and the observed zenith angle of the second ocean pixel satisfies ; In this paragraph The observed zenith angle representing the first ocean pixel, in this paragraph The zenith angle represents the observation angle of the second ocean pixel.

[0040] Step 104: Determine the atmospheric parameters corresponding to the first DCC pixel set, the second DCC pixel set, the first ocean pixel set, and the second ocean pixel set, and determine the DCC emissivity spectrum and the ocean emissivity spectrum.

[0041] In step 104, the atmospheric parameters of each DCC cell in the first DCC cell set are calculated, the atmospheric parameters of each DCC cell in the second DCC cell set are calculated, the atmospheric parameters of each ocean cell in the first ocean cell set are calculated, and the atmospheric parameters of each ocean cell in the second ocean cell set are calculated.

[0042] In step 104, atmospheric parameters for each sample are calculated using the MODTRAN radiative transfer model and reanalysis atmospheric profile data (such as ERA5 data from the European Centre for Medium-Range Weather Forecasts (ECMWF)). For each DCC and ocean sample pixel within the observation area, corresponding atmospheric profile information is obtained based on its observation time and geographical location, and atmospheric transmittance in the thermal infrared target band is calculated using the MODTRAN radiative transfer model. Atmospheric uplift radiation and atmospheric downward radiation Here, "sample" refers to a pixel, such as a DCC pixel or an ocean pixel.

[0043] In other words, calculating the atmospheric parameters of each DCC pixel in the first DCC pixel set includes: obtaining the atmospheric profile information corresponding to the first DCC pixel based on the observation time and geographical location of the first DCC pixel in the first DCC pixel set; and calculating the atmospheric parameters of the first DCC pixel using the MODTRAN radiative transfer model based on the atmospheric profile information corresponding to the first DCC pixel; wherein the atmospheric parameters of the first DCC pixel include atmospheric transmittance, atmospheric upward radiation, and atmospheric downward radiation.

[0044] In step 104, the emissivity spectrum of each DCC pixel in the first DCC pixel set can be retrieved using the temperature-emissivity separation method, and the emissivity spectrum of each DCC pixel in the second DCC pixel set can be retrieved using the temperature-emissivity separation method.

[0045] According to Planck's law of blackbody radiation, the radiance of a blackbody can be expressed as:

[0046]

[0047] In the formula, C1 and C2 are Planck's constants, C1 = 1.191 × 10⁻⁶. 8 W / (m 2 ∙sr∙μm -4 C2 = 1.439 × 10 4 μm∙K, T is temperature in Kelvin (K), and λ is wavelength in micrometers (μm).

[0048] The emissivity of an object is defined as the ratio of its radiance to that of a blackbody at the same temperature. The emissivity spectrum of a DCC can be expressed as:

[0049]

[0050] In the formula, The radiance of DCC, The blackbody radiation brightness. For DCC at wavelength Emissivity at that location

[0051] Since the emissivity spectrum of DCC cannot be directly obtained from laboratory spectral libraries like that of general surface materials, it needs to be inverted. Therefore, this application uses hyperspectral thermal infrared data and employs a temperature-emissivity separation method to invert the emissivity spectrum of DCC. .

[0052] The operations for inverting the DCC emissivity spectrum include, for example: Figure 2 Operations 1041-1043 shown

[0053] In operation 1041, hyperspectral thermal infrared data corresponding to the DCC samples are acquired based on the observation time and geographical location, such as data from the Atmospheric Infrared Detector (AIRS) and the Infrared Atmospheric Sounding Interferometer (IASI), to obtain the observed radiance of DCC pixels in the hyperspectral thermal infrared data. .

[0054] In operation 1042, atmospheric parameters corresponding to DCC pixels in the hyperspectral thermal infrared data were calculated using the MODTRAN radiative transfer model and reanalysis atmospheric profile data. Atmospheric correction was then applied to obtain the cloud top upward radiation.

[0055]

[0056] In the formula, To observe radiance, This is upward atmospheric radiation. Atmospheric transmittance, The upward radiation from the cloud top is after atmospheric correction.

[0057] In operation 1043, the emissivity spectrum of each DCC sample is retrieved using the temperature-emissivity separation method. The emissivity spectrum of the samples is averaged and smoothed according to the spectral continuity characteristics to reduce the influence of inversion noise and local outliers, finally obtaining a representative DCC emissivity spectrum. The specific algorithm is as follows:

[0058] Assumption =0.99, estimated Genting self-radiation :

[0059]

[0060] For each channel within the target band, the brightness temperature can be calculated using the inverse Planck function. :

[0061]

[0062] In the formula, and The Planck constant term is used, and the maximum brightness temperature within the target band is taken as the estimated temperature. :

[0063]

[0064] get Then, the blackbody radiance corresponding to this temperature can be further calculated. This updates the emissivity of each band accordingly. The expression is:

[0065]

[0066] After obtaining the new emissivity spectrum, the target's own radiation can be reconstructed by combining it with the currently estimated temperature, and then proceed to the next iteration:

[0067]

[0068] In the formula, Indicates the number of iterations. For the first The emissivity spectrum obtained in the next iteration To re-estimate the first Sub-cloud top self-radiance. Then, using Calculate the brightness temperature again:

[0069]

[0070] Take the maximum brightness temperature as the new estimated temperature:

[0071]

[0072] Based on this, the blackbody radiance at the corresponding temperature is further calculated. And update the emissivity of each band:

[0073]

[0074] This forms an iterative process of "target intrinsic radiance - brightness temperature inversion - maximum brightness temperature - emissivity". The above iterations are repeated until the change in the target intrinsic radiance between two adjacent iterations is less than a preset threshold, or the number of iterations reaches a set upper limit N. If the slope of the change in radiance with the number of iterations increases, the algorithm terminates and outputs the initial emissivity spectrum after the iteration is completed.

[0075] After obtaining the initial emissivity spectrum, calculate the target's relative emissivity spectrum. :

[0076]

[0077] In the formula, For the emissivity spectrum obtained through iterative inversion, This is the average emissivity across all bands. It is a relative emissivity spectrum, which reflects the relative fluctuation shape of the DCC emissivity spectrum.

[0078] Calculate the maximum-minimum relative emissivity difference (MMD) based on the relative emissivity spectrum:

[0079]

[0080] After obtaining the MMD, the minimum emissivity of the target is estimated using the empirical relationship between the minimum emissivity and the MMD. This empirical relationship can be expressed as:

[0081]

[0082] In the formula, a and b are empirical coefficients that can be fitted using a ground feature spectral library.

[0083] Finally, the DCC emissivity spectrum was calculated based on the minimum emissivity. :

[0084]

[0085] In other words, the method of using temperature-emissivity separation to retrieve the emissivity spectrum of each DCC pixel in the first DCC pixel set includes: acquiring hyperspectral thermal infrared data corresponding to the first DCC pixel based on the observation time and geographical location of the first DCC pixel in the first DCC pixel set, wherein the hyperspectral thermal infrared data is data observed by an atmospheric infrared detector (AIRS) or an infrared atmospheric sounding interferometer (IASI); performing atmospheric correction on the hyperspectral thermal infrared data based on the atmospheric parameters of the first DCC pixel; and retrieving the emissivity spectrum of the first DCC pixel set using the temperature-emissivity separation method based on the atmospherically corrected hyperspectral thermal infrared data.

[0086] In step 104, the DCC emissivity spectrum can be determined based on the emissivity spectrum of each DCC pixel in the first DCC pixel set and the emissivity spectrum of each DCC pixel in the second DCC pixel set.

[0087] Specifically, the emissivity spectrum of each DCC pixel in the first DCC pixel set and the average of the emissivity spectrum of each DCC pixel in the second DCC pixel set can be calculated to obtain the DCC emissivity spectrum.

[0088] Step 105: Based on the DCC emissivity spectrum, the ocean emissivity spectrum, the atmospheric parameters, the spectral response function of the payload to be calibrated, and the spectral response function of the reference payload, determine the spectral matching factor between the payload to be calibrated and the reference payload.

[0089] In step 105, the DCC top radiance of the channel corresponding to the payload to be calibrated can be simulated based on the atmospheric parameters of each DCC pixel in the first DCC pixel set and the DCC emissivity spectrum; the DCC top radiance of the channel corresponding to the reference payload can be simulated based on the atmospheric parameters of each DCC pixel in the second DCC pixel set and the DCC emissivity spectrum; the ocean top radiance of the channel corresponding to the payload to be calibrated can be simulated based on the atmospheric parameters of each ocean pixel in the first ocean pixel set and the ocean emissivity spectrum; and the ocean top radiance of the channel corresponding to the reference payload can be simulated based on the atmospheric parameters of each ocean pixel in the second ocean pixel set and the ocean emissivity spectrum. Next, based on the DCC emissivity spectrum, the ocean emissivity spectrum, the spectral response function of the payload to be calibrated, the spectral response function of the reference payload, the DCC top radiance of the corresponding channel of the payload to be calibrated, the DCC top radiance of the corresponding channel of the reference payload, the ocean top radiance of the corresponding channel of the payload to be calibrated, and the ocean top radiance of the corresponding channel of the reference payload, the spectral matching factor SBAF between the payload to be calibrated and the reference payload is calculated.

[0090] In step 105, the DCC emissivity spectrum and the ocean emissivity spectrum obtained from the ASTER spectral library are substituted into the radiative transfer equation, respectively. Combining the spectral response functions of the reference load and the load to be calibrated, the spectral matching factor SBAF is calculated to correct for systematic errors caused by differences in the spectral response functions of different load channels. The formula for calculating SBAF is:

[0091]

[0092] In the formula, K is the spectral matching factor; and These are the spectral response functions of the two loads (i.e., the load to be calibrated and the reference load), and their integration intervals [λ1, λ2] and [λ3, λ4] are the wavelength ranges of their respective bands. To simulate the top-of-atmosphere radiance, based on the thermal infrared radiation transfer theory, and ignoring solar reflection or nighttime conditions, the radiance of the target scene at the top of the atmosphere (TOA) can be expressed as:

[0093]

[0094] In the formula, Indicates the target temperature. Indicates atmospheric transmittance. This indicates upward radiation from the atmosphere. Indicates downward atmospheric radiation; Indicates the emissivity.

[0095] For DCC targets, the DCC emissivity spectrum obtained from step 104 is used, and multiple temperature points (e.g., 180-220 K, step size 5 K) are selected within the DCC temperature range. For each temperature point, the DCC top radiance at that temperature is simulated using the MODTRAN radiative transfer model. Then, the top radiance is convolved with the spectral response functions of the payload to be calibrated and the reference payload, respectively, to obtain the simulated top radiance of the corresponding channels for the two payloads at that temperature point. Linear fitting is performed on the two sets of simulated channel radiances for all temperature points, and the slope of the fitted line is the spectral matching factor SBAF. For marine targets, the marine emissivity spectrum from the ASTER spectral library is used, and the SBAF calculation is performed using the same method.

[0096] Step 106: Based on the spectral matching factor, correct the DCC observation radiance and ocean observation radiance of the reference payload respectively to obtain the DCC equivalent reference radiance and ocean equivalent reference radiance respectively.

[0097] In step 106, the DCC observation radiance of the reference payload can be corrected according to the spectral matching factor between the payload to be calibrated and the reference payload to obtain the DCC equivalent reference radiance; and the ocean observation radiance of the reference payload can be corrected according to the spectral matching factor between the payload to be calibrated and the reference payload to obtain the ocean equivalent reference radiance. In step 109, the equivalent reference radiance of the reference payload DCC sample and the ocean sample after SBAF correction are calculated respectively:

[0098]

[0099] By comparing the observed radiance with the load to be calibrated, the relative deviation is calculated to verify the load accuracy. The relative deviation RB can be expressed as:

[0100] .

[0101] Step 107: Determine the calibration coefficient of the payload to be calibrated based on the DCC equivalent reference radiance, the ocean equivalent reference radiance, the DN value of the DCC pixels in the first DCC pixel set, and the DN value of the ocean pixels in the first ocean pixel set.

[0102] In step 109, all DCC and ocean samples are used for calibration coefficient fitting, and the calibration relationship can be expressed as:

[0103]

[0104] In the formula, This is the gain coefficient. This is the bias coefficient.

[0105] The infrared load cross-calibration method provided in this application establishes a thermal infrared cross-calibration technical scheme that takes into account both low-temperature and normal-temperature ranges, given the lack of effective calibration references at the low-temperature end of thermal infrared calibration. DCC is selected to provide calibration references for the low-temperature range, and the ocean provides calibration references for the normal-temperature range, constructing a cross-calibration system that considers both low-temperature and normal-temperature ranges, thereby improving the applicability of calibration coefficients in different temperature zones.

[0106] In addition, the infrared payload cross-calibration method provided in this application is based on hyperspectral thermal infrared observation data. It uses the temperature-emissivity separation method to invert the DCC emissivity spectrum, and then obtains the representative DCC emissivity spectrum through sample averaging and spectral smoothing, providing data support for calculating the SBAF of DCC targets.

[0107] In other words, this application constructs a cross-calibration system that takes into account both cryogenic and ambient temperature ranges. By using DCC as a cryogenic reference target and the ocean as an ambient temperature reference target in combination for cross-calibration, it can effectively compensate for the lack of cryogenic reference targets in cross-calibration of the thermal infrared band and reduce the nonlinear errors caused by extrapolation of calibration relationships in the cryogenic region. Compared with schemes that rely solely on a single type of target, this application can improve the reliability and applicability of calibration results under different scenarios.

[0108] Furthermore, this application solves the problem of emissivity acquisition in the calibration process of DCC, extending the application scope of DCC from thermal infrared band evaluation to cross-calibration. In cross-calibration, SBAF needs to be calculated to eliminate differences in spectral response functions of different loads, and SBAF calculation requires the target's emissivity spectrum. DCC is located at high altitudes, and its emissivity spectrum cannot be obtained from laboratory spectral libraries like that of surface materials. Therefore, existing research only uses DCC for thermal infrared band stability assessment. This application, based on hyperspectral thermal infrared data, uses a temperature-emissivity separation method to invert the DCC emissivity spectrum, solving the problem of DCC emissivity acquisition and providing a feasible path for the application of DCC in infrared band cross-calibration.

[0109] This application also provides an infrared load cross-calibration device 300. For example... Figure 3 As shown, the device 300 includes:

[0110] The acquisition unit 310 is used to acquire the first thermal infrared data observed by the load to be calibrated, and to acquire the second thermal infrared data observed by the reference load.

[0111] The first filtering unit 320 is used to filter deep convection cloud DCC pixels in the first thermal infrared data and the second thermal infrared data respectively, and obtain the first DCC pixel set and the second DCC pixel set respectively.

[0112] The second filtering unit 330 is used to filter ocean pixels in the first thermal infrared data and the second thermal infrared data respectively, and obtain a first ocean pixel set and a second ocean pixel set respectively.

[0113] The first determining unit 340 is used to determine the atmospheric parameters corresponding to the first DCC pixel set, the second DCC pixel set, the first ocean pixel set, and the second ocean pixel set, and to determine the DCC emissivity spectrum and the ocean emissivity spectrum.

[0114] The second determining unit 350 is used to determine the spectral matching factor between the payload to be calibrated and the reference payload based on the DCC emissivity spectrum, the ocean emissivity spectrum, the atmospheric parameters, the spectral response function of the payload to be calibrated, and the spectral response function of the reference payload.

[0115] The correction unit 360 is used to correct the DCC observation radiance and ocean observation radiance of the reference payload according to the spectral matching factor, respectively, to obtain the DCC equivalent reference radiance and the ocean equivalent reference radiance.

[0116] The third determining unit 370 is used to determine the calibration coefficient of the payload to be calibrated based on the DCC equivalent reference radiance, the ocean equivalent reference radiance, the DN value of the DCC pixels in the first DCC pixel set, and the DN value of the ocean pixels in the first ocean pixel set.

[0117] The functions of each functional unit of device 300 can be referred to the above text. Figure 1 or Figure 2 The implementation of the method shown in the examples will not be described in detail here.

[0118] See Figure 4 This application also provides a computing device 400, including a memory 410 and a processor 420; wherein, the memory 410 stores computing instructions; and the processor 420 executes the computing instructions, causing the computing device 400 to perform... Figure 1 The method shown.

[0119] This application also provides a computer-readable storage medium including computer program instructions, which, when executed by a computing device, are executed by the computing device. Figure 1 The method shown.

[0120] This application also provides a computer program product, including a computer program, which, when executed by a processor, implements the following: Figure 1 The method shown.

[0121] The specific embodiments described above further illustrate the purpose, technical solution, and beneficial effects of this application. It should be understood that the above description is only a specific embodiment of this application and is not intended to limit the scope of protection of this application. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of this application should be included within the scope of protection of this application.

Claims

1. A method for cross-calibrating infrared loads, characterized in that, The method includes: Acquire the first thermal infrared data observed by the load to be calibrated, and acquire the second thermal infrared data observed by the reference load; Deep convection cloud DCC pixels are selected from the first thermal infrared data and the second thermal infrared data respectively to obtain the first DCC pixel set and the second DCC pixel set. Ocean pixels are selected from the first thermal infrared data and the second thermal infrared data respectively to obtain the first set of ocean pixels and the second set of ocean pixels. Determine the atmospheric parameters corresponding to the first DCC pixel set, the second DCC pixel set, the first ocean pixel set, and the second ocean pixel set, and determine the DCC emissivity spectrum and the ocean emissivity spectrum; Based on the DCC emissivity spectrum, the ocean emissivity spectrum, the atmospheric parameters, the spectral response function of the payload to be calibrated, and the spectral response function of the reference payload, the spectral matching factor between the payload to be calibrated and the reference payload is determined. Based on the spectral matching factor, the DCC observation radiance and ocean observation radiance of the reference payload are corrected respectively to obtain the DCC equivalent reference radiance and ocean equivalent reference radiance. The calibration coefficients of the payload to be calibrated are determined based on the DCC equivalent reference radiance, the ocean equivalent reference radiance, the DN values ​​of DCC pixels in the first DCC pixel set, and the DN values ​​of ocean pixels in the first ocean pixel set.

2. The method according to claim 1, characterized in that, The step of filtering deep convection cloud DCC pixels from the first thermal infrared data and the second thermal infrared data respectively to obtain a first DCC pixel set and a second DCC pixel set includes: DCC pixels are selected from the first thermal infrared data and the second thermal infrared data based on the brightness temperature threshold and spatial uniformity. The brightness temperature threshold includes: the brightness temperature of the 11μm thermal infrared channel is less than 205K; The spatial uniformity includes: a standard deviation of brightness temperature of less than 1K for the 3×3 pixel thermal infrared channel and a standard deviation of reflectance of less than 3% for the visible light channel.

3. The method according to claim 1 or 2, characterized in that, DCC pixels in the first DCC pixel set and DCC pixels in the second DCC pixel set are matched; ocean pixels in the first ocean pixel set and ocean pixels in the second ocean pixel set are matched. Among them, matching pixels correspond to the same observation area, and the time difference between the observation times of matching pixels is less than or equal to 10 minutes.

4. The method according to claim 3, characterized in that, The observed zenith angle of the DCC pixels in the first DCC pixel set and the DCC pixels in the second DCC pixel set is less than 40°. The first DCC pixel in the first DCC pixel set and the second DCC pixel in the second DCC pixel set are matched; wherein, the relationship between the observed zenith angle of the first DCC pixel and the observed zenith angle of the second DCC pixel satisfies ;in, The observed zenith angle representing the first DCC pixel, This represents the observed zenith angle of the second DCC pixel.

5. The method according to claim 1, characterized in that, The determination of atmospheric parameters corresponding to the first DCC pixel set, the second DCC pixel set, the first ocean pixel set, and the second ocean pixel set includes: Based on the observation time and geographical location of each pixel in the first DCC pixel set, the second DCC pixel set, the first ocean pixel set, and the second ocean pixel set, obtain the corresponding atmospheric profile information; Based on the atmospheric profile information, atmospheric parameters for each pixel are calculated using a radiative transfer model; The atmospheric parameters include atmospheric transmittance, upward atmospheric radiation, and downward atmospheric radiation.

6. The method according to claim 1, characterized in that, The determination of the DCC emissivity spectrum includes: Based on the observation time and geographical location of the DCC pixels in the first DCC pixel set and the DCC pixels in the second DCC pixel set, hyperspectral thermal infrared data corresponding to the DCC pixels in the first DCC pixel set and the DCC pixels in the second DCC pixel set are obtained; atmospheric correction is performed on the hyperspectral thermal infrared data based on the atmospheric parameters of the DCC pixels; and the DCC emissivity spectrum is obtained by inverting the temperature-emissivity separation method based on the atmospherically corrected hyperspectral thermal infrared data.

7. The method according to claim 1, characterized in that, Determining the spectral matching factor between the load to be calibrated and the reference load includes: Based on the DCC emissivity spectrum, the ocean emissivity spectrum, and the atmospheric parameters, simulate the DCC atmospheric top radiance and the ocean atmospheric top radiance; The DCC atmospheric top radiance and the ocean atmospheric top radiance are convolved with the spectral response function of the payload to be calibrated and the spectral response function of the reference payload, respectively, to obtain the corresponding simulated channel radiance. The spectral matching factor is determined based on the simulated channel radiance.

8. An infrared load cross-calibration device, characterized in that, The device includes: The acquisition unit is used to acquire the first thermal infrared data observed by the load to be calibrated, and to acquire the second thermal infrared data observed by the reference load. The first filtering unit is used to filter deep convection cloud DCC pixels from the first thermal infrared data and the second thermal infrared data respectively, and obtain the first DCC pixel set and the second DCC pixel set respectively. The second filtering unit is used to filter ocean pixels from the first thermal infrared data and the second thermal infrared data respectively, and obtain a first ocean pixel set and a second ocean pixel set respectively. The first determining unit is used to determine the atmospheric parameters corresponding to the first DCC pixel set, the second DCC pixel set, the first ocean pixel set, and the second ocean pixel set, and to determine the DCC emissivity spectrum and the ocean emissivity spectrum. The second determining unit is used to determine the spectral matching factor between the payload to be calibrated and the reference payload based on the DCC emissivity spectrum, the ocean emissivity spectrum, the atmospheric parameters, the spectral response function of the payload to be calibrated, and the spectral response function of the reference payload. The correction unit is used to correct the DCC observation radiance and ocean observation radiance of the reference payload according to the spectral matching factor, respectively, to obtain the DCC equivalent reference radiance and the ocean equivalent reference radiance. The third determining unit is used to determine the calibration coefficient of the payload to be calibrated based on the DCC equivalent reference radiance, the ocean equivalent reference radiance, the DN value of the DCC pixels in the first DCC pixel set, and the DN value of the ocean pixels in the first ocean pixel set.

9. A computing device, characterized in that, Including memory and processor; The memory stores calculation instructions; The processor executes the computation instructions, causing the computing device to perform the method of any one of claims 1-7.

10. A computer-readable storage medium, characterized in that, It includes computer program instructions, which, when executed by a computing device, perform the method of any one of claims 1-7.

Images