Method and system for optimizing pre-emission spectral dispersion coefficient of greenhouse gas spectrometer

Abstract

The application discloses a greenhouse gas spectrometer emission front spectrum dispersion coefficient optimization method and system, the method comprises the following steps: obtaining an instrument line function, an initial dispersion coefficient and a laboratory direct sunlight observation spectrum; simulating an atmospheric absorption spectrum based on a radiation transfer model and performing convolution processing by using the instrument line function; screening a position-robust reference absorption line through environmental parameter sensitivity analysis; performing fitting peak searching on the observation spectrum and the simulated spectrum, extracting a central wavelength and calculating a wavelength shift; establishing a polynomial fitting relationship of the wavelength shift to obtain a wavelength shift fitting coefficient; performing difference operation on the initial dispersion coefficient and the wavelength shift fitting coefficient, keeping the high-order term unchanged, and outputting an optimized dispersion coefficient. The application can effectively improve the wavelength calibration precision of the spectrometer before emission.

Description

Technical Field

[0001] This invention belongs to the field of spectral calibration technology, specifically relating to a method for optimizing the spectral dispersion coefficient of a greenhouse gas spectrometer before launch, which is applicable to high-precision spectral calibration of satellite remote sensing instruments before launch. Background Technology

[0002] High-precision spectral calibration is a core prerequisite for ensuring the quality of satellite greenhouse gas remote sensing data and improving the accuracy of CO2 column concentration inversion. Even a small shift in the center wavelength can significantly reduce the accuracy of radiance inversion. To achieve an XCO2 inversion accuracy better than 1% (approximately 4 ppm), greenhouse gas monitoring satellites such as OCO-2, TanSat, and GOSAT all require on-orbit wavelength calibration errors to be less than one-tenth of the spectral resolution. This requirement must be met in advance during the pre-launch calibration phase.

[0003] Greenhouse gas spectrometers generally employ high spectral resolution designs; for example, GOSAT has a maximum resolution of 0.025 nm, while OCO2 / OCO3 and TanSat have maximum resolutions of around 0.04 nm. This extremely high spectral resolution places stringent requirements on the accuracy of spectral calibration.

[0004] Current greenhouse gas spectrometers typically employ a tunable laser in conjunction with a wavelength meter as the calibration light source and reference for pre-emission spectral calibration. The instrument's linear dispersion function (ILS) is acquired through band-by-band, pixel-by-pixel scanning, and then polynomial fitting is used to establish the dispersion relationship between detector pixels and wavelengths, yielding the initial dispersion coefficient. However, during long-term data acquisition across the entire spectrum, fluctuations in laser output stability and changes in ambient temperature can introduce systematic wavelength drift errors, leading to a decrease in the accuracy of the initial dispersion coefficient and making it difficult to meet picometer-level calibration requirements.

[0005] Current methods for optimizing dispersion coefficients generally rely on high-resolution Fourier transform spectrometers (FTS) at ground-based TCCON (Total Carbon Column Observing Network) sites to achieve wavelength calibration through synchronous solar observation spectral comparison. However, TCCON sites are sparsely distributed globally, and this method is significantly limited by location. Most laboratory calibration sites lack available TCCON sites nearby, and spatial differences in atmospheric environment caused by distant sites introduce additional biases. Furthermore, the portable Bruker EM27 spectrometer has a resolution lower than the target instrument's operating wavelength, failing to meet the requirements for high-precision dispersion coefficient optimization.

[0006] The existing technology has the following main drawbacks:

[0007] 1. High dependence on external equipment: Traditional calibration methods rely on the high-resolution Fourier transform spectrometer (FTS) at the TCCON site. When there is no such equipment near the calibration site, this calibration scheme is difficult to apply.

[0008] 2. Effects of laser stability fluctuations: The long-term data acquisition process during the laboratory spectral calibration stage is susceptible to the effects of tunable laser stability fluctuations and changes in environmental parameters, which introduce wavelength drift errors, resulting in a decrease in the calibration accuracy of the initial dispersion coefficient and making it difficult to meet the picometer-level requirements.

[0009] 3. Lack of independent calibration capability without relying on external equipment: Current technology has not yet established an independent calibration system for the laboratory stage based on the radiative transfer model, and cannot get rid of the dependence on external high-precision observation equipment.

[0010] In summary, there is an urgent need to develop a pre-launch spectral dispersion coefficient optimization method that does not rely on external high-precision observation equipment, in order to adapt to the calibration scenarios and accuracy requirements of this instrument. Summary of the Invention

[0011] To address the shortcomings of the existing technologies, this invention provides a method for optimizing the spectral dispersion coefficient of a greenhouse gas spectrometer before emission based on a radiative transfer model. Using a high-precision spectrum simulated by SCIATRAN as a reference, the wavelength shift fitting coefficient is obtained by fitting the center shift results of the observed spectrum and the simulated spectrum at the same reference absorption line. The optimized dispersion coefficient is obtained by subtracting the corresponding order terms of the initial dispersion coefficient of the spectrometer from those of the wavelength shift fitting coefficient, while keeping higher-order terms unchanged. Using this dispersion coefficient, the wavelength calibration accuracy can meet the requirement of one-tenth of the spectral resolution.

[0012] A method for optimizing the spectral dispersion coefficient of a greenhouse gas spectrometer before emission includes the following steps:

[0013] Obtain the instrument linearity function, initial dispersion coefficient, and solar direct observation spectrum collected in the laboratory from the greenhouse gas spectrometer;

[0014] Atmospheric absorption spectra are simulated based on a radiative transfer model, and the simulated spectra are convolved using the instrument's linear function to obtain simulated spectra that match the instrument's characteristics.

[0015] Environmental parameter sensitivity analysis was performed on the simulated spectrum to screen out positionally robust reference absorption lines;

[0016] The observed and simulated spectra are fitted to the reference absorption line positions to find peaks, the center wavelength is extracted, and the wavelength shift is calculated.

[0017] Establish a polynomial fitting relationship for the wavelength offset to obtain the wavelength offset fitting coefficients;

[0018] The initial dispersion coefficient and the wavelength shift fitting coefficient are compared by performing a difference operation, keeping the higher-order terms unchanged, and the optimized dispersion coefficient is output.

[0019] Specifically, obtaining the initial dispersion coefficient of the greenhouse gas spectrometer includes:

[0020] A tunable laser was used as a light source in conjunction with a wavelength meter to perform a full-image scan of a greenhouse gas spectrometer and obtain the instrument's linear function data.

[0021] Based on the instrument's linear function data, a fifth-order polynomial relationship between the detector pixel number and the wavelength is established to obtain the initial dispersion coefficient.

[0022] Specifically, obtaining the solar direct observation spectrum collected in the laboratory includes:

[0023] A greenhouse gas spectrometer was placed inside a vacuum chamber to simulate an on-orbit environment, and sunlight was introduced to collect spectra.

[0024] The collected spectral data were normalized, and spectral regions with high transmittance were selected as the observed spectra.

[0025] Specifically, the simulation of atmospheric absorption spectra based on the radiative transfer model includes:

[0026] The TSIS-1 HSRS solar spectrum was used as the incident solar flux at the top of the atmosphere;

[0027] Replace the atmospheric pressure, temperature, and water vapor vertical profiles in the basic atmospheric parameters with ERA5 reanalysis data.

[0028] The pressure profile was scaled and corrected by combining measured surface pressure data, and the simulated geometric angles were set to be consistent with the actual acquisition state of the observed spectrum.

[0029] Specifically, the step of performing environmental parameter sensitivity analysis on the simulated spectrum to screen out positionally robust reference absorption lines includes:

[0030] Preset errors are applied to the pressure profile and temperature profile respectively, and the water vapor profile is scaled by a preset ratio.

[0031] The effects of changes in air pressure, temperature, and water vapor on the position of simulated spectral wavelengths were analyzed, and sensitive absorption lines were eliminated.

[0032] The absorption line that is insensitive to changes in any of the three types of environmental parameters is retained as the wavelength calibration reference.

[0033] Specifically, the step of fitting the observed spectrum and the simulated spectrum to find peaks at the reference absorption line positions includes:

[0034] A fitted dataset is formed by selecting a preset number of data points centered on the valley value of the absorption line.

[0035] The center wavelength of the absorption line is extracted by performing least-squares fitting on the fitted dataset using an asymmetric Gaussian function.

[0036] Specifically, establishing the polynomial fitting relationship for the wavelength offset includes:

[0037] The optimal order of the wavelength offset polynomial fitting is determined using leave-one-out cross-validation (LOOCV).

[0038] For the O2A band, weak CO2 band, and strong CO2 band, based on the LOOCV results, the relationship between wavelength shift and center wavelength is established by fitting with the optimal order (e.g., fourth order).

[0039] Specifically, the step of performing a difference calculation between the initial dispersion coefficient and the wavelength shift fitting coefficient includes:

[0040] Subtract the lower-order terms of the initial dispersion coefficient from the corresponding terms of the wavelength shift fitting coefficient;

[0041] By keeping the value of the highest-order term of the initial dispersion coefficient unchanged, the optimized dispersion coefficient is obtained.

[0042] Among them, the application in wavelength calibration before the emission of a greenhouse gas spectrometer includes three detection channels: O2A band, weak CO2 band, and strong CO2 band.

[0043] This invention also proposes an optimization system for the spectral dispersion coefficient of a greenhouse gas spectrometer before emission, comprising:

[0044] The data acquisition module is used to acquire the instrument linear function, initial dispersion coefficient, and solar direct observation spectrum collected in the laboratory from the greenhouse gas spectrometer.

[0045] The spectral simulation module is used to simulate atmospheric absorption spectra based on radiative transfer models and generate simulated spectra that match the instrument characteristics using instrument linearity functions.

[0046] The benchmark screening module is used to perform environmental parameter sensitivity analysis on the simulated spectrum and screen out the position-robust reference absorption lines.

[0047] The parameter extraction module is used to fit the observed spectrum and the simulated spectrum to the reference absorption line position to find the peak, extract the center wavelength and calculate the wavelength shift.

[0048] The coefficient optimization module is used to establish a polynomial fitting relationship for the wavelength offset, calculate the difference between the initial dispersion coefficient and the wavelength offset fitting coefficient, and output the optimized dispersion coefficient.

[0049] Therefore, this invention provides a method for optimizing the spectral dispersion coefficient of a greenhouse gas spectrometer before emission based on a radiative transfer model. The method determines the target calibration band based on instrument parameters and calibration requirements, obtains the initial dispersion coefficient and observed spectral data from laboratory measurements, simulates the atmospheric absorption spectrum using the SCIATRAN radiative transfer model, convolves the simulated spectrum with the actual instrument response function to obtain a simulated spectrum matching the instrument characteristics, selects a positionally robust reference absorption line based on multi-parameter sensitivity analysis, extracts the center wavelength of the observed and simulated spectra using asymmetric Gaussian fitting, calculates the wavelength shift and uses polynomial fitting to obtain the wavelength shift fitting coefficient, and calculates the optimized dispersion coefficient based on the initial dispersion coefficient and the wavelength shift fitting coefficient. This method enables the laboratory spectral dispersion coefficient optimization method provided by this invention to achieve picometer-level wavelength calibration accuracy without relying on external high-precision FTS instruments. It effectively introduces the radiative transfer model into the laboratory stage of spectral calibration, overcomes the shortcomings of existing calibration schemes that depend on external equipment, and solves the wavelength drift problem caused by laser stability fluctuations and changes in environmental parameters.

[0050] The present invention has the following beneficial effects:

[0051] a) Eliminating dependence on external equipment: This invention introduces the SCIATRAN radiative transfer model to build a calibration benchmark, eliminating the need to rely on high-precision ground-based FTS instruments such as TCCON. This solves the calibration problem when there are no high-precision observation devices around the calibration site or the distance is too far, significantly improving the autonomy and flexibility of pre-launch spectral calibration.

[0052] b) Achieving picometer-level wavelength calibration accuracy: Through a closed-loop process of "model simulation - benchmark screening - fitting peak finding - coefficient calibration", the wavelength shifts of the O2A band, weak CO2 band and strong CO2 band are reduced from the initial 12 picometers, 60 picometers and 80 picometers to less than 4 picometers, 7 picometers and 9 picometers respectively. The wavelength calibration accuracy is better than one-tenth of the spectral resolution, meeting the stringent calibration requirements of greenhouse gas spectrometers.

[0053] c) Ensure the robustness of the calibration benchmark: By conducting sensitivity analysis on the three major environmental parameters of air pressure, temperature and water vapor, 46 robust reference absorption lines that are not sensitive to changes in environmental parameters were selected. This effectively avoids the impact of environmental parameter errors on the calibration results in simulation and actual measurement, and improves the reliability and adaptability of the calibration method.

[0054] d) The method has strong universality: This invention is not only applicable to the laboratory calibration of greenhouse gas spectrometers, but can also be extended to the optimization of spectral dispersion coefficients of other hyperspectral remote sensing instruments, and has broad engineering application value. Attached Figure Description

[0055] The above and other objects, features, and advantages of exemplary embodiments of the present disclosure will become readily apparent upon reading the following detailed description with reference to the accompanying drawings. In the drawings, several embodiments of the present disclosure are illustrated by way of example and not limitation, and like or corresponding reference numerals denote like or corresponding parts, wherein:

[0056] Figure 1 This is a flowchart of the pre-emission spectral dispersion coefficient optimization method provided in an embodiment of the present invention;

[0057] Figure 2 These are the reference absorption lines selected for three bands in a specific example of the present invention. From top to bottom, they are the selected reference absorption lines for the O2A band (17 lines), the weak CO2 band (13 lines), and the strong CO2 band (16 lines) (marked with black crosses).

[0058] Figure 3 This is a study on the optimal order selection for fitting the wavelength offset of three bands in a specific example of the present invention; the left figure shows the fitting results of the O2A band, weak CO2 band and strong CO2 band using the optimal order (4th order) under different spatial footprints (FP); the right figure shows the mean ± standard deviation (STD) of the root mean square error (RMSE) of the leave-one-out cross-validation method (LOOCV) for the O2A band, weak CO2 band and strong CO2 band using different polynomial orders, and the selected optimal order is marked with a red dashed line;

[0059] Figure 4 This is a diagram showing the center wavelength shift before and after dispersion coefficient optimization in a specific embodiment of the present invention; where FP is the spatial footprint. The left diagram, from top to bottom, shows the center wavelength difference results of the O2A band, weak CO2 band, and strong CO2 band in the range of FP15-FP25 before dispersion coefficient optimization; the right diagram, from top to bottom, shows the center wavelength difference results of the corresponding three spectral bands after dispersion coefficient optimization.

[0060] Figure 5 This is a comparison of the observed spectrum (blue) and the simulated spectrum (orange) before and after dispersion coefficient optimization in a specific embodiment of the present invention. The left panel shows a full-spectrum comparison of the O2A band, the weak CO2 band, and the strong CO2 band. The middle panel shows a magnified view of the area indicated by the black dashed box. The right panel shows the corresponding optimized results.

[0061] Figure 6This represents the residuals between the normalized observed and simulated spectra before and after optimization of the dispersion coefficients for the O2A, weak CO2, and strong CO2 bands in a specific embodiment of this invention. The first to third rows show the residuals for these three bands, with the left and right subplots corresponding to the results before and after optimization, respectively. Colored curves represent different spatial coverage ranges (FP15–FP25), and the black dashed line indicates a residual of zero. The fourth row shows the root mean square error (RMSE) box plots before and after optimization.

[0062] Figure 7 This is the result of verifying the optimized spectral dispersion coefficients for the O2A band, weak CO2 band, and strong CO2 band in a specific embodiment of this invention. The left subplot (box plot with solid borders and light gray fill) shows the unoptimized wavelength shift distribution, and the right subplot (box plot with solid borders and dark gray fill) shows the corresponding optimized results. The three rows correspond to these three spectral bands respectively. The box plot summarizes the shift of five independent validation datasets at the selected reference line within the FP15-FP25 range. The black dashed line represents a 10% spectral resolution tolerance.

[0063] Figure 8 This is a flowchart of the pre-emission spectral dispersion coefficient optimization method in a specific embodiment of the present invention. Detailed Implementation

[0064] To make the objectives, technical solutions, and advantages of this invention clearer, the invention will be further described in detail below with reference to the accompanying drawings. Obviously, the described embodiments are only a part of the embodiments of this invention, and not all of them. Based on the embodiments of this invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of this invention.

[0065] The terminology used in the embodiments of this invention is for the purpose of describing particular embodiments only and is not intended to limit the invention. The singular forms “a,” “the,” and “the” as used in the embodiments of this invention and the appended claims are also intended to include the plural forms, and “multiple” generally includes at least two unless the context clearly indicates otherwise.

[0066] It should be understood that although the terms first, second, third, etc., may be used to describe... in the embodiments of the present invention, these... should not be limited to these terms. These terms are only used to distinguish... For example, first... may also be referred to as second... without departing from the scope of the embodiments of the present invention, and similarly, second... may also be referred to as first...

[0067] It should be understood that the term "and / or" used in this article is merely a description of the relationship between related objects, indicating that three relationships can exist. For example, A and / or B can represent: A existing alone, A and B existing simultaneously, and B existing alone. Additionally, the character " / " in this article generally indicates that the preceding and following related objects have an "or" relationship.

[0068] Depending on the context, the words “if” or “suppose” as used here can be interpreted as “when” or “in response to determination” or “in response to detection.” Similarly, depending on the context, the phrases “if determination” or “if detection (of the stated condition or event)” can be interpreted as “when determination” or “in response to determination” or “when detection (of the stated condition or event)” or “in response to detection (of the stated condition or event).”

[0069] It should also be noted that the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that an article or device that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such an article or device. Without further limitation, an element defined by the phrase "comprising one..." does not exclude the presence of other identical elements in the article or device that includes said element.

[0070] This invention provides a method for optimizing the spectral dispersion coefficient of a greenhouse gas spectrometer before launch based on a radiative transfer model. According to instrument parameters and calibration requirements, the target calibration band and laboratory wavelength calibration accuracy requirements are determined. Based on the target instrument, the instrument linearity function (ILS), initial dispersion coefficient, and spectral data from direct solar observation under clear skies are obtained. Using the SCIATRAN radiative transfer model, the atmospheric absorption spectrum received at the ground is simulated. Based on the instrument spectral response function tested in the laboratory, the simulated spectrum is convolved with the instrument response function to obtain a simulated spectrum matching the instrument characteristics. Based on the sensitivity analysis results of atmospheric pressure, temperature, and water vapor environmental parameters, positionally robust reference absorption lines are selected from the simulated spectrum. According to the asymmetric Gaussian fitting method, the observed spectrum and the simulated spectrum are fitted and peaked at the reference absorption line position respectively to extract their respective center wavelengths. Based on the wavelength offset calculation results, the wavelength offset fitting coefficient is obtained by polynomial fitting. Based on the initial dispersion coefficient and the wavelength offset fitting coefficient, the optimized dispersion coefficient is calculated, thereby optimizing the spectral dispersion coefficient in the laboratory. Thus, the pre-emission spectral dispersion coefficient optimization method provided by this invention does not rely on external high-precision FTS instruments, achieving picometer-level wavelength calibration accuracy. It effectively introduces the radiative transfer model into the spectral calibration at the laboratory stage, making up for the shortcomings of existing calibration schemes that rely on external equipment, and solving the wavelength drift problem caused by laser stability fluctuations and changes in environmental parameters.

[0071] The optional embodiments of the present invention will now be described in detail with reference to the accompanying drawings.

[0072] Example 1

[0073] This invention discloses a method for optimizing the spectral dispersion coefficient of a greenhouse gas spectrometer before emission, comprising the following steps:

[0074] Obtain the instrument linearity function, initial dispersion coefficient, and solar direct observation spectrum collected in the laboratory from the greenhouse gas spectrometer;

[0075] Atmospheric absorption spectra are simulated based on a radiative transfer model, and the simulated spectra are convolved using the instrument's linear function to obtain simulated spectra that match the instrument's characteristics.

[0076] Environmental parameter sensitivity analysis was performed on the simulated spectrum to screen out positionally robust reference absorption lines;

[0077] The observed and simulated spectra are fitted to the reference absorption line positions to find peaks, the center wavelength is extracted, and the wavelength shift is calculated.

[0078] Establish a polynomial fitting relationship for the wavelength offset to obtain the wavelength offset fitting coefficients;

[0079] The initial dispersion coefficient and the wavelength shift fitting coefficient are compared by performing a difference operation, keeping the higher-order terms unchanged, and the optimized dispersion coefficient is output.

[0080] Specifically, obtaining the initial dispersion coefficient of the greenhouse gas spectrometer includes:

[0081] A tunable laser was used as a light source in conjunction with a wavelength meter to perform a full-image scan of a greenhouse gas spectrometer and obtain the instrument's linear function data.

[0082] Based on the instrument's linear function data, a fifth-order polynomial relationship between the detector pixel number and the wavelength is established to obtain the initial dispersion coefficient.

[0083] Specifically, obtaining the solar direct observation spectrum collected in the laboratory includes:

[0084] A greenhouse gas spectrometer was placed inside a vacuum chamber to simulate an on-orbit environment, and sunlight was introduced to collect spectra.

[0085] The collected spectral data were normalized, and spectral regions with high transmittance were selected as the observed spectra.

[0086] Specifically, the simulation of atmospheric absorption spectra based on the radiative transfer model includes:

[0087] The TSIS-1 HSRS solar spectrum was used as the incident solar flux at the top of the atmosphere;

[0088] Replace the atmospheric pressure, temperature, and water vapor vertical profiles in the basic atmospheric parameters with ERA5 reanalysis data.

[0089] The pressure profile was scaled and corrected by combining measured surface pressure data, and the simulated geometric angles were set to be consistent with the actual acquisition state of the observed spectrum.

[0090] Specifically, the step of performing environmental parameter sensitivity analysis on the simulated spectrum to screen out positionally robust reference absorption lines includes:

[0091] Preset errors are applied to the pressure profile and temperature profile respectively, and the water vapor profile is scaled by a preset ratio.

[0092] The effects of changes in air pressure, temperature, and water vapor on the position of simulated spectral wavelengths were analyzed, and sensitive absorption lines were eliminated.

[0093] The absorption line that is insensitive to changes in any of the three types of environmental parameters is retained as the wavelength calibration reference.

[0094] Specifically, the step of fitting the observed spectrum and the simulated spectrum to find peaks at the reference absorption line positions includes:

[0095] A fitted dataset is formed by selecting a preset number of data points centered on the valley value of the absorption line.

[0096] The center wavelength of the absorption line is extracted by performing least-squares fitting on the fitted dataset using an asymmetric Gaussian function.

[0097] Specifically, establishing the polynomial fitting relationship for the wavelength offset includes:

[0098] The optimal order of the polynomial fitting for wavelength offset is determined using leave-one-out cross-validation (LOOCV). For the O2A band, weak CO2 band, and strong CO2 band, the relationship between wavelength offset and center wavelength is established by fitting with the optimal order (e.g., fourth order) based on the LOOCV results.

[0099] Specifically, the step of performing a difference calculation between the initial dispersion coefficient and the wavelength shift fitting coefficient includes:

[0100] Subtract the lower-order terms of the initial dispersion coefficient from the corresponding terms of the wavelength shift fitting coefficient;

[0101] By keeping the value of the highest-order term of the initial dispersion coefficient unchanged, the optimized dispersion coefficient is obtained.

[0102] The optimized dispersion coefficient is used to correct the wavelength calibration accuracy of the greenhouse gas spectrometer to better than one-tenth of the spectral resolution.

[0103] The optimized dispersion coefficients result in the following: the wavelength shift of the O2A band is reduced to below 4 picometers, the wavelength shift of the weak CO2 band is reduced to below 7 picometers, and the wavelength shift of the strong CO2 band is reduced to below 9 picometers.

[0104] Example 2

[0105] The present invention proposes a method for optimizing the spectral dispersion coefficient of a greenhouse gas spectrometer before emission, comprising the following steps:

[0106] Figure 1 This is a flowchart of the laboratory spectral dispersion coefficient optimization method provided in the embodiments of the present invention.

[0107] The following is based on Figure 1 The laboratory spectral dispersion coefficient optimization method of the present invention is described, such as... Figure 1 As shown, the method for optimizing the spectral dispersion coefficient in the laboratory includes the following steps:

[0108] Step 101: Based on the instrument parameters and calibration requirements, determine the target calibration band and the required laboratory spectral calibration accuracy.

[0109] In practical applications, for example, for a three-channel grating-type high-resolution greenhouse gas spectrometer, the target calibration bands for the O2A band, weak CO2 band, and strong CO2 band are defined as 757.5–772.5 nm, 1595–1625 nm, and 2040–2080 nm, respectively, with spectral resolutions of 0.04 nm (O2A band), 0.07 nm (weak CO2 band), and 0.09 nm (strong CO2 band). The wavelength calibration accuracy requirement is better than one-tenth of the spectral resolution, i.e., O2A band ≤ 4 picometers, weak CO2 band ≤ 7 picometers, and strong CO2 band ≤ 9 picometers. The detector's spectral dimension pixel count is 1242 (O2A band), 1304 (weak CO2 band), and 1360 (strong CO2 band), respectively, with a spatial dimension pixel count of 488 for all bands. This is just an example.

[0110] It should be noted that different instruments have different band settings and resolution parameters, and the target calibration band and calibration accuracy requirements are also different.

[0111] Step 102: Based on the target instrument, obtain the instrument linearity function (ILS), initial dispersion coefficient, and spectral data of direct solar radiation under clear skies obtained from laboratory measurements;

[0112] A tunable laser output was used as the laboratory calibration light source. After beam splitting, a portion of the laser light entered the wavelength meter path for precise monitoring of the output wavelength, another portion entered the power meter path for real-time monitoring of the output light response, and the remaining portion was homogenized by a rotating integrating sphere before entering a vacuum chamber where it was collected by a greenhouse gas spectrometer to obtain laboratory spectral calibration data. Through pixel-by-pixel scanning, the instrument response function (ILS) of the entire image plane was obtained, totaling 140,616 ILS.

[0113] Based on ILS measurement data, a dispersion curve between wavelength values ​​and detector pixel numbers was determined using fifth-order polynomial fitting, and the initial laboratory dispersion coefficient was established.

[0114] ;

[0115] Where, λ p It is the center wavelength of each pixel, a i p is the spectral dispersion coefficient calibrated in the laboratory. i The initial dispersion coefficient represents the spectral pixel number p raised to the power of i, where n is the polynomial fitting order. This initial dispersion coefficient is obtained by fitting the whole-image scanning data of a tunable laser.

[0116] Atmospheric absorption spectra collected by a greenhouse gas spectrometer under clear-sky conditions were selected as observation data. The spectrometer was placed inside a vacuum chamber to simulate the on-orbit vacuum thermal environment. Sunlight was introduced into the vacuum chamber through a heliostat and a series of mirrors to complete the spectral acquisition. Spectra with relatively high intermediate transmittance were selected for subsequent calibration analysis.

[0117] It should be noted that the initial dispersion coefficient and the observed atmospheric absorption spectrum are different for different instruments.

[0118] Step 103: Based on the SCIATRAN radiative transfer model, simulate the incident radiation at the top of the atmosphere and the atmospheric absorption spectrum.

[0119] The high-resolution TSIS-1 HSRS solar spectrum recommended by the Global Space Cross-Calibration System (GSICS) was used as the atmospheric top incident solar flux for the SCIATRAN model. This TSIS-1 HSRS spectrum covers 202–2730 nm with a spectral resolution of approximately 0.001–0.01 nm. Mid-latitude spring model atmospheric data from the Bremen Two-Dimensional Atmospheric Model (B2D) was used as the basic atmospheric parameters. ERA5 reanalysis data, after spatiotemporal interpolation, replaced the atmospheric pressure, temperature, and water vapor vertical profiles in the B2D model. The pressure profile was scaled and corrected by incorporating measured surface pressure data from meteorological stations. Parameters such as time, latitude and longitude, solar zenith angle, and azimuth angle were kept consistent with the actual acquisition conditions of the observed spectra to simulate the atmospheric absorption spectrum of the target band.

[0120] In practical applications, for example, the model uses SCIATRAN version 4.7.6, and the simulation process employs a planar parallel atmospheric model (pp_scat) and the discrete coordinate method (DOM_S). The radiation intensity is calculated using the line-by-line (LBL) integration method. This is just an example.

[0121] Step 104: Based on the instrument spectral response function tested in the laboratory, the simulated spectrum is convolved with the instrument response function to obtain a simulated spectrum that matches the instrument characteristics.

[0122] Specifically, the high-precision spectrum simulated by the SCIATRAN model is convolved with the instrument spectral response function (ISRF) tested in the laboratory to obtain the convolved simulated spectrum I. sim The calculation formula is as follows:

[0123] ;

[0124] Among them, I (scitran,λ)This represents the high-precision spectrum obtained based on the SCIATRAN model simulation. ISRF(λ) is the spectral response function of the greenhouse gas spectrometer as measured in the laboratory. dλ is an infinitesimal wavelength element, representing the response of I in a continuous wavelength space. (scitran,λ) Integrate the product ISRF(λ).

[0125] It should be noted that convolution processing avoids the optimization error of spectral dispersion coefficient caused by the difference between spectral resolution and sampling interval, so that the simulated spectrum and the observed spectrum have matching instrument response characteristics.

[0126] Step 105: Based on the sensitivity analysis results of air pressure, temperature and water vapor environmental parameters, select positionally robust reference absorption lines from the simulated spectrum.

[0127] Specifically, based on the ±2hPa pressure uncertainty, pressure errors ranging from -2hPa to +2hPa are applied to each layer of the pressure profile. The influence of the vertical pressure profile on the simulated spectral wavelength position is analyzed, and absorption lines with an absolute value of center wavelength offset greater than 0.1% of the corresponding band spectral resolution are removed.

[0128] Based on a ±2 K temperature error, a temperature error of -2 K to +2 K is applied to each layer of the temperature profile. The influence of the temperature profile on the wavelength position of the simulated spectrum is analyzed, and absorption lines with an absolute value of center wavelength offset greater than 0.1% of the spectral resolution of the corresponding band are removed.

[0129] Based on a water vapor profile error of ±0.2 g / cm², the water vapor profile was scaled after applying a water vapor error of -0.2 g / cm² to +0.2 g / cm² to the real-time water vapor content of 2.07 g / cm² calculated in ERA5. The influence of water vapor content on the simulated spectral wavelength position was analyzed, and absorption lines with an absolute value of center wavelength offset greater than 0.1% of the spectral resolution of the corresponding band were removed.

[0130] Absorption lines that are insensitive to changes in the three types of environmental parameters were retained as wavelength calibration benchmarks, and finally 17 reference absorption lines were determined for the O2A band (758.808~770.107 nm), 13 weak CO2 bands (1596.44~1623.61 nm), and 16 strong CO2 bands (2047.87~2074.24 nm).

[0131] Figure 2 This is the result of selecting reference absorption lines for three bands in a specific embodiment of the present invention;

[0132] In practical applications, for example, if a weak CO2 band has one absorption line with a wavelength shift exceeding 0.1% of spectral resolution in temperature sensitivity analysis, that absorption line at that location will be removed; if a strong CO2 band has three absorption lines with wavelength shifts exceeding 0.1% of spectral resolution in water vapor sensitivity analysis, those three absorption lines will be removed; if a weak CO2 band has three absorption lines with wavelength shifts exceeding 0.1% of spectral resolution in water vapor sensitivity analysis, those three absorption lines will be removed (one of which is the same as the absorption line with a wavelength shift exceeding 0.1% of spectral resolution in temperature sensitivity analysis); if a strong CO2 band has six absorption lines with wavelength shifts exceeding 0.1% of spectral resolution in water vapor sensitivity analysis, those six absorption lines will be removed. This is just one example. In this embodiment, the quantitative criterion for determining whether an absorption line is a "sensitive absorption line" is as follows: after applying pressure error (±2 hPa), temperature error (±2 K), and water vapor error (±0.2 g / cm²), if the absolute value of the center wavelength shift of the absorption line exceeds 0.1% of the spectral resolution of the corresponding band under any of these conditions, it is determined to be a sensitive absorption line and is discarded. Only absorption lines that are insensitive to changes in the three types of environmental parameters are retained as wavelength calibration benchmarks.

[0133] Step 106: Based on the asymmetric Gaussian fitting method, the observed spectrum and the simulated spectrum are fitted and peaked at the reference absorption line position respectively, and their respective center wavelengths are extracted.

[0134] Using the valley value of the absorption line as the center, take 3 data points on each side of the O2A band for a total of 7 data points, take 2 data points on each side of the weak CO2 band for a total of 5 data points, and take 3 data points on each side of the strong CO2 band for a total of 7 data points to form the fitting dataset.

[0135] An asymmetric Gaussian function was used for fitting, and the amplitude, center wavelength, standard deviation, asymmetry factor, and background intensity parameters were determined through an optimization algorithm to minimize the residual between the fitted value and the actual data points, thereby extracting the center wavelength of the absorption line.

[0136] Specifically, the formula for the asymmetric Gaussian function is:

[0137] ;

[0138] Where A represents amplitude, x represents the wavelength value used for fitting, μ corresponds to the center wavelength position of the absorption line, σ is the standard deviation, α is the asymmetry factor (when α=0, it is a symmetric Gaussian function), and B represents the background intensity. During the fitting process, amplitude A, center wavelength μ, standard deviation σ, asymmetry factor α, and background intensity B are all used as parameters to be optimized. The optimal values ​​of each parameter are determined using least squares optimization methods such as the Levenberg-Marquardt algorithm, minimizing the sum of squared residuals between the fitted curve and the actual spectral data points.

[0139] In practical applications, for example, using the Normalized Root Mean Square Error (NRMSE) and Bayesian Information Criterion (BIC) to characterize the fitting effect, asymmetric Gaussian fitting exhibits lower NRMSE and BIC (lower values ​​indicate better performance) across all three bands compared to the ordinary Gaussian model and the Gaussian-super-Gaussian mixture model. This is just one example.

[0140] Step 107: Based on the wavelength offset calculation results, the wavelength offset fitting coefficients are obtained by polynomial fitting.

[0141] Calculate the difference in center wavelength between the observed and simulated spectra at each reference absorption line. The optimal fitting order for the wavelength shifts of the O2A band, weak CO2 band, and strong CO2 band was determined to be fourth order using leave-one-out cross-validation (LOOCV). Therefore, the optimal order (e.g., fourth order) was uniformly used to fit the wavelength shifts of the three bands, establishing a polynomial relationship between the wavelength shift and the center wavelength.

[0142] Specifically, the formula for polynomial fitting of wavelength offset is:

[0143] ;

[0144] in, This is the difference between the observed spectrum and the simulated spectrum, where m is the fitting order (m=4), and b i These are the fitting coefficients, p i It represents the spectral pixel number p raised to the power of i.

[0145] Figure 3 This is the fitting result of the wavelength offset of the three bands in a specific example of the present invention;

[0146] In practical applications, for example, according to Leave-One-out Cross-Validation (LOOCV), using a fourth-order fit for the wavelength shifts of the three bands yields the lowest RMSE. Therefore, fitting the three bands with the optimal order (e.g., fourth-order) yields better results. This is just one example.

[0147] Step 108: Based on the initial dispersion coefficient and the wavelength shift fitting coefficient, the calibrated dispersion coefficient is calculated to optimize the laboratory spectral dispersion coefficient.

[0148] Subtracting the initial laboratory dispersion coefficient from the wavelength offset fitting coefficient by the corresponding order terms, while keeping higher-order terms unchanged, yields the calibrated dispersion coefficient, reducing the wavelength deviation between the observed spectrum and the simulated spectrum to within one-tenth of the spectral resolution.

[0149] Specifically, the formula for the calibrated dispersion coefficient is:

[0150]

[0151] Among them, c i It is the calibrated dispersion coefficient, a i It is the initial dispersion coefficient before calibration, b i Here, c5 represents the center wavelength shift fitting coefficient, n is the number of fitting iterations before calibration (n=5), and m is the number of fitting iterations for the wavelength shift (m=4). It should be noted that "higher-order terms" refer to terms in the initial dispersion coefficients whose order is greater than the wavelength shift fitting order m. In this embodiment, the initial dispersion coefficients are fifth-order polynomials (n=5), and the wavelength shift fitting is a fourth-order polynomial (m=4). Therefore, the coefficient a5 for i=5 remains unchanged, i.e., c5=a5; for coefficients from i=0 to i=4, c5 is adjusted according to c5. i = a i - b i Update accordingly. If n ≠ m, then the coefficients for orders higher than m remain unchanged.

[0152] Figure 4 This is a diagram showing the center wavelength shift before and after dispersion coefficient optimization in a specific embodiment of the present invention;

[0153] The calibration results show that the wavelength shifts of the O2A band, weak CO2 band, and strong CO2 band decreased from the initial 12 picometers, 60 picometers, and 80 picometers to less than 4 picometers, 7 picometers, and 9 picometers, respectively. The wavelength calibration accuracy is better than one-tenth of the spectral resolution, meeting the high-precision index requirements for laboratory spectral calibration.

[0154] Figure 5 This is a comparison of the spectra before and after the dispersion coefficient optimization in a specific example of the present invention;

[0155] Without optimization, the wavelength mismatch was very obvious, while the optimized spectrum showed a significantly improved line alignment effect.

[0156] Figure 6 This is a residual plot between the normalized observed spectrum and the simulated spectrum before and after the dispersion coefficient optimization in a specific example of the present invention.

[0157] The results show that the residual calculated from the difference between the normalized observed spectrum and the simulated spectrum after the dispersion coefficient optimization is reduced, and the root mean square error (RMSE) box plot shows that the errors in all three bands show a continuous downward trend.

[0158] Figure 7 This is a verification diagram of the spectral dispersion coefficient optimization in a specific embodiment of the present invention;

[0159] To independently evaluate the effectiveness of the optimization coefficients, five validation datasets that were not involved in the optimization were selected. These datasets were acquired at least 30 minutes after the optimization datasets were collected. Figure 7 As shown, the light gray box plot on the left represents the unoptimized wavelength shift distribution, while the dark gray box plot on the right represents the optimized result. These plots respectively illustrate the statistical distribution of wavelength shifts from FP15 to FP25 at selected absorption lines in the O2A, weak CO2, and strong CO2 bands. The dashed lines in the figure represent a 10% spectral resolution tolerance. After applying the optimized spectral dispersion coefficient, all wavelength shifts are controlled within 10% of the spectral resolution of the corresponding spectral band.

[0160] Figure 8 This is a flowchart of the pre-emission spectral dispersion coefficient optimization method in a specific embodiment of the present invention. Taking a greenhouse gas spectrometer as an example, the above process is illustrated in the flowchart shown in Figure 5.

[0161] Step 1: Data Preparation. Obtain the ILS and initial dispersion coefficient α measured by the spectrometer. i In addition, solar observation spectra were collected under clear skies in mid-April 2025 and normalized. At the same time, high-precision spectra were simulated using SCIATRAN and convolved with the ILS of the spectrometer to obtain simulated spectra consistent with the instrument response characteristics. The simulated spectra were then normalized.

[0162] Step 2: Select a stable reference absorption line as the wavelength reference. By analyzing the sensitivity of the SCIATRAN simulated spectrum to pressure, temperature, and water vapor, a reference absorption line that is insensitive to changes in these three parameters is selected as the wavelength reference.

[0163] Step 3: Determine the center wavelength. Use the asymmetric Gaussian fitting method to determine the center wavelength at the reference absorption line for both the simulated and observed spectra.

[0164] Step 4: Optimize the dispersion coefficient. First, calculate the center wavelength shift of the observed and simulated spectra at the reference absorption line. Then, use a polynomial to fit the wavelength shift to obtain the wavelength shift polynomial fitting coefficient. Finally, subtract the corresponding order terms of the initial dispersion coefficient of the spectrometer from the wavelength shift fitting coefficient, keeping higher-order terms unchanged, to obtain the optimized dispersion coefficient.

[0165] Example 3

[0166] This invention also proposes an optimization system for the spectral dispersion coefficient of a greenhouse gas spectrometer before emission, comprising:

[0167] The data acquisition module is used to acquire the instrument linear function, initial dispersion coefficient, and solar direct observation spectrum collected in the laboratory from the greenhouse gas spectrometer.

[0168] The spectral simulation module is used to simulate atmospheric absorption spectra based on radiative transfer models and generate simulated spectra that match the instrument characteristics using instrument linearity functions.

[0169] The benchmark screening module is used to perform environmental parameter sensitivity analysis on the simulated spectrum and screen out the position-robust reference absorption lines.

[0170] The parameter extraction module is used to fit the observed spectrum and the simulated spectrum to the reference absorption line position to find the peak, extract the center wavelength and calculate the wavelength shift.

[0171] The coefficient optimization module is used to establish a polynomial fitting relationship for the wavelength offset, calculate the difference between the initial dispersion coefficient and the wavelength offset fitting coefficient, and output the optimized dispersion coefficient.

[0172] Therefore, this invention provides a method and system for optimizing the spectral dispersion coefficient of a greenhouse gas spectrometer before launch based on a radiative transfer model. According to instrument parameters and calibration requirements, the target calibration band and laboratory wavelength calibration accuracy requirements are determined. Based on the target instrument, the instrument linearity function (ILS), initial dispersion coefficient, and spectral data from direct solar observation under clear skies are obtained. Using the SCIATRAN radiative transfer model, the atmospheric absorption spectrum received at the ground is simulated. Based on the instrument spectral response function tested in the laboratory, the simulated spectrum is convolved with the instrument response function to obtain a simulated spectrum matching the instrument characteristics. Based on the sensitivity analysis results of atmospheric pressure, temperature, and water vapor environmental parameters, a position-robust reference absorber is selected from the simulated spectrum. Based on the asymmetric Gaussian fitting method, the observed spectrum and the simulated spectrum are fitted and peaked at the reference absorption line position to extract their respective center wavelengths. Based on the wavelength shift calculation results, a polynomial fitting is used to obtain the wavelength shift fitting coefficient. Based on the initial dispersion coefficient and the wavelength shift fitting coefficient, the calibrated dispersion coefficient is calculated, thereby optimizing the laboratory spectral dispersion coefficient. This allows the laboratory spectral dispersion coefficient optimization method provided by this invention to be independent of external high-precision FTS instruments, achieving picometer-level wavelength calibration accuracy. It effectively introduces the radiative transfer model into the spectral calibration at the laboratory stage, making up for the shortcomings of existing calibration schemes that rely on external equipment, and solving the wavelength drift problem caused by laser stability fluctuations and changes in environmental parameters.

[0173] In addition, reference absorption lines are selected based on the sensitivity analysis of multiple parameters such as air pressure, temperature and water vapor, thereby ensuring the robustness of the calibration benchmark, avoiding the influence of environmental parameter errors on the calibration results, and improving the reliability and accuracy of spectral dispersion coefficient optimization.

[0174] In another embodiment of the present invention, in addition to the SCIATRAN model, other high-precision radiative transfer models such as LBLRTM and MODTRAN can be used to simulate atmospheric absorption spectra, as long as they can accurately simulate the absorption line characteristics of molecules such as CO2 and O2.

[0175] In another embodiment of the present invention, in addition to symmetrical Gaussian fitting, methods such as super-Gaussian or Gaussian mixture line fitting can be used to extract the center wavelength of the absorption line. The optimal method should be selected according to the actual absorption line characteristics.

[0176] Example 4

[0177] This disclosure provides a non-volatile computer storage medium storing computer-executable instructions that can perform the steps described in the above embodiments.

[0178] It should be noted that the computer-readable medium described in this disclosure can be a computer-readable signal medium or a computer-readable storage medium, or any combination thereof. A computer-readable storage medium can be, for example,—but not limited to—an electrical, magnetic, optical, electromagnetic, infrared, or semiconductor system, apparatus, or device, or any combination thereof. More specific examples of a computer-readable storage medium may include, but are not limited to: an electrical connection having one or more wires, a portable computer disk, a hard disk, random access memory (RAM), read-only memory (ROM), erasable programmable read-only memory (EPROM or flash memory), optical fiber, portable compact disk read-only memory (CD-ROM), optical storage device, magnetic storage device, or any suitable combination thereof. In this disclosure, a computer-readable storage medium can be any tangible medium containing or storing a program that can be used by or in conjunction with an instruction execution system, apparatus, or device. In this disclosure, a computer-readable signal medium can include a data signal propagated in baseband or as part of a carrier wave, carrying computer-readable program code. Such propagated data signals can take various forms, including but not limited to electromagnetic signals, optical signals, or any suitable combination thereof. A computer-readable signal medium can be any computer-readable medium other than a computer-readable storage medium, which can send, propagate, or transmit a program for use by or in connection with an instruction execution system, apparatus, or device. The program code contained on the computer-readable medium can be transmitted using any suitable medium, including but not limited to: wires, optical fibers, RF (radio frequency), etc., or any suitable combination thereof.

[0179] The aforementioned computer-readable medium may be included in the aforementioned electronic device; or it may exist independently and not assembled into the electronic device.

[0180] Computer program code for performing the operations of this disclosure can be written in one or more programming languages ​​or a combination thereof, including object-oriented programming languages ​​such as Java, Smalltalk, and C++, and conventional procedural programming languages ​​such as the "C" language or similar programming languages. The program code can be executed entirely on the user's computer, partially on the user's computer, as a standalone software package, partially on the user's computer and partially on a remote computer, or entirely on a remote computer or server. In cases involving remote computers, the remote computer can be connected to the user's computer via any type of network—including a local area network (AN) or a wide area network (WAN)—or can be connected to an external computer (e.g., via the Internet using an Internet service provider).

[0181] The flowcharts and block diagrams in the accompanying drawings illustrate the architecture, functionality, and operation of possible implementations of systems, methods, and computer program products according to various embodiments of this disclosure. In this regard, each block in a flowchart or block diagram may represent a module, segment, or portion of code containing one or more executable instructions for implementing a specified logical function. It should also be noted that in some alternative implementations, the functions indicated in the blocks may occur in a different order than those indicated in the drawings. For example, two consecutively indicated blocks may actually be executed substantially in parallel, and they may sometimes be executed in reverse order, depending on the functions involved. It should also be noted that each block in the block diagrams and / or flowcharts, and combinations of blocks in the block diagrams and / or flowcharts, can be implemented using a dedicated hardware-based system that performs the specified function or operation, or using a combination of dedicated hardware and computer instructions.

[0182] The units described in the embodiments of this disclosure can be implemented in software or hardware. The names of the units are not, in some cases, intended to limit the specific unit.

[0183] The preferred embodiments of the present invention have been described above to make the spirit of the present invention clearer and easier to understand, and are not intended to limit the present invention. All modifications, substitutions, and improvements made within the spirit and principles of the present invention should be included within the protection scope summarized by the appended claims.

Claims

1. A method for optimizing the spectral dispersion coefficient of a greenhouse gas spectrometer before emission, characterized in that, Includes the following steps: Obtain the instrument linearity function, initial dispersion coefficient, and solar direct observation spectrum collected in the laboratory from the greenhouse gas spectrometer; Atmospheric absorption spectra are simulated based on a radiative transfer model, and the simulated spectra are convolved using the instrument's linear function to obtain simulated spectra that match the instrument's characteristics. Environmental parameter sensitivity analysis was performed on the simulated spectrum to screen out positionally robust reference absorption lines. Specifically, based on ±2hPa pressure uncertainty, pressure errors of -2hPa to +2hPa were applied to each layer of the pressure profile. The influence of the vertical pressure profile on the wavelength position of the simulated spectrum was analyzed, and absorption lines with an absolute value of center wavelength offset greater than 0.1% of the spectral resolution of the corresponding band were removed. Based on a ±2 K temperature error, a temperature error of -2 K to +2 K is applied to each layer of the temperature profile. The influence of the temperature profile on the wavelength position of the simulated spectrum is analyzed, and absorption lines with an absolute value of center wavelength offset greater than 0.1% of the spectral resolution of the corresponding band are removed. Based on a water vapor profile error of ±0.2 g / cm², the water vapor profile was scaled after applying a water vapor error of -0.2 g / cm² to +0.2 g / cm² to the real-time water vapor content of 2.07 g / cm² calculated in ERA5. The influence of water vapor content on the simulated spectral wavelength position was analyzed, and absorption lines with an absolute value of center wavelength offset greater than 0.1% of the spectral resolution of the corresponding band were removed. Absorption lines that are insensitive to changes in the three types of environmental parameters were retained as wavelength calibration benchmarks, and 17 reference absorption lines for the O2A band, 13 for the weak CO2 band, and 16 for the strong CO2 band were finally determined. The observed and simulated spectra are fitted to the reference absorption line positions to find peaks, the center wavelength is extracted, and the wavelength shift is calculated. Establish a polynomial fitting relationship for the wavelength offset to obtain the wavelength offset fitting coefficients; The initial dispersion coefficient and the wavelength shift fitting coefficient are compared by performing a difference operation, keeping the higher-order terms unchanged, and the optimized dispersion coefficient is output.

2. The method as described in claim 1, characterized in that, The acquisition of the initial dispersion coefficient of the greenhouse gas spectrometer includes: A tunable laser was used as a light source in conjunction with a wavelength meter to perform a full-image scan of a greenhouse gas spectrometer and obtain the instrument's linear function data. Based on the instrument's linear function data, a fifth-order polynomial relationship between the detector pixel number and the wavelength is established to obtain the initial dispersion coefficient.

3. The method as described in claim 1, characterized in that, The acquisition of the solar direct observation spectrum collected in the laboratory includes: A greenhouse gas spectrometer was placed inside a vacuum chamber to simulate an on-orbit environment, and sunlight was introduced to collect spectra. The collected spectral data were normalized, and spectral regions with high transmittance were selected as the observed spectra.

4. The method as described in claim 1, characterized in that, The simulation of atmospheric absorption spectra based on the radiative transfer model specifically includes: The TSIS-1 HSRS solar spectrum was used as the incident solar flux at the top of the atmosphere; Replace the atmospheric pressure, temperature, and water vapor vertical profiles in the basic atmospheric parameters with ERA5 reanalysis data. The pressure profile was scaled and corrected by combining measured surface pressure data, and the simulated geometric angles were set to be consistent with the actual acquisition state of the observed spectrum.

5. The method as described in claim 1, characterized in that, The process of fitting and peak-finding the observed and simulated spectra at the reference absorption line positions specifically includes: A fitted dataset is formed by selecting a preset number of data points centered on the valley value of the absorption line. The center wavelength of the absorption line is extracted by performing least-squares fitting on the fitted dataset using an asymmetric Gaussian function.

6. The method as described in claim 1, characterized in that, The establishment of the polynomial fitting relationship for the wavelength offset specifically includes: The optimal order of the wavelength offset polynomial fitting is determined using leave-one-out cross-validation (LOOCV). For the O2A band, weak CO2 band, and strong CO2 band, the relationship between wavelength offset and center wavelength is established by using optimal order fitting based on the LOOCV results.

7. The method as described in claim 1, characterized in that, The step of performing a difference calculation between the initial dispersion coefficient and the wavelength shift fitting coefficient specifically includes: Subtract the lower-order terms of the initial dispersion coefficient from the corresponding terms of the wavelength shift fitting coefficient; By keeping the value of the highest-order term of the initial dispersion coefficient unchanged, the optimized dispersion coefficient is obtained.

8. The application of the method according to any one of claims 1 to 7 in the pre-emission wavelength calibration of a greenhouse gas spectrometer, wherein the greenhouse gas spectrometer includes three detection channels: an O2A band, a weak CO2 band, and a strong CO2 band.

9. A system for optimizing the spectral dispersion coefficient of a greenhouse gas spectrometer before emission, based on the optimization method for the spectral dispersion coefficient of a greenhouse gas spectrometer according to any one of claims 1-7, comprising: The data acquisition module is used to acquire the instrument linear function, initial dispersion coefficient, and solar direct observation spectrum collected in the laboratory from the greenhouse gas spectrometer. The spectral simulation module is used to simulate atmospheric absorption spectra based on radiative transfer models and generate simulated spectra that match the instrument characteristics using instrument linearity functions. The benchmark screening module is used to perform environmental parameter sensitivity analysis on the simulated spectrum and screen out the position-robust reference absorption lines. The parameter extraction module is used to fit the observed spectrum and the simulated spectrum to the reference absorption line position to find the peak, extract the center wavelength and calculate the wavelength shift. The coefficient optimization module is used to establish a polynomial fitting relationship for the wavelength offset, calculate the difference between the initial dispersion coefficient and the wavelength offset fitting coefficient, and output the optimized dispersion coefficient.

Images