Method and device for constructing a model for correcting the direct effect of sunlight-induced chlorophyll fluorescence

Abstract

The application discloses a sunlight-induced chlorophyll fluorescence directionality effect correction model construction method and device, comprising: establishing a SIF anisotropic distribution equation by referring to a reflectivity BRDF model structure, and calibrating and locally verifying parameters by using ground-based multi-angle measured data; subsequently, acquiring global TROPOMI SIF remote sensing products, fitting global BRDF parameters under a 0.1*0.1 grid through a 16-day time window; and based on the fitting result, converting instantaneous observation SIF to standard observation modes such as a nadir, a hotspot and a hemisphere integral, and checking correlation by using global site GPP data. By adopting the technical scheme of the application, geometric deviation in SIF observation is effectively eliminated, vegetation physiological emission characteristics are restored, and consistency and precision of global vegetation productivity monitoring are significantly improved.

Description

Technical Field

[0001] This invention belongs to the field of vegetation remote sensing technology, specifically relating to a method and apparatus for constructing a correction model for the directionality effect of sunlight-induced chlorophyll fluorescence (SIF), used to correct the directional deviation of SIF signals under different observation geometry conditions. Background Technology

[0002] Solar-induced chlorophyll fluorescence (SIF), as an important tool for monitoring vegetation physiological processes and dynamic changes in photosynthesis, has gradually become a key data source in global carbon cycle research and Gross Primary Productivity (GPP) estimation. Because SIF has a direct mechanistic link with vegetation light response processes, it exhibits greater sensitivity and more clearly defined physiological significance in characterizing the instantaneous physiological state of vegetation compared to traditional vegetation indices. However, in practical applications using satellite and ground-based observations, SIF signals often exhibit significant directional effects due to the combined influence of the complexity of the three-dimensional structure of the vegetation canopy and changes in observation geometry.

[0003] This directional effect primarily stems from the spatial heterogeneity of the light environment within the canopy. As the observation angle changes, the proportion of signals contributed by sun-exposed and shaded leaves received by the sensor also alters. Since sun-exposed leaves receive significantly more direct solar radiation than shaded leaves, the difference in the visible proportion of sun-exposed and shaded leaves caused by changes in observation geometry directly leads to significant changes in the observed SIF intensity. Therefore, the observed SIF signal cannot be simply equated to the actual fluorescence emission level of the vegetation. Simultaneously, the SIF signal undergoes reabsorption and multiple scattering during its upward transmission within the canopy, and its escape probability is jointly controlled by the canopy structure and the observation direction, further exhibiting significant anisotropy.

[0004] Currently, although various satellite payloads can provide global-scale SIF datasets, these data are typically acquired at specific transit times and under specific observation geometries, lacking a unified directional standard and comparable benchmark. Due to the absence of a dedicated bidirectional reflection distribution function (BRDF) model capable of accurately characterizing the directional effects of SIF, existing SIF data struggles to be uniformly converted and standardized across different observation modes; for example, there is still a lack of consistent quantitative correlation between hotspot directional observations, nadir observations, and hemispherical integral observations. This systematic bias caused by directional differences severely limits the inversion accuracy of SIF in long-term GPP monitoring. Summary of the Invention

[0005] To address the problems existing in the prior art, this invention provides a method and apparatus for constructing a correction model for the directionality effect of sunlight-induced chlorophyll fluorescence (SIF). By establishing an anisotropic distribution model of SIF, the multi-angle observation signals are standardized, thereby reducing the impact of observational geometric differences and improving the reliability of remote sensing monitoring of vegetation physiological status and GPP estimation.

[0006] To achieve the above objectives, the present invention provides the following solution: A method for constructing a correction model for the directionality effect of sunlight-induced chlorophyll fluorescence includes: Step S1: Establish the SIF anisotropy expression equation through the reflectance BRDF model structure; Step S2: Use multi-angle measured data of the foundation to calibrate the parameters of the SIF spatial anisotropy expression equation, and verify the local accuracy of the model by comparing the goodness of fit between the simulated values ​​and the measured values. Step S3: Acquire global satellite SIF remote sensing products and combine them with contemporaneous radiometric data to invert BRDF weight parameters globally within a preset spatial grid scale and time sliding window. Step S4: Based on the weight parameters obtained by inversion, convert the original SIF value under the instantaneous satellite observation geometry to the estimated value under the standard observation mode; Step S5: Verify the feasibility of the SIF estimate under the standard observation mode with the measured vegetation productivity (GPP) data from the flux station.

[0007] Preferably, in step S1, the SIF signal under different observation geometry conditions is deconstructed into a weighted combination of isotropic components, volume scattering components, and geometric optical scattering components.

[0008] Preferably, in step S2, multi-angle SIF measured data of a long-term series are obtained using a ground-based automated observation platform, and the parameters of the SIF spatial anisotropy expression equation are calibrated.

[0009] As a preferred option, in step S3, global TROPOMI satellite SIF remote sensing products are acquired and combined with ERA5 solar radiation data. Using a 16-day time window and a spatial grid scale of 0.1°×0.1°, the least squares method is used to invert the BRDF parameters globally.

[0010] Preferably, the standard observation modes include nadir observation mode, hotspot observation mode, and hemispherical integral observation mode.

[0011] This invention also provides a device for constructing a correction model for the directionality effect of sunlight-induced chlorophyll fluorescence, comprising: The first processing module is used to establish the SIF anisotropy expression equation through the reflectivity BRDF model structure; The second processing module is used to calibrate the parameters of the SIF spatial anisotropy expression equation using multi-angle measured data of the foundation, and to verify the local accuracy of the model by comparing the goodness of fit between the simulated values ​​and the measured values. The third processing module is used to acquire global satellite SIF remote sensing products and combine them with contemporaneous radiometric data to invert BRDF weight parameters globally within a preset spatial grid scale and time sliding window. The fourth processing module is used to convert the original SIF values ​​under the instantaneous satellite observation geometry to the estimated values ​​under the standard observation mode based on the weight parameters obtained by inversion. The fifth processing module is used to verify the feasibility of the SIF estimate under the standard observation mode with the measured vegetation productivity (GPP) data from the flux station.

[0012] Preferably, the first processing module decomposes the SIF signal under different observation geometry conditions into a weighted combination of isotropic components, volume scattering components, and geometric optical scattering components.

[0013] Preferably, the second processing module uses a ground-based automated observation platform to acquire long-term series multi-angle SIF measured data and calibrates the parameters of the SIF spatial anisotropy expression equation.

[0014] As a preferred option, the third processing module acquires global TROPOMI satellite SIF remote sensing products and combines them with ERA5 solar radiation data. Using a 16-day time window and a 0.1°×0.1° spatial grid scale, it employs the least squares method to invert BRDF parameters globally.

[0015] Preferably, the standard observation modes include nadir observation mode, hotspot observation mode, and hemispherical integral observation mode.

[0016] Compared with the prior art, the beneficial effects of the present invention are as follows: 1. High accuracy in tower base verification and strong model reliability: This invention verifies the SIFBRDF model, constructed by drawing on the reflectivity approach, through multi-angle measured data from the tower base, demonstrating extremely high simulation accuracy. Taking the corn canopy as an example, the goodness of fit R² between the simulated and measured values ​​reaches 0.90, ensuring the correctness of the model in terms of basic physical logic.

[0017] 2. Significantly improves the accuracy and consistency of global GPP monitoring: According to the global-scale fitting verification of this invention, the SIF estimated under the three standard observation modes of Nadir, Hotspot and Hemis has a better correlation with the actual GPP measured at the site than the original observation data without BRDF correction.

[0018] 3. Achieved cross-platform unification of multi-source SIF data: By converting instantaneous observation signals to a standard geometric mode, the problem of signal incomparability caused by differences in observation geometry between different satellite sensors and at different transit times was solved, providing technical support for building a globally unified standard SIF long-term sequence dataset. Attached Figure Description

[0019] To more clearly illustrate the technical solution of the present invention, the drawings used in the embodiments are briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0020] Figure 1 This is a flowchart illustrating the method for constructing a correction model for the directionality effect of sunlight-induced chlorophyll fluorescence according to an embodiment of the present invention. Figure 2 This is a comparison chart of the multi-angle SIF of corn simulated based on the RTLT model and the measured SIF in an embodiment of the present invention; Figure 3 This is a schematic diagram of the global spatial distribution of parameters obtained by inversion based on the RPV model in an embodiment of the present invention; Figure 4 This is a comparison chart of the correlation between satellite SIF and flux station GPP under different observation conditions based on the BRDF model in this embodiment of the invention. Detailed Implementation

[0021] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0022] To make the above-mentioned objects, features and advantages of the present invention more apparent and understandable, the present invention will be further described in detail below with reference to the accompanying drawings and specific embodiments.

[0023] Example 1 like Figure 1 As shown, this invention provides a method for constructing a correction model for the directionality effect of sunlight-induced chlorophyll fluorescence, comprising: Step S1: Establish the SIF anisotropy expression equation through the reflectance BRDF model structure; Step S2: Use multi-angle measured data of the foundation to calibrate the parameters of the SIF spatial anisotropy expression equation, and verify the local accuracy of the model by comparing the goodness of fit between the simulated values ​​and the measured values. Step S3: Acquire global satellite SIF remote sensing products and combine them with contemporaneous radiometric data to invert BRDF weight parameters globally within a preset spatial grid scale and time sliding window. Step S4: Based on the weight parameters obtained by inversion, convert the original SIF value under the instantaneous satellite observation geometry to the estimated value under the standard observation mode; Step S5: Verify the feasibility of the SIF estimate under the standard observation mode with the measured vegetation productivity (GPP) data from the flux station.

[0024] In one embodiment of the present invention, in step S1, observation data at different scales are fitted to a semi-empirical or physical model to extract core coefficients describing the directional distribution of SIF. Specifically, the present invention introduces the RossThick-LiTransit (RTLT) kernel-driven model and the Rahman-Pinty-Verstraete (RPV) model for comparative fitting.

[0025] Within the RTLT model framework, the directional distribution of SIF is expressed as a linear combination of isotropic scattering kernels, volume scattering kernels, and geometric optics scattering kernels. The core task is to solve for the weighting coefficients corresponding to these three scattering kernels (respectively...). , and The model formula is shown below: ; in, , and These are the solar zenith angle, the observed zenith angle, and the relative azimuth angle, respectively. For short Wave incident radiation; and These are respectively volume scattering kernels and geometric scattering kernels; , and The weight coefficients to be solved are denoted as . It can be calculated using the following formula: ; in, This is the phase angle. It is calculated by the following formula: ; ; Here, h / b and b / r are set to 2 and 1, respectively.

[0026] By constructing a target cost function and using the linear least squares method to iteratively optimize the multi-angle observation sequence, the sum of squared residuals between the model simulation values ​​and the actual observation values ​​is minimized, thereby accurately obtaining the kernel coefficient dataset that characterizes the anisotropy of the vegetation canopy SIF.

[0027] Furthermore, for vegetation types with more pronounced nonlinear characteristics, this embodiment of the invention simultaneously employs the RPV model for parameter inversion. Unlike the linear kernel-driven model, the nonlinear RPV model characterizes the spatial distribution of SIF using four parameters (total scattering intensity ρ0, anisotropic intensity k, forward and backward scattering intensity Θ, and hotspot effect ρc). The model formula is shown below: ; in, It is shortwave incident radiation; (Right now ) represents the amplitude of the fluorescent BRDF; , and These are used to describe the overall shape, asymmetry, and hotspot effect of the BRDF, and the relevant calculation formulas are as follows: ; Among them, parameters (0-1) is used to describe the bowl-shaped shape of the BRDF; (0-1) is used to describe the hotspot effect; the smaller the value, the more pronounced the hotspot effect. Furthermore, formula (12) can be simplified as follows: ; Among them, parameters Used to describe the asymmetry of BRDF.

[0028] In this process, embodiments of the present invention employ the Gauss-Newton method or the Levenberg-Marquardt (LM) nonlinear optimization algorithm to substitute multi-angle scan data into the RPV equation to search for the optimal solution of parameters.

[0029] As one embodiment of the present invention, in step S2, before the model is applied globally, the present invention first conducts quantitative verification of the model's fitting ability at a local scale, aiming to establish the physical reliability of the SIF BRDF model through high-precision ground-based measured data. This verification process relies on long-term, multi-angle SIF datasets acquired by a ground-based automated spectroscopic observation platform. In the specific execution process, the system collects observation samples covering the entire vegetation growth cycle and various typical weather conditions, and specifically extracts observation sequences containing key geometric features such as hotspot direction, forward scattering, and backscattering.

[0030] In the parameter fitting stage, this embodiment of the invention uses instantaneous solar geometric information (solar zenith angle, azimuth angle) recorded by the ground-based observation system and sensor observation geometric information as input, and substitutes them into the aforementioned constructed SIF BRDF core equation. Using linear least squares or other nonlinear optimization algorithms, the kernel weight coefficients within the observation grid are iteratively solved to obtain the parameter combination that optimally characterizes the current spatial distribution features of the vegetation canopy. The calibration process strictly follows physical constraints to ensure that the weights of each component not only achieve optimal mathematical fitting but also physically conform to the contribution patterns of yin-yang leaves within the canopy.

[0031] To quantitatively evaluate the model's performance, this embodiment of the invention performs a 1:1 regression analysis between the calibrated BRDF model's simulated values ​​and the measured SIF values ​​of the foundation. Experimental results ( Figure 2 The results show that, taking the maize canopy as an example, the SIF BRDF model constructed based on the RTLT framework exhibits excellent simulation stability, with its simulated values ​​showing a high degree of consistency with the measured values, and the coefficient of determination R0 is high. 2 It remains stable at around 0.90, and the root mean square error (RMSE) remains at an extremely low level.

[0032] As one embodiment of the present invention, in step S3, after completing the mechanism verification at the base scale, the present invention further applies the model to the SIF data standardization processing at the global satellite scale. First, TROPOMI satellite SIF remote sensing products from around the world are acquired as the core input source. This payload possesses high spatial resolution and multi-angle observation characteristics, providing a rich data foundation for BRDF parameter inversion. Simultaneously, ERA5 surface solar radiation downward (SSRD) data of the same phase are acquired to normalize the photosynthetically active radiation of the SIF signal, thereby eliminating interference caused by incident light fluctuations in directional effect calibration.

[0033] To capture seasonal changes in vegetation phenology while ensuring computational efficiency, this embodiment of the invention employs a 16-day sliding time window strategy and establishes a global 0.1° × 0.1° high-resolution geospatial grid. Within each grid cell, the system automatically retrieves all valid satellite observation samples within the window period and extracts the corresponding observation zenith angle, solar zenith angle, and relative azimuth angle information. By constructing a large-scale matrix equation, the BRDF kernel weight parameters within each grid are solved independently using the least squares method. During parameter fitting, this embodiment of the invention introduces spatial continuity constraints and robust estimation algorithms to effectively eliminate anomalous values ​​caused by cloud remnants or surface heterogeneity. Specifically, within each grid cell, this embodiment of the invention constructs an observation vector from the n instantaneous satellite observation SIF values ​​acquired within the window period. And based on the corresponding observed geometric parameters, the n×m design matrix is ​​calculated. (Where m is the number of kernel functions in the BRDF model). To ensure the robustness and spatial consistency of parameter inversion, this invention constructs an objective function that includes weighted robust estimation and spatial constraints. Solve the following: ; In the above objective function, As a diagonal weight matrix, a lower weight is assigned to observation samples with large residuals through an iterative update mechanism, thereby effectively removing residual cloud and surface heterogeneous noise. For regularization terms, where As a regularization factor, For the prior parameter weights obtained based on the spatiotemporal neighborhood, this term introduces a spatial continuity constraint to ensure that the retrieved parameter weight field has good smoothness and physical coherence in its spatial distribution. Finally, by adjusting the objective function... By taking the partial derivative and setting it to zero, the optimal weight parameter set for this grid is obtained analytically using the least squares method. Through the above calculations, this embodiment of the invention finally constructs a globally covered, spatiotemporally continuous SIF anisotropic parameter dataset. This dataset uses 0.1° as a spatial reference and accurately records the isotropic scattering, volume scattering, and geometric optical scattering characteristics (e.g., ...) of different vegetation types worldwide over specific time periods. Figure 3 (As shown).

[0034] In one embodiment of the present invention, in step S4, after obtaining the global BRDF model weight parameter set, the present invention establishes a unified geometric reference benchmark through mathematical transformation, converting the instantaneous observation signal with anisotropic characteristics into a SIF estimate under a standardized mode. Specifically, the specific calculation process for each standard observation mode is as follows: 1. Nadir observation mode: Set the observation zenith angle relative azimuth All are 0, solar zenith angle Keeping the original observations, the geometric parameters mentioned above are substituted into the aforementioned physical model formula to calculate the estimated SIF value under the nadir model. Nadir .

[0035] 2. Hotspot observation mode: Set the observation zenith angle. Equal to the solar zenith angle And relative azimuth angle Setting the value to 0 ensures that the observation direction perfectly coincides with the solar incidence direction. Substituting this geometric configuration into the physical model formula, the estimated SIF value under the hotspot mode is calculated. Hotspot .

[0036] 3. Hemispherical Integral Observation Mode: This mode calculates the total observed fluorescence intensity emitted from the canopy by integrating the observation angles over the entire hemispherical spatial range. The calculation process follows the integral equation below: ; in, SIF model This represents the aforementioned RTLT or RPV model function.

[0037] Because the observation zenith angle and relative azimuth angle of satellite payloads such as Tropomi are dynamically changing during transit, the original SIF products contain complex directional noise. To eliminate this geometric bias, this embodiment of the invention utilizes isotropic weights, volume scattering weights, and geometric optical weights obtained through inversion to resimulate the SIF intensity under specific standard geometric conditions, thereby achieving normalization and comparability of multi-source remote sensing data across spatiotemporal scales.

[0038] In the specific operation of this invention embodiment, three standard observation modes with typical biophysical significance are constructed: First, the Nadir observation mode, which fixes the observation zenith angle and relative azimuth angle in the model to 0° to simulate the fluorescence intensity of the vertically downward field of view, which plays an important role in unifying the satellite overpass field of view; second, the Hotspot observation mode, which calculates the SIF intensity when the contribution of the canopy light-receiving leaves (sun-receiving leaves) is maximized by setting the observation geometry to completely coincide with the solar incidence geometry (i.e., the observation zenith angle is equal to the solar zenith angle and the relative azimuth angle is 0°), which can effectively characterize the maximum photosynthetic potential of vegetation; and finally, the Hemispherical integral mode, which calculates the total observed fluorescence intensity emitted outward by the canopy by integrating the observation angles of the entire hemispherical space. This index eliminates the limitations of all single-angle observations and best represents the overall physiological emission level of vegetation.

[0039] In one embodiment of the present invention, step S5 evaluates the effectiveness of the BRDF correction method in practical applications using vegetation physiological indicators. The verification process uses the representative Total Primary Productivity (GPP) from global carbon cycle research as a reference benchmark. During data processing, this embodiment of the present invention collects measured GPP data from stations covering typical ecosystems in Ameriflux (including information from stations with complex canopy structures such as ENF-evergreen coniferous forests and DBF-deciduous broadleaf forests). To ensure the temporal compatibility between satellite remote sensing observations and ground station data, this embodiment of the present invention introduces a daily correction factor to extend the temporal scale of instantaneous SIF observations while performing spatiotemporal registration. This processing step aims to reduce instantaneous signal fluctuations caused by differences in solar altitude angle at sampling times, thereby physically and logically enhancing the comparability between daily-scale SIF and daily total GPP.

[0040] In the comparative analysis, this invention established linear regression relationships between the original SIF observation sequences before correction and the SIF sequences obtained after correction using the method in Nadir, Hotspot, and Hemispherical modes, and synchronous GPP data. As shown in Figure 4, the experimental results show that the original SIF signal contains more anisotropic biases caused by observation geometry, resulting in a relatively discrete scatter distribution with GPP. In contrast, the SIF data under the three standard observation modes after BRDF model correction exhibit higher consistency in their linear fitting relationship with GPP, and the data dispersion is reduced.

[0041] The conclusion analysis shows that, regardless of whether it is the nadir model from a vertical perspective, the hotspot model from a specific geometric angle, or the hemispherical integral model representing total characteristics, the corrected regression determination coefficient (R²) remains consistent. 2 The results show varying degrees of improvement compared to the original observations, and the root mean square error (RMSE) exhibits a decreasing trend. This confirms that the SIF BRDF model constructed in this invention can correct observational biases caused by changes in the geometric relationship between the sun, target, and sensor at a physical level, enabling the corrected signal to more objectively reflect the physiological emission characteristics of vegetation. This standardized transformation process enhances the consistency and reliability of SIF data in monitoring vegetation productivity on a global scale, providing standardized data support for subsequent related scientific research. Example 2 This invention also provides a device for constructing a correction model for the directionality effect of sunlight-induced chlorophyll fluorescence, comprising: The first processing module is used to establish the SIF anisotropy expression equation through the reflectivity BRDF model structure; The second processing module is used to calibrate the parameters of the SIF spatial anisotropy expression equation using multi-angle measured data of the foundation, and to verify the local accuracy of the model by comparing the goodness of fit between the simulated values ​​and the measured values. The third processing module is used to acquire global satellite SIF remote sensing products and combine them with contemporaneous radiometric data to invert BRDF weight parameters globally within a preset spatial grid scale and time sliding window. The fourth processing module is used to convert the original SIF values ​​under the instantaneous satellite observation geometry to the estimated values ​​under the standard observation mode based on the weight parameters obtained by inversion. The fifth processing module is used to verify the feasibility of the SIF estimate under the standard observation mode with the measured vegetation productivity (GPP) data from the flux station.

[0042] As one embodiment of the present invention, the first processing module deconstructs the SIF signal under different observation geometry conditions into a weighted combination of isotropic components, volume scattering components, and geometric optical scattering components.

[0043] As one embodiment of the present invention, the second processing module uses a ground-based automated observation platform to acquire multi-angle SIF measured data of a long-term series and calibrates the parameters of the SIF spatial anisotropy expression equation.

[0044] As one embodiment of the present invention, the third processing module acquires global TROPOMI satellite SIF remote sensing products and combines them with ERA5 solar radiation data. Using a 16-day time window and a spatial grid scale of 0.1°×0.1°, the least squares method is used to invert the BRDF parameters globally.

[0045] As one embodiment of the present invention, the standard observation modes include nadir observation mode, hotspot observation mode, and hemispherical integral observation mode.

[0046] The embodiments described above are merely preferred embodiments of the present invention and are not intended to limit the scope of the present invention. Various modifications and improvements made to the technical solutions of the present invention by those skilled in the art without departing from the spirit of the present invention should fall within the protection scope defined by the claims of the present invention.

Claims

1. A method for constructing a correction model for the directionality effect of sunlight-induced chlorophyll fluorescence, characterized in that, include: Step S1: Establish the SIF anisotropy expression equation through the reflectance BRDF model structure; Step S2: Use multi-angle measured data of the foundation to calibrate the parameters of the SIF spatial anisotropy expression equation, and verify the local accuracy of the model by comparing the goodness of fit between the simulated values ​​and the measured values. Step S3: Acquire global satellite SIF remote sensing products and combine them with contemporaneous radiometric data to invert BRDF weight parameters globally within a preset spatial grid scale and time sliding window. Step S4: Based on the weight parameters obtained by inversion, convert the original SIF value under the instantaneous satellite observation geometry to the estimated value under the standard observation mode; Step S5: Verify the feasibility of the SIF estimate under the standard observation mode with the measured vegetation productivity (GPP) data from the flux station.

2. The method for constructing a correction model for the directionality effect of sunlight-induced chlorophyll fluorescence as described in claim 1, characterized in that, In step S1, the SIF signal under different observation geometry conditions is deconstructed into a weighted combination of isotropic components, volume scattering components, and geometric optical scattering components.

3. The method for constructing a correction model for the directionality effect of sunlight-induced chlorophyll fluorescence as described in claim 2, characterized in that, In step S2, the ground-based automated observation platform is used to acquire long-term series multi-angle SIF measured data, and the parameters of the SIF spatial anisotropy expression equation are calibrated.

4. The method for constructing a correction model for the directionality effect of sunlight-induced chlorophyll fluorescence as described in claim 3, characterized in that, In step S3, global TROPOMI satellite SIF remote sensing products are acquired and combined with ERA5 solar radiation data. Using a 16-day time window and a spatial grid scale of 0.1°×0.1°, the least squares method is used to invert the BRDF parameters globally.

5. The method for constructing a correction model for the directionality effect of sunlight-induced chlorophyll fluorescence as described in claim 4, characterized in that, The standard observation modes include nadir observation mode, hotspot observation mode, and hemispherical integral observation mode.

6. A device for constructing a correction model for the directionality effect of sunlight-induced chlorophyll fluorescence, characterized in that, include: The first processing module is used to establish the SIF anisotropy expression equation through the reflectivity BRDF model structure; The second processing module is used to calibrate the parameters of the SIF spatial anisotropy expression equation using multi-angle measured data of the foundation, and to verify the local accuracy of the model by comparing the goodness of fit between the simulated values ​​and the measured values. The third processing module is used to acquire global satellite SIF remote sensing products and combine them with contemporaneous radiometric data to invert BRDF weight parameters globally within a preset spatial grid scale and time sliding window. The fourth processing module is used to convert the original SIF values ​​under the instantaneous satellite observation geometry to the estimated values ​​under the standard observation mode based on the weight parameters obtained by inversion. The fifth processing module is used to verify the feasibility of the SIF estimate under the standard observation mode with the measured vegetation productivity (GPP) data from the flux station.

7. The apparatus for constructing a correction model for the directionality effect of sunlight-induced chlorophyll fluorescence as described in claim 6, characterized in that, The first processing module decomposes the SIF signal under different observation geometry conditions into a weighted combination of isotropic components, volume scattering components, and geometric optical scattering components.

8. The apparatus for constructing a correction model for the directionality effect of sunlight-induced chlorophyll fluorescence as described in claim 7, characterized in that, The second processing module uses a ground-based automated observation platform to acquire long-term series multi-angle SIF measured data and calibrates the parameters of the SIF spatial anisotropy expression equation.

9. The apparatus for constructing a correction model for the directionality effect of sunlight-induced chlorophyll fluorescence as described in claim 8, characterized in that, The third processing module acquires global TROPOMI satellite SIF remote sensing products and combines them with ERA5 solar radiation data. Using a 16-day time window and a 0.1°×0.1° spatial grid scale, it uses the least squares method to invert BRDF parameters globally.

10. The apparatus for constructing a correction model for the directionality effect of sunlight-induced chlorophyll fluorescence as described in claim 9, characterized in that, The standard observation modes include nadir observation mode, hotspot observation mode, and hemispherical integral observation mode.

Images