Carbon flux metering inversion method and system for Saggoke new energy base

Abstract

The invention provides a carbon flux metering inversion method and system for a Saggohm new energy base, and the method comprises the steps: obtaining the physical parameters, topographic data and image data of the Saggohm new energy base, and building a three-dimensional geometric model through a three-dimensional reconstruction technology; performing space grid processing on the model area to generate an LBM computational grid; inputting the image data into the trained ground parameter model, and outputting a ground boundary parameter and a ground source item parameter of each grid unit; cloud layer information is monitored through an all-sky imager, and ground irradiation intensity distribution is calculated by adopting a Monte Carlo method in combination with a sun trajectory tracking model; acquiring an atmosphere input boundary parameter and a source item parameter; solving the LBM model to obtain carbon dioxide concentration, flow velocity, temperature and pollutant concentration distribution; and the carbon flux of the base is calculated based on the simulation result, so that high-precision, low-cost and spatially continuous inversion calculation of the large-range carbon flux of the Sagomean region is realized.

Description

Technical Field

[0001] This invention relates to the field of environmental monitoring technology, and in particular to a method and system for measuring and retrieving carbon flux in the Shagohuang New Energy Base. Background Technology

[0002] For new energy bases built in special environments such as deserts, Gobi, and wastelands, carbon flux monitoring is of great significance for assessing the environmental impact of the project.

[0003] Traditional carbon flux monitoring methods mainly rely on direct measurements from ground-based flux towers. This approach faces several challenges in the desert region: First, the desert region is vast, and deploying flux towers across the entire area would result in excessively high construction costs; second, due to the complexity of the terrain and the difficulty of equipment maintenance, the density of flux towers is often insufficient, making it difficult to accurately reflect the spatial distribution characteristics of carbon flux over a large area and ensuring monitoring accuracy.

[0004] Therefore, how to reduce costs while ensuring monitoring accuracy has become a core technical problem that urgently needs to be solved. Summary of the Invention

[0005] This invention provides a carbon flux measurement and inversion method and system for the Shagohuang New Energy Base, which solves the defects of high carbon flux detection cost and low monitoring accuracy in the existing technology.

[0006] On the one hand, this invention provides a method for carbon flux measurement and inversion in a desert renewable energy base, which includes: The physical parameters, terrain data, and image data of the Shagohuang New Energy Base were obtained, and a three-dimensional geometric model of the Shagohuang New Energy Base was established using three-dimensional reconstruction technology. The region represented by the three-dimensional geometric model is spatially latticized to generate the computational grid required for the LBM model; The image data is input into a pre-trained ground parameter model for identification, and the ground boundary parameters and ground source term parameters corresponding to each grid cell in the computational grid are output to set the ground boundary conditions and ground material source terms of the LBM model. By monitoring cloud information with an all-sky imager and combining it with a high-precision solar trajectory tracking model, the Monte Carlo method is used to calculate the ground irradiance distribution under cloud cover conditions, so as to set the ground thermal radiation source term of the LBM model. The input boundary parameters and input source term parameters of the atmosphere within the monitoring area are obtained in order to set the atmospheric boundary conditions and atmospheric material source terms of the LBM model. solving the LBM model, obtaining detailed simulation results of carbon dioxide concentration distribution, flow velocity distribution, temperature distribution and pollutant concentration distribution in the Shagehuang new energy base, and calculating the carbon flux of the Shagehuang new energy base based on the detailed simulation results.

[0007] In another aspect, the present application also provides a Shagehuang new energy base carbon flux measurement inversion system, which comprises: A construction module is configured to obtain physical parameters, terrain data and image data of the Shagehuang new energy base, so as to establish a three-dimensional geometric model of the Shagehuang new energy base by using a three-dimensional reconstruction technology; A generation module is configured to perform spatial lattice processing on a region represented by the three-dimensional geometric model, so as to generate a calculation grid required by the LBM model; A first setting module is configured to input the image data into a pre-trained ground parameter model for identification, and output ground boundary parameters and ground source term parameters corresponding to each grid element in the calculation grid, so as to set a ground boundary condition of the LBM model and a ground material source term of the LBM model; A second setting module is configured to monitor cloud information by using an all-sky imager, combine a high-precision sun trajectory tracking model, and calculate a ground irradiance intensity distribution under a cloud layer shielding condition by using a Monte Carlo method, so as to set a ground thermal radiation source term of the LBM model; A third setting module is configured to obtain input boundary parameters and input source term parameters of the atmosphere in the monitoring region, so as to set an atmospheric boundary condition of the LBM model and an atmospheric material source term of the LBM model; A solving module is configured to solve the LBM model, obtain detailed simulation results of carbon dioxide concentration distribution, flow velocity distribution, temperature distribution and pollutant concentration distribution in the Shagehuang new energy base, and calculate the carbon flux of the Shagehuang new energy base based on the detailed simulation results.

[0008] The Shagehuang new energy base carbon flux measurement inversion method and system provided by the present application use remote sensing images and a three-dimensional reconstruction technology to construct a fine geometric model of the base, automatically identify the surface characteristics and set the boundary conditions by using a ground parameter model, accurately simulate the actual irradiance distribution by combining all-sky cloud monitoring and a Monte Carlo radiation transfer model, finally solve the atmospheric transmission process of the multi-physical field coupling by using the LBM, output the high-resolution spatial distribution results of carbon dioxide concentration, flow velocity, temperature and pollutants, and realize the high-precision, low-cost and spatially continuous inversion calculation of the carbon flux in the Shagehuang region. The present application effectively overcomes the limitations of high deployment cost and poor spatial representativeness of the traditional flux tower, and significantly improves the accuracy and economy of the carbon flux monitoring in the Shagehuang environment. BRIEF DESCRIPTION OF DRAWINGS

[0009] In order to more clearly illustrate the technical solutions in the present application or the prior art, the following will briefly introduce the drawings needed to be used in the embodiments or prior art description. Obviously, the drawings in the following description are some embodiments of the present application, and other drawings can also be obtained by those skilled in the art without creative effort on the basis of these drawings.

[0010] Figure 1 is a flowchart of a carbon flux measurement inversion method for a Shagehuang new energy base provided by an embodiment of the present application. Figure 2 is a structural schematic diagram of a carbon flux measurement inversion system for a Shagehuang new energy base provided by an embodiment of the present application. DETAILED DESCRIPTION

[0011] In order to make the objects, technical solutions and advantages of the present application clearer, the technical solutions in the present application will be described clearly and completely in the following with reference to the drawings in the present application. Obviously, the described embodiments are some embodiments of the present application, rather than all the embodiments. Based on the embodiments in the present application, all other embodiments obtained by those skilled in the art without creative effort belong to the protection scope of the present application.

[0012] In the prior art, carbon flux monitoring of new energy bases constructed in special environments such as deserts, Gobi and deserts mainly relies on direct measurement of ground flux towers. The area of such regions is vast, and the overall deployment of flux towers leads to high construction cost; the complexity of the terrain and the difficulty of equipment maintenance limit the deployment density of the flux towers, and it is difficult to accurately reflect the spatial distribution characteristics of the carbon flux in a large area, and the monitoring accuracy is limited. The traditional method cannot balance the cost and accuracy, and an efficient and economical solution is urgently needed.

[0013] In order to solve the above problems, it is found through research that it is necessary to break through the limitation of single reliance on physical monitoring equipment and instead combine numerical simulation with limited measured data. Through analysis, it is found that the carbon flux in Shagehuang area is affected by multiple factors such as surface features, atmospheric flow and thermal radiation, and a high-precision three-dimensional model needs to be constructed to reflect the complex environment. Further considering using the lattice Boltzmann method (LBM) to simulate the flow field and material transport, combining with remote sensing data to dynamically correct the model parameters, and finally forming a low-cost monitoring system based on simulation inversion.

[0014] Therefore, the present application proposes the following technical solutions: Figure 1 is a flowchart of a carbon flux measurement inversion method for a Shagehuang new energy base provided by an embodiment of the present application.

[0015] As Figure 1As shown, the execution subject of the carbon flux measurement and inversion method of the Shagehuang new energy base provided by the embodiments of the present application can be an electronic device, and the method mainly comprises the following steps: 101. Obtain the physical parameters, terrain data and image data of the Shagehuang new energy base to establish a three-dimensional geometric model of the Shagehuang new energy base by using a three-dimensional reconstruction technology; 102. Perform spatial lattice processing on the area represented by the three-dimensional geometric model to generate a calculation grid required by an LBM model; 103. Input the image data into a pre-trained ground parameter model for identification, output ground boundary parameters and ground source term parameters corresponding to each grid element in the calculation grid, and set the ground boundary conditions of the LBM model and the ground material source term of the LBM model; 104. Monitor cloud information by using an all-sky imager, combine a high-precision sun trajectory tracking model, and calculate the ground irradiance intensity distribution under the condition of cloud layer shielding by using a Monte Carlo method to set the ground thermal radiation source term of the LBM model; 105. Obtain the input boundary parameters and input source term parameters of the atmosphere in the monitoring area to set the atmospheric boundary conditions of the LBM model and the atmospheric material source term of the LBM model; 106. Solve the LBM model to obtain detailed simulation results of the carbon dioxide concentration distribution, flow velocity distribution, temperature distribution and pollutant concentration distribution in the Shagehuang new energy base, and calculate the carbon flux of the Shagehuang new energy base based on the detailed simulation results.

[0016] In a specific implementation process, the three-dimensional geometric model refers to a three-dimensional digital model that integrates terrain elevation, facility geometric shape and surface cover, and can be generated by using a Poisson surface reconstruction algorithm after registering unmanned aerial vehicle aerial point cloud data and satellite multispectral image, and is used to accurately represent surface features such as wind turbine blade height and photovoltaic panel inclination. The LBM calculation grid refers to a grid system that discretizes the simulation area into regular cubic elements, and can specifically use an octree adaptive grid division method to encrypt the grid to a meter-level resolution in a facility-dense area to ensure the capture accuracy of the wake vortex. The ground parameter model refers to a multi-label classification model based on deep learning, and can specifically be trained by using a ResNet-50 architecture, the input being a slice of remote sensing image or unmanned aerial vehicle aerial image matched with the grid size, and the output being the ground roughness, thermal conductivity and material source term parameters of each grid, thereby solving the problem of low efficiency of traditional manual calibration. The Monte Carlo ray tracing algorithm refers to a probability model that simulates the interaction between sunlight and cloud layers, and can specifically use a Buie solar intensity distribution model combined with cloud transmittance inversion data to calculate the irradiance intensity in the shadow area by sampling one million rays, thereby eliminating the interference of cloud layer dynamic shielding on thermal radiation calculation.

[0017] Specifically, first, the unmanned aerial vehicle is equipped with a laser radar to obtain centimeter-level terrain point cloud data, and the surface cover type is extracted in combination with a Sentinel-2 satellite image to construct a three-dimensional model containing a photovoltaic panel inclination and a fan tower height. The simulation area is divided into a cubic grid with adjustable side length, and a 0.5-meter fine grid is used in the edge area of the photovoltaic array to capture the change of the turbulent boundary layer. The grid corresponding remote sensing image slice is input into a pre-trained ResNet-50 model to automatically identify the area ratio of rock, vegetation, concrete and other materials in the grid, and the roughness parameter of each grid is calculated by weighting. The cloud distribution map is collected by the all-sky imager every 5 minutes, and the direct and scattered radiation ratio of the photovoltaic panel surface is calculated in real time in combination with the solar elevation angle, the Monte Carlo method is used to simulate 100,000 light paths, and the effective radiation energy received by each grid is counted as a heat source term. Finally, the LBM equation is solved by coupling the atmospheric boundary condition to output the hourly updated carbon dioxide concentration field and three-dimensional flow field data to provide spatially continuous input parameters for carbon flux calculation.

[0018] The scheme realizes the space-time continuous simulation of the whole area through the LBM model, and quantifies the actual influence of the cloud shape on the irradiance intensity through the ray tracing algorithm. In addition, the scheme also improves the parameter identification efficiency through the fusion processing of the unmanned aerial vehicle and the satellite image.

[0019] Through the above technical scheme, the present application realizes high-precision inversion of the carbon flux of Shaguo barren base, only a small number of flux towers are needed to be deployed for model verification, and the monitoring cost is significantly reduced. The fine construction of the three-dimensional model effectively captures the influence of new energy facilities on the local microenvironment, the Monte Carlo irradiation calculation improves the simulation accuracy under cloudy weather, and the automatic parameter identification of multi-source data fusion ensures the space-time consistency of the model input, which finally effectively reduces the error of the global carbon flux calculation. That is, the present application effectively overcomes the limitations of high deployment cost and poor spatial representativeness of traditional flux towers, and significantly improves the precision and economy of carbon flux monitoring in the Shaguo barren environment.

[0020] In some embodiments, the present application further proposes a method for calculating the carbon flux of the Shagehuang new energy base based on detailed simulation results, comprising: dividing the Shagehuang new energy base into multiple partition types; for each partition type, calculating a first carbon flux by a gradient diffusion method according to the carbon dioxide concentration distribution and the flow velocity distribution, and calculating a second carbon flux by an eddy correlation method; performing a weighted summation on the first carbon flux and the second carbon flux to obtain a basic carbon flux corresponding to each partition type; correcting the basic carbon flux based on a correction coefficient of each partition type to obtain a corrected carbon flux of each partition type; performing an area-weighted summation on the corrected carbon fluxes of all partition types, and obtaining the carbon flux of the Shagehuang new energy base by dividing the obtained sum by the total area.

[0021] In a specific implementation process, the partition type refers to different functional areas divided according to the terrain structure, vegetation coverage, and distribution characteristics of new energy facilities, which can be realized by comprehensively judging multiple source data such as digital elevation model data, multispectral remote sensing images, hyperspectral remote sensing images, wind turbine coordinates, wind resource meteorological data, and equipment deployment drawings.

[0022] Specifically, the embodiment provides a specific implementation mode for calculating a terrain complexity index based on a digital elevation model (DEM) and applying it to partition judgment. The DEM data can be obtained from spaceborne radar or LiDAR measurement, and the spatial resolution is preferably 5-30 meters.

[0023] Specifically, the digital elevation model data can be preprocessed to calculate the terrain complexity index of each calculation grid, and the calculation grid can be preliminarily partitioned and judged based on a preset terrain threshold to obtain a first partition result. In order to subsequently comprehensively logically judge each calculation grid based on the partition result obtained by other methods, determine the final partition type to which each calculation grid belongs. The specific implementation process is as follows: S11, data preprocessing: Obtain the DEM data of the Shagehuang new energy base. First, preprocess the DEM data to eliminate data holes and outliers, which includes smoothing and denoising by Gaussian filtering or median filtering, and filling invalid data areas by natural neighbor method or Kriging interpolation method.

[0024] S12, calculate the slope value (Slope): calculate the slope value in the specified first moving window by using the maximum elevation difference method or Horn algorithm: Specifically, based on the pre-processed DEM data, the maximum elevation difference method or the Horn algorithm can be used to calculate the slope value in a specified moving window (for example, a 3x3 pixel window) per pixel, in degrees or percentage. The slope is used to represent the degree of inclination of the ground surface.

[0025] The calculation formula (take the maximum elevation difference method as an example):

[0026] Wherein, represents the slope value, is the elevation change rate in the east-west direction, is the elevation change rate in the south-north direction.

[0027] The partition application corresponding to the slope value: compare the slope value of the calculation grid with the preset threshold value to obtain the partition. For example, the rules of the partition can include but are not limited to the following rules, which can be set according to actual needs: If the average slope is less than 5°, the terrain of the region is flat, and it is suitable to be divided into a photovoltaic power generation area.

[0028] If the average slope is between 5° and 15°, the region needs to be further identified in combination with other indicators.

[0029] If the average slope is greater than 15°, the terrain of the region is steep, and the construction and maintenance cost is high, so it is preferred to be divided into a virgin desert area or an ecological protection area, and excluded from the core construction area of the photovoltaic area and the wind power generation area.

[0030] S13, calculate the aspect value (Aspect): use the preset aspect calculation formula to calculate the aspect value in the first moving window: Specifically, based on the pre-processed DEM data, the same moving window as used when calculating the slope value can be used to calculate the aspect value, in degrees, from 0° to 360° (usually north is 0°).

[0031] The aspect calculation formula is as follows:

[0032] Wherein, represents the slope value. The partition application corresponding to the slope value: the aspect is mainly used to analyze the amount of solar radiation received.

[0033] The south slope (for example, 135° to 225°) receives the strongest solar radiation in the middle and high latitudes of the hemisphere, which has a positive effect on photovoltaic power generation efficiency, and can be used as a preferred factor for photovoltaic area in partition identification.

[0034] The north slope receives less radiation, which may cause snow or vegetation differences.

[0035] S14, Calculate the relief amplitude of the terrain: using a preset relief amplitude calculation formula, calculate the relief amplitude of the terrain within a specified second moving window: The relief amplitude of the terrain is an index that describes the extreme difference in elevation within a unit, and can effectively reflect the macroscopic roughness of the terrain. Its calculation needs to be performed in a larger analysis window (such as 500m × 500m).

[0036] The relief amplitude calculation formula is:

[0037] Among them, to calculate the relief amplitude of the terrain, and are the maximum and minimum elevation values in the analysis window, respectively.

[0038] The application of the corresponding partition of the relief amplitude of the terrain: If the relief amplitude of the terrain is less than 30 meters, it indicates that the terrain is flat or hilly, and is suitable for large-scale new energy facility construction.

[0039] If the relief amplitude of the terrain is between 30 meters and 100 meters, it indicates that the terrain is hilly, and careful evaluation of facility layout is required, which may be divided into a mixed use area.

[0040] If the relief amplitude of the terrain is greater than 100 meters, it indicates that the terrain is mountainous or deeply cut, and is divided into a primitive desert area or a restricted development area in principle, as it has a complex impact on wind fields and construction difficulty is extremely great.

[0041] S15, Comprehensive discrimination: For each calculation grid, the first partition result obtained by the slope, slope direction, and terrain relief amplitude indicators is superimposed and fused with its own data such as image spectral data (such as normalized vegetation index value (NDVI)), device deployment information (such as wind turbine foundation positions extracted from construction drawings), or other partition results obtained, for example, a priority level for different partition results can be set, and a corresponding fusion strategy can be set for different priority levels for fusion, wherein the fusion strategy can be set according to actual needs. For example: A unit may have a flat slope (<5°), a small relief amplitude (<30m), a very low NDVI value, and be located within the planning range of a photovoltaic plant, and is finally determined to be a photovoltaic power generation area.

[0042] A unit has a steep slope (>10°), a large relief amplitude (>50m), but NDVI shows natural vegetation coverage, and is finally determined to be a primitive desert area.

[0043] A unit has no special restrictions on terrain indicators, but high-spectral image identifies concrete foundation features, and construction drawings indicate a booster station, and is finally determined to be an electrical facility area.

[0044] The above examples are only illustrative, and the present example does not limit other fusion strategies, which are not illustrated one by one.

[0045] Through the above embodiments, the method realizes quantitative calculation and automatic partition discrimination of the terrain complexity index, and effectively improves the objectivity, precision and efficiency of the partition of the Shagehuang new energy base.

[0046] In a specific implementation process, NDVI can also be calculated based on multispectral remote sensing images, and the specific implementation mode of dividing the vegetation coverage level according to a preset threshold. The multispectral remote sensing image data source can be derived from a sentinel or a satellite image with higher resolution, and at least contains a red band (Red, about 0.64-0.67 μm) and a near-infrared band (NIR, about 0.85-0.88 μm).

[0047] Specifically, the multispectral remote sensing image can be radiometrically calibrated and atmospherically corrected, NDVI can be calculated pixel by pixel, and the vegetation coverage level can be divided according to a preset vegetation index threshold. The vegetation coverage level is combined to preliminarily discriminate the calculation grid to obtain a second partition result, so as to subsequently comprehensively logically discriminate each calculation grid with the partition result obtained by other methods, and determine the final partition type to which each calculation grid belongs. The specific implementation process is as follows: S21, data preprocessing and NDVI calculation: Multispectral remote sensing images covering the Shagehuang new energy base are obtained. First, the images are radiometrically calibrated and atmospherically corrected (models such as FLASSH and 6S can be used) to eliminate the interference of aerosols, water vapor and other factors in the atmosphere on the reflectivity of ground objects, and obtain the true reflectivity data of the ground surface. Then, the normalized difference vegetation index value is calculated using a preset normalized difference vegetation index calculation formula: ; wherein, is the reflectivity of the near-infrared band, is the reflectivity of the red light band. The theoretical value range of NDVI is [-1, 1].

[0048] S22, determine the vegetation coverage level threshold, to determine the vegetation coverage level and the second partition result corresponding to the vegetation coverage level according to the range to which the normalized difference vegetation index value belongs: Due to the unique environment and sparse vegetation coverage in the Shagehuang area, the NDVI threshold range is different from that in general dense vegetation areas. Based on statistical analysis of historical field investigation samples and remote sensing images, the following vegetation coverage level discrimination threshold is set for the base: Bare soil / sand: NDVI≤0.1; This area is mainly bare sand, gobi, rock or construction site. NDVI value is low, near-infrared reflectance is close to red light reflectance.

[0049] Sparse Vegetation: 0.1 < NDVI ≤ 0.2; This area is mainly sparse desert herbaceous plants, mosses or seedlings in early growth stage. Vegetation canopy coverage is usually less than 15%.

[0050] Moderate Vegetation: 0.2 < NDVI ≤ 0.4; This area is mainly well-grown shrubs, artificially planted sand-fixing grass blocks, or higher coverage natural grassland. Vegetation canopy coverage is usually between 15% and 50%.

[0051] Dense Vegetation: NDVI > 0.4; This area is mainly artificially maintained green belt, shelter forest, or rare natural oasis. Vegetation canopy coverage is higher than 50%.

[0052] Optionally, S23, post-processing and optimization: The preliminary calculated NDVI raster layer is post-processed to eliminate noise and outliers. Majority filtering or morphological opening and closing operation can be used to smooth the data, for example, using a moving window of 3x3 or 5x5 pixels, to remove false positives caused by single pixel anomalies.

[0053] S24, comprehensive discrimination and zoning application: The calculated vegetation coverage level map is superimposed and fused with the results of other zoning standards (terrain complexity, facility type, etc.) to participate in the decision of the final zoning type. For example: If a calculation grid is judged as no vegetation coverage area, and its terrain is flat (slope < 5°), and it is located in the photovoltaic planning area, then its final zoning type is photovoltaic power generation area.

[0054] If a calculation grid is judged as high vegetation coverage area, regardless of its terrain, it should be given priority to be divided into artificial green area.

[0055] If a calculation grid is judged as low vegetation coverage area, and its terrain undulation is > 50 meters, and no facilities are planned, then its final zoning type is original desert area.

[0056] If a computing grid is identified as a medium vegetation coverage area, but a booster station is built on it (identified from construction drawings), the final zoning type of the computing grid is electrical facility area, and the vegetation on it is considered as a green component within the facility area.

[0057] Through the above examples, the method realizes the objective and quantitative classification of vegetation coverage in the special environment of Shagehuang, and provides key data support for ecological evaluation, green engineering effect monitoring and zoning planning of new energy bases.

[0058] In a specific implementation process, the hyperspectral remote sensing image can also be radiometrically calibrated and atmospherically corrected, the characteristic high reflection peak of the photovoltaic panel in the near-infrared band is extracted, the characteristic high reflection peak is enhanced, the photovoltaic panel pixels are identified by combining the pre-set reflection peak threshold, and the preliminary range of the photovoltaic power generation area is formed through post-processing. The computing grid is preliminarily zoned and identified to obtain a third zoning result, so as to be combined with the zoning results obtained by other methods in the subsequent process, and each computing grid is comprehensively logically identified to determine the final zoning type of each computing grid. The hyperspectral remote sensing data source is preferably a sensor with high spectral resolution, and the wavelength setting needs to cover at least the near-infrared region of 760 nm to 900 nm. The specific implementation process is as follows: S31, data preprocessing and spectral library construction: A hyperspectral image cube covering the Shagehuang new energy base is obtained. First, the data is radiometrically calibrated and atmospherically corrected (advanced atmospheric correction models such as MODTRAN and FLAASH can be used) to accurately restore the surface reflectance spectrum of the ground object.

[0059] At the same time, a standard spectral feature library of photovoltaic panels is constructed. Through field measurement (using ASD ground object spectrometer) or from verified public / commercial spectral libraries, typical reflection spectrum curves of photovoltaic panels of different materials, different degrees of aging, and different installation angles are obtained as reference benchmarks for identification.

[0060] S32, characteristic reflection peak extraction and enhancement: Due to the anti-reflective film (AR Coating) and semiconductor material (such as silicon) on the surface of the photovoltaic panel, a very significant and stable high reflection peak is formed in the near-infrared band (especially near 850 nm), which is the core spectral fingerprint that distinguishes it from other natural ground objects (such as soil, vegetation, and rock).

[0061] The spectral curve of each pixel in the image is operated as follows to enhance the feature: 1. Calculate the first derivative spectrum: to eliminate the influence of background noise and light conditions, and accurately lock the wavelength position of the reflection peak (the point where the derivative value is zero corresponds to the peak value position of the reflection peak).

[0062] 2. Calculate the normalized index (optional): An index like (R850 - R700) / (R850 + R700) can be constructed to quantify the intensity of the reflection peak, where R850 and R700 represent the reflectance values at 850 nm and 700 nm, respectively.

[0063] S33, Set the discrimination threshold and identification: Based on the standard spectral feature library and statistical analysis of a large number of samples, the following discrimination threshold is set for the base: Reflection peak intensity threshold: the reflectance value at 850 nm is greater than 0.35.

[0064] Reflection peak shape threshold: within the wavelength range of 760 nm to 900 nm, there must be a continuous and prominent reflection peak with a half-width (FWHM) within a certain range (e.g. 30-50 nm), and the spectral curve has no obvious absorption valley in this interval (to distinguish from the "red edge" feature of vegetation).

[0065] Auxiliary exclusion condition: the NDVI value of the pixel should be less than 0.2 to exclude the possibility of dense vegetation; at the same time, the blue band reflectance should not be too high to exclude water bodies.

[0066] Compare the spectral curve of each pixel with the above threshold. If the reflection peak intensity threshold and the reflection peak shape threshold are met at the same time, the pixel is preliminarily determined as a photovoltaic panel pixel.

[0067] S34, Post-processing and object optimization: Post-processing of the preliminary identified scattered pixels to form a complete photovoltaic power generation area: 1. Morphological processing: using clustering analysis and morphological closing operation (first expansion and then corrosion), the adjacent scattered pixels are aggregated into complete contiguous areas, and the internal pores are filled and the boundaries are smoothed.

[0068] 2. Spatial constraint: combined with the photovoltaic array planning boundary extracted from the base construction drawings, the identification results are spatially constrained and verified, and the misidentified spots obviously located outside the planning area are removed.

[0069] 3. Logical verification: if a region is identified as a photovoltaic panel, but its average slope is greater than 15°, it needs to be confirmed by manual visual interpretation to exclude false positives caused by steep bare rock.

[0070] S35, Comprehensive discrimination and zoning confirmation: The photovoltaic panel coverage area identified by the hyperspectral feature recognition above is directly assigned the partition type of "photovoltaic power generation area". This is one of the high-priority discrimination criteria. Its boundary is superimposed with the partition results based on other criteria (such as terrain, NDVI), ensuring the logical consistency of the partition map.

[0071] In a specific implementation process, in order to realize the accurate quantification and spatial division of the influence range of the wake effect of the wind turbine in the Shaguo Huang new energy base, the embodiment provides a wake area demarcation method coupled with the geographic location of the wind turbine, fluid dynamics model and geographic information system (GIS) spatial analysis technology. This method not only considers geometric distance, but also introduces wind resource distribution characteristics and special environmental factors in Shaguo Huang, thereby dynamically and accurately defining the influence range of the wake.

[0072] Specifically, the wind turbine coordinates, wind resource meteorological data and digital elevation model data can be used to calculate the influence range of the wind turbine wake using a customized wake model combined with a wind speed loss threshold, to preliminarily partition and distinguish the calculation grid, and obtain a fourth partition result. In order to subsequently comprehensively logically distinguish each calculation grid based on the partition results obtained by other methods, determine the final partition type of each calculation grid. The specific implementation process is as follows: S41, data preparation and preprocessing: The following data of the Shaguo Huang new energy base is obtained: Wind turbine coordinates: From the construction drawings, as-built drawings provided by the wind turbine manufacturer or on-site GPS measurement, the accurate geographic coordinates (latitude and longitude), hub height and rotor diameter of each wind turbine are extracted.

[0073] Wind resource meteorological data: Obtain long-term wind observation data of the base, and analyze to obtain the dominant wind direction (such as prevailing northwest wind) and wind speed frequency distribution. In particular, parameters for characterizing atmospheric stability, such as the Richardson number (Ri) calculated from the temperature profile or the turbulence intensity data directly from the wind tower, need to be obtained.

[0074] High-precision DEM data: used for subsequent analysis of the potential impact of terrain on the wake. A surface roughness map can be generated based on high-precision DEM data.

[0075] S42, build and customize wind turbine wake model: The wind turbine wake is a complex fluid dynamics phenomenon. The Jensen wake model (also known as the Park model) widely verified in the industry is used as the core calculation engine, and the key parameters are customized for the Shaguo Huang environment.

[0076] The core function of the Jensen model is as follows:

[0077] wherein, V(x) represents the wind speed at the downwind distance x, V0 represents the incoming wind speed, Ct represents the thrust coefficient of the wind turbine, D represents the rotor diameter of the wind turbine, and x represents the downwind distance, k represents the wake decay coefficient, which is dynamically adjusted according to the atmospheric stability. Wherein, after obtaining the difference between and , the wind speed deficit can be obtained.

[0078] S43, special adaptation of Shaguohe environment - dynamic wake decay coefficient k: The atmospheric stratification in Shaguohe area is unstable (intense convective flow due to strong heating of the ground surface during the day) and the background turbulence intensity is high (due to the ground roughness and thermal effect), which will cause the recovery speed of the wake to accelerate and the influence range to shrink significantly compared with the neutral stability condition.

[0079] Adaptive design: The wake decay coefficient k is a function dynamically related to the atmospheric stability. Based on the observation and research of the atmospheric boundary layer characteristics in Shaguohe area, the following relationship is used for dynamic setting in this embodiment: Unstable atmospheric conditions (usually occur during sunny periods during the day): k = 0.05 to 0.06 (fast recovery of wake, small influence range).

[0080] Neutral atmospheric conditions (usually occur during overcast or windy periods): k = 0.04 to 0.05 (standard case).

[0081] Stable atmospheric conditions (usually occur during clear sky periods at night): k = 0.03 to 0.04 (slow recovery of wake, large influence range).

[0082] By statistical analysis of wind frequency under different stability conditions, a weighted average k value is assigned to each wind direction sector, or a comprehensive result is synthesized after running multiple stability scenarios.

[0083] S44, define wake influence threshold and buffer zone generation: Set the wind speed deficit threshold: define when the wind speed of a certain point is reduced to 90% of the incoming wind speed (i.e. 10% wind speed deficit) due to the influence of the upstream wind turbine wake, the point is considered to be in the significantly influenced wake area. This threshold can be adjusted according to the specific requirements of power generation efficiency or turbulence intensity.

[0084] Calculate dynamic wake range: Take a single wind turbine as the center, along the dominant wind direction and the secondary wind direction (usually covering 16 wind direction sectors), calculate the furthest distance that the wind speed deficit reaches the wind speed deficit threshold as the dynamic wake range according to the customized Jensen model (using dynamic k value) under each wind direction sector.

[0085] Generate wind direction sector buffer zone: Calculate the furthest distance that the wind speed deficit reaches the wind speed deficit threshold as the dynamic wake range under each wind direction sector. The azimuth angle of each sector is determined by the wind direction frequency, and the radius R is determined by the furthest influence distance of the wake under this wind direction (R = x_max).

[0086] Generate comprehensive wake influence area: Spatially superimpose all wind direction sector buffer zones of all wind turbines to generate a comprehensive wake influence area layer for the entire base.

[0087] S45, spatial superposition and partition discrimination: Superimpose the generated comprehensive wake influence area layer with the calculation grid to obtain the fourth partition result.

[0088] For calculation grids located entirely or partially within the comprehensive wake influence area, mark or partially mark their partition type as "wind turbine wake influence area".

[0089] This area is a key input area for subsequent high-precision turbulence intensity calculation and carbon flux transport simulation.

[0090] S46, wind erosion risk identification: Wind turbine wake vortex can significantly enhance the wind speed shear force in the near-surface layer, intensify soil wind erosion on sandy surfaces, trigger dust, and indirectly affect carbon flux (dust deposition) and the surrounding ecological environment.

[0091] Adaptive design: Based on the delineation of the wake area, superimpose the surface soil type map (which can be interpreted from remote sensing images or geological maps). If the wake influence area overlaps with the loose sandy soil area, further identify the overlapping area as a "high wind erosion risk area" and highlight it in the final partition map to provide a precise target area for wind-blown particulate matter source items.

[0092] In a specific implementation process, to achieve the objectivity, accuracy and engineering practicability of the Shagehuang new energy base partition, the embodiment provides a method of comprehensively utilizing the coordinate information of the equipment deployment drawing, the remote sensing image recognition result and the geographic information system (GIS) spatial analysis technology to spatially constrain, verify and refine the partition result. This method ensures that the partition boundary is highly consistent with the actual engineering facilities.

[0093] Specifically, the coordinates of the facility elements in the equipment deployment drawing are converted to the coordinate system consistent with the multi-source data, a facility element vector layer is generated, preliminary partition discrimination is performed on the calculation grid to obtain a fifth partition result, so as to be subsequently combined with the partition results obtained by other methods, and logical discrimination is performed on each calculation grid to determine the final partition type to which each calculation grid belongs; the specific implementation process is as follows: S51, data preparation and coordinate system: Obtain the equipment deployment design drawing or as-built drawing of the Shaguo Barren New Energy Base. The drawing is usually in CAD format (such as.dwg or.dxf) and contains the spatial range, boundary and attribute information of various facilities.

[0094] Data extraction: parse and extract the vector boundary coordinates of the following key facility elements from the drawing: Photovoltaic array area: the precise boundary range of each photovoltaic array.

[0095] Wind turbine: center point coordinates of the wind turbine tower and its foundation range.

[0096] Electrical facilities: boundary range of booster stations, collection stations and box-type transformers.

[0097] Road: center line or boundary line of the construction and maintenance road.

[0098] Other structures: office area, fence, etc.

[0099] Coordinate conversion and projection: the original coordinate system in the drawing (which may be a local coordinate system or a construction coordinate system) is accurately converted through public control points to be unified into the same geodetic coordinate system (such as WGS-84) and projection coordinate system (such as UTM) as the remote sensing image and DEM data. This is the premise of all spatial overlay analysis.

[0100] S52, construction of authoritative vector layer of facility elements: Convert the converted facility elements of various types to generate corresponding vector layers (such as face-shaped “.shp” files) in GIS software: Photovoltaic array area (PV_Design.shp); Turbine foundation range (Turbine_Design.shp); Substation range (Substation_Design.shp); Road surface area (Road_Design.shp); These vector layers constitute the “authoritative spatial reference” for partition.

[0101] S53, spatial constraint and forced assignment - highest priority discrimination: Overlay Analysis of the above authoritative vector layers and the preliminary zoning results interpreted by remote sensing, and the following forced assignment logic is executed to ensure that the map information has the highest priority: Direct assignment area: Any calculation grid within the PV_Design.shp face domain, regardless of its spectral, topographic features, must be forcibly divided into "Photovoltaic Power Generation Area".

[0102] Any calculation grid within the Turbine_Design.shp face domain must be forcibly divided into "Wind Power Generation Area".

[0103] Any calculation grid within the Substation_Design.shp or Road_Design.shp face domain must be forcibly divided into "Electrical Facilities Area" or "Road Area".

[0104] This step ensures that the zoning results are completely consistent with the engineering design, avoiding any possible errors in remote sensing identification.

[0105] S54, spatial verification and conflict processing: For areas outside the range of authoritative vector layers, mainly rely on remote sensing interpretation results. However, for areas adjacent to the interpretation results and map information, consistency verification is performed: False Negative: If remote sensing interpretation divides a certain area into "Original Desert Area", but the area is mostly covered by PV_Design.shp, it means that the remote sensing identification missed the photovoltaic array. The system should record the conflict and correct it according to the map.

[0106] False Positive: If remote sensing interpretation divides a certain area into "Photovoltaic Power Generation Area", but the area is completely not within the range of any design map, it means that remote sensing identification may misjudge bare rock, water accumulation, etc. as photovoltaic panels. The system should record the conflict and start manual interactive inspection or re-identify combined with other evidence (such as time sequence images).

[0107] S55, define and identify "Facility Proximity Area": Special consideration: In Shaguo Desert New Energy Base, there is a unique microenvironment and human activity around the facilities, which needs to be distinguished.

[0108] Around Substation_Design.shp, Road_Design.shp, etc., generate a certain width (such as 10-20 meters) buffer (Buffer).

[0109] The grid cells located in the buffer zone and not divided into the core facility area are identified as the "facility adjacent area". This area may be frequently disturbed by human activities, and the surface type is complex (such as compacted land, temporary piles), and its carbon flux characteristics may be significantly different from the "original desert area" far away. This refined classification improves the fineness of the partition and the accuracy of the model.

[0110] S56, generate the most authoritative partition map: Comprehensive of the above steps: The vector surface domain generated by the device deployment coordinates is used as the core framework for forced assignment.

[0111] Outside this framework, rules such as remote sensing spectrum and terrain are applied for supplementary discrimination.

[0112] Logical consistency checks are performed on all areas (such as wind turbine foundations should not appear in the photovoltaic area), and the final digital partition map corresponding to the field engineering is generated.

[0113] Through the above embodiments, the method closely combines abstract remote sensing image features with real-world engineering coordinates, forming a "top-down" (paper constraints) and "bottom-up" (remote sensing inversion) combined double-driven partition process, which maximizes the accuracy, reliability and engineering application value of the partition results, and provides a data basis for subsequent accurate carbon flux measurement.

[0114] It should be noted that after obtaining each partition result, how to comprehensively logically discriminate with other partition results to determine the final partition type of each calculation grid is only an example, and the comprehensive logic can be set according to actual needs, which will not be described one by one.

[0115] In a specific implementation process, the gradient diffusion method refers to a method for calculating vertical material transport based on turbulent diffusion theory, which can be specifically implemented by analyzing the product of carbon dioxide concentration gradient and eddy diffusion coefficient, and is used to quantify the contribution of vertical diffusion process to carbon flux. The eddy correlation method refers to a method for calculating material flux based on the characteristics of atmospheric turbulence fluctuations, which can be specifically implemented by calculating the correlation between the vertical wind speed fluctuation value and the concentration fluctuation value through covariance, and is used to capture the characteristics of turbulent instantaneous exchange. Weighted summation refers to the comprehensive processing of the calculation results of different methods, which can be specifically implemented by using a dynamic weight distribution strategy combined with an error analysis model, and is used to balance the systematic errors of different methods. The correction coefficient is an adjustment parameter reflecting the influence of a specific partition environment, which is used to eliminate the calculation bias caused by the difference in surface type.

[0116] Specifically, first, the new energy base is divided into a photovoltaic power generation area, a wind power generation area, an electrical facility area, a road area, a virgin desert area, and an artificial greening area, and each division is based on spatial attributes including a vegetation index threshold and a facility type. Gradient diffusion and eddy correlation methods are simultaneously performed for each division to calculate carbon flux. Gradient diffusion obtains flux values by multiplying the concentration gradient of three vertical height layers by the eddy diffusion coefficient, and the eddy correlation method obtains flux values by calculating the covariance of high-frequency fluctuation data. The results of the two methods are weighted according to the determination coefficient to form a fused basic flux value. On this basis, a correction coefficient is applied to the underlying surface characteristics of each division, and the global carbon flux is finally calculated by area weighting.

[0117] In a specific implementation process, the flux correction logic of each division is as follows: the flux of the virgin desert area is dynamically adjusted based on a soil respiration correction coefficient, the flux of the photovoltaic power generation area is introduced into a shading correction coefficient, the flux of the wind power generation area is superimposed with a heat disturbance term, the flux of the artificial greening area is superimposed with a photosynthetic absorption flux, the flux of the electrical facility area is superimposed with an inverter emission term, and the flux of the road area can be superimposed with a vehicle emission term.

[0118] The virgin desert area flux correction coefficient is a dynamic adjustment factor based on changes in soil temperature and moisture content, which can be realized by a linear combination formula of temperature and moisture content, and is used to reflect the nonlinear response of soil respiration to environmental changes. The photovoltaic panel shading rate is the shading proportion of the photovoltaic array to the ground, which can be realized by analyzing the unmanned aerial vehicle image or calculating the irradiance intensity, and is used to quantify the shading effect of the photovoltaic facility on the ground carbon exchange process. The tip speed of the fan blade is the linear speed of the fan blade tip, which can be realized by multiplying the rotational speed and the blade length, and is used to represent the influence degree of the fan operation on the surrounding turbulence intensity. The photosynthetically active radiation is the solar radiation energy in the wave band available for plant photosynthesis, which can be realized by measuring with a spectral sensor or converting by a solar radiation model, and is used to quantify the absorption contribution of plant photosynthesis to carbon flux. The leaf area index is the total area of the vegetation leaf per unit ground surface area, which can be realized by multispectral image inversion or laser radar scanning, and is used to reflect the influence of the vegetation canopy structure on the photosynthetic efficiency. The inverter emission term is the carbon dioxide emission amount generated by the operation of the power facility, which can be realized by calculating the power and the emission coefficient model, and is used to quantify the additional influence of the device operation on the carbon flux. The road area vehicle emission coefficient is the carbon dioxide emission amount generated per vehicle flow or per driving distance, which can be realized by matching the vehicle type classification statistics with the emission factor database, and is used to quantify the additional influence of the road passing vehicles on the carbon flux.

[0119] Specifically, the method divides the photovoltaic power generation area, the wind power generation area, the electrical facility area, the road area, the original desert area and the artificial greening area into multiple areas, and establishes a differentiated correction model for the underlying surface characteristics and facility interference of each area. The flux correction of the original desert area considers the synergistic effect of soil temperature and moisture content on respiration, and dynamically adjusts the basic flux value by using a linear combination formula; the flux correction of the photovoltaic array area quantitatively evaluates the shading effect of the photovoltaic panel based on the shading rate, and reduces the basic flux value by using a proportional coefficient; the flux correction of the fan surrounding area combines the blade tip speed to calculate the turbulence enhancement effect, and superimposes a constant heat disturbance term to reflect the heat emission of the fan operation; the flux correction of the vegetation area introduces a product term of photosynthetically active radiation and leaf area index to represent the absorption capacity of vegetation photosynthesis to carbon dioxide; the flux correction of the facility area adjusts the basic flux by using a temperature-dependent coefficient, and superimposes an inverter emission term inversely proportional to the facility area; and the flux of the road area can superimpose a vehicle emission term. The parameters of each correction term are derived from the multi-field data output by the LBM simulation, and the fine calculation of the partitioned carbon flux is realized by coupling the formulas.

[0120] The scheme effectively solves the adaptability problem of carbon flux calculation in facility-intensive areas by using the multi-attribute gridding partition mechanism, combining the complementary advantages of the gradient diffusion method and the eddy correlation method, and introducing a dynamic correction coefficient. Compared with the eddy correlation method which is easily affected by the intermittency of turbulence, or the gradient diffusion method which is difficult to capture transient transport, the scheme makes the root mean square error of the two methods lower by using a weight fusion mechanism.

[0121] Through the above technical scheme, the present application realizes the fine modeling of the complex underlying surface conditions in the new energy base, and overcomes the calculation deviation of the traditional method in the facility-intensive area. Through the partition correction mechanism, the interaction between natural carbon cycle and artificial facilities is effectively distinguished, so that the carbon flux calculation error under special conditions such as photovoltaic array shading effect and fan wake disturbance is controlled within 5%. The Monte Carlo method is used for uncertainty analysis, which can quantify the sensitivity of different correction coefficients to the final result, and provide data support for the optimization deployment of the carbon flux monitoring network.

[0122] In some embodiments, the present application further proposes a method for calculating the first carbon flux by the gradient diffusion method, comprising calculating the vertical concentration gradient of carbon dioxide and the eddy diffusion coefficient based on the turbulent diffusion theory, combining the concentration distribution of carbon dioxide and the flow velocity distribution, taking the negative value of the product of the vertical concentration gradient of carbon dioxide and the eddy diffusion coefficient to obtain the first carbon flux. The calculation process of the vertical concentration gradient of carbon dioxide includes selecting a plurality of turbulent stable vertical height layers from a preset height range, wherein the height difference of each vertical height layer is a preset difference value, extracting the average value of the concentration of carbon dioxide of each vertical height layer, and calculating the concentration difference of carbon dioxide of adjacent two vertical height layers, and taking the average value of all the concentration differences of carbon dioxide to obtain the vertical concentration gradient of carbon dioxide. The calculation process of the eddy diffusion coefficient includes selecting the first flow velocity of the data point at the first height and the second flow velocity of the data point at the second height based on the flow velocity distribution, determining the height difference between the second height and the first height and the flow velocity difference between the first flow velocity and the second flow velocity, and taking the product of the height difference, the flow velocity difference and the von Karman constant to obtain the eddy diffusion coefficient.

[0123] The vertical concentration gradient of carbon dioxide refers to the difference in the concentration of carbon dioxide between different vertical height layers, which can be specifically obtained by selecting a plurality of turbulent stable vertical height layers from a preset height range, extracting the average value of the concentration of carbon dioxide of each vertical height layer, and calculating the concentration difference of carbon dioxide of adjacent two vertical height layers; and taking the average value of all the concentration differences of carbon dioxide to obtain the vertical concentration gradient of carbon dioxide. The height difference of each vertical height layer is a preset difference value. The vertical concentration gradient of carbon dioxide reflects the diffusion trend of carbon dioxide in the vertical direction and is the core parameter for calculating the carbon flux by the gradient diffusion method.

[0124] The eddy diffusion coefficient refers to a parameter representing the ability of turbulent motion to diffuse matter, which can be specifically obtained by selecting the first flow velocity of the data point at the first height and the second flow velocity of the data point at the second height based on the flow velocity distribution; determining the height difference between the second height and the first height and the flow velocity difference between the first flow velocity and the second flow velocity; and taking the product of the height difference, the flow velocity difference and the von Karman constant to obtain the eddy diffusion coefficient. The eddy diffusion coefficient reflects the contribution of atmospheric turbulence to the vertical transport of carbon dioxide and directly affects the accuracy of carbon flux calculation.

[0125] Specifically, in the implementation process, first, a plurality of turbulence stable vertical height layers are selected from the LBM simulation data, for example, three height layers are selected in the range of 10 meters to 30 meters, and the height difference between adjacent layers is 10 meters. After extracting the average concentration of carbon dioxide of each layer, the concentration difference between adjacent layers is calculated, and the average value is taken as the vertical concentration gradient. At the same time, based on the flow velocity distribution data of the same area, two different height wind speed values are selected, the height difference and the wind speed difference are calculated, and the eddy diffusion coefficient is calculated combined with the von Karman constant. For example, when the first height is 10 meters and the second height is 20 meters, the measured wind speed is 5 meters / second and 6 meters / second respectively, and the eddy diffusion coefficient is calculated by the formula. Finally, the vertical concentration gradient and the eddy diffusion coefficient are multiplied and taken as negative, and the first carbon flux is obtained.

[0126] The method can effectively reduce the measurement deviation caused by local turbulence fluctuation by selecting a plurality of turbulence stable height layers and calculating the average concentration difference. At the same time, the method of calculating the eddy diffusion coefficient based on the wind speed difference of two heights is more accurate than the traditional single-point wind speed estimation to reflect the actual turbulence diffusion capacity.

[0127] Through the above technical solutions, the application can improve the calculation accuracy of the vertical concentration gradient of carbon dioxide and the eddy diffusion coefficient, thereby ensuring the reliability of the first carbon flux data obtained by the gradient diffusion method. The method effectively suppresses the influence of random errors of single-point measurement on the final result by using multi-height layer data averaging and double-point wind speed calculation, and provides more accurate basic input parameters for subsequent carbon flux fusion calculation.

[0128] In some embodiments, the application further proposes to set differentially according to each partition type and meteorological interference factors to obtain a probability distribution model of the eddy diffusion coefficient, the vertical concentration gradient of carbon dioxide and the correction coefficient; wherein the meteorological interference factors at least include aerosol optical interference caused by strong wind sand and severe near-surface thermal turbulence caused by strong solar radiation; perform Monte Carlo random disturbance iteration based on the probability distribution model of the differential setting to obtain a plurality of candidate carbon fluxes; after sorting the plurality of candidate carbon fluxes, extract the first n candidate carbon fluxes and the last m candidate carbon fluxes, and the minimum candidate carbon flux and the maximum candidate carbon flux in the remaining candidate carbon fluxes constitute the confidence interval of the carbon flux of the Shagehuang New Energy Base.

[0129] The difference setting refers to different distribution modes for different partition types, and adjusting the standard deviation of each partition type for different meteorological interference factors. For example, for the turbulent diffusion coefficient, the normal distribution can be used for the unobstructed area such as the road area, and the skew distribution can be used for the photovoltaic power generation area and the wind power generation area due to the uneven distribution of turbulence caused by equipment obstruction. When strong solar radiation is monitored, the standard deviation of each partition type can be increased, and when there is no strong radiation, the standard deviation of each partition type can not be adjusted. For the carbon dioxide vertical concentration gradient and the correction coefficient, similar principles can be used to make different settings according to actual needs, which will not be illustrated one by one here.

[0130] The Monte Carlo method refers to a numerical simulation method based on probability statistics, which can be implemented by random sampling and repeated calculation. A large number of random parameter combinations are generated to simulate the distribution of carbon flux under different scenarios. The turbulent diffusion coefficient refers to a parameter representing the ability of atmospheric turbulence to diffuse matter, which can be calculated by the wind speed gradient and the von Karman constant, and is used to quantify the transmission efficiency of carbon dioxide in the vertical direction. The carbon dioxide vertical concentration gradient refers to the rate of change of carbon dioxide concentration per unit height, which can be calculated by taking the average of the concentration difference of multiple height layers, and reflects the distribution characteristics of carbon dioxide in space. The correction coefficient refers to an adjustment factor set according to the characteristics of the underlying surface in different functional partitions, which can be dynamically calculated according to soil temperature, moisture content, and equipment operation parameters, and is used to compensate for the deviation of the basic carbon flux model in specific areas. The confidence interval refers to the range representing the reliability of the results in statistics, which can be obtained by removing extreme values and retaining the region with the highest intermediate probability density, and is used to quantify the uncertainty of the carbon flux calculation results.

[0131] Specifically, the implementation process of the Monte Carlo method includes determining the disturbance range of key parameters, such as allowing ±15% fluctuation for the turbulent diffusion coefficient, allowing ±10% deviation for the carbon dioxide vertical concentration gradient, and allowing ±8% adjustment for the correction coefficient. Through a computer program, 1000 independent iterations are calculated, each iteration randomly selects parameter values from the preset disturbance range and recalculates the carbon flux. After sorting all iteration results by numerical value, removing the extreme data of the first 2.5% and the last 2.5%, the interval formed by the minimum value and the maximum value in the remaining 95% data is the confidence interval. This process can effectively reflect the influence of model parameter error on the final result.

[0132] Compared with the prior art, the traditional carbon flux evaluation method usually only uses the calculation result of a single model, without considering the influence of parameter uncertainty on the result. The existing technology lacks error propagation analysis means for multi-field coupled models in Shagehuang area, which leads to the inability to quantify the reliability boundary of the monitoring results. The present scheme introduces a parameter disturbance mechanism and a probability statistics method to construct a comprehensive evaluation system covering model errors and measurement errors.

[0133] Through the technical solution, the application can accurately quantify the uncertainty range of the calculation of the carbon flux of the Shagehuang new energy base, and provide decision makers with confidence interval data with statistical significance. The method effectively solves the problem of insufficient reliability of the calculation results of the traditional single model, and significantly improves the credibility and engineering applicability of the carbon flux monitoring results, especially under the conditions of complex underlying surface and multi-field coupling.

[0134] In some embodiments, the application further provides a method for calculating a second carbon flux by the eddy correlation method, comprising extracting high-frequency data of a plurality of data points within a preset time length based on the atmospheric turbulence fluctuation characteristics, the sampling time interval of the high-frequency data being a preset time interval, the high-frequency data including high-frequency vertical flow velocity extracted from the flow velocity distribution and carbon dioxide concentration extracted from the carbon dioxide concentration distribution, and the sampling frequency being not less than 10 Hz; performing quality marking and preprocessing on a high-frequency original data sequence corresponding to the high-frequency data, at least including eliminating abnormal values of the carbon dioxide concentration signal caused by abnormal aerosol concentration due to strong wind and sand in the Shagehuang environment; obtaining the fluctuation value of the high-frequency vertical flow velocity of each data point by differencing the high-frequency vertical flow velocity of each data point and the mean vertical flow velocity, and obtaining the fluctuation value of the carbon dioxide concentration of each data point by differencing the carbon dioxide concentration of each data point and the mean carbon dioxide concentration of all data points; and calculating the covariance according to the fluctuation value of the high-frequency vertical flow velocity of each data point and the fluctuation value of the carbon dioxide concentration of each data point as the second carbon flux.

[0135] Among them, the high-frequency data refers to sampling data with a smaller sampling time interval. The fluctuation value refers to the instantaneous deviation of a physical quantity relative to its average value, which can be calculated by sliding average processing of time series data, for example, using a 30-minute sliding window to calculate the mean value, which is used to eliminate background trends and extract turbulence fluctuation components. The covariance refers to the statistical correlation of the fluctuation components of two variables, which can be realized by time-averaging the product of the vertical wind speed fluctuation and the concentration fluctuation, for example, taking the average of the sum of the products of 18000 consecutive sampling points, which is used to quantify the contribution of vertical turbulent transport to the carbon flux.

[0136] Specifically, in the Shaguo Huang new energy base, the vertical transport of carbon dioxide caused by atmospheric turbulence has significant spatiotemporal variation characteristics. By deploying high-frequency sampling equipment, such as installing an ultrasonic anemometer and an open-path infrared gas analyzer on a flux tower, vertical wind speed and carbon dioxide concentration data are continuously collected at a sampling time interval of 20 Hz. For the data in each 30-minute time window, first, the 30-minute average value of the vertical wind speed is calculated, and the instantaneous measurement value is subtracted from the average value to obtain the vertical wind speed fluctuation sequence; the 30-minute average value of the carbon dioxide concentration is calculated synchronously to generate the concentration fluctuation sequence. Then, the two fluctuation sequences are multiplied point by point and accumulated, and then divided by the total number of data points to obtain the covariance value. The covariance directly represents the net vertical flux caused by turbulent motion, for example, when the updraft carries high concentration of carbon dioxide, it produces a positive contribution, and when the downdraft carries low concentration, it produces a negative contribution, and through statistical averaging, random errors can be eliminated and the carbon flux can be accurately reflected.

[0137] Through the above technical solution, the present application can accurately quantify the instantaneous contribution of turbulent motion to carbon flux, avoiding systematic bias caused by low-frequency sampling. For example, in the photovoltaic array area, due to the turbulent enhancement effect caused by photovoltaic panels, the accuracy of carbon flux of traditional methods is poor, while the present method captures the turbulent enhancement characteristics through high-frequency data, making the calculation results closer to the real physical process. At the same time, this method is complementary to the gradient diffusion method, considering both molecular diffusion and turbulent transport mechanisms by weighted fusion, significantly improving the overall accuracy of carbon flux inversion.

[0138] In some embodiments, the image data is a multi-temporal remote sensing image sequence covering the Shaguo Huang new energy base; and the ground parameter model is a spatiotemporal sequence prediction model constructed based on a Transformer architecture. The present application further proposes inputting the image data into a pre-trained ground parameter model for identification, and outputting the ground boundary parameters and the ground source term parameters corresponding to each grid element in the calculation grid, specifically including: extracting the dynamic change characteristics of each land cover over time from the multi-temporal remote sensing image sequence, and outputting the ground boundary parameters and the ground source term parameters corresponding to each calculation grid.

[0139] In one specific implementation process, the Transformer can include an encoder, a decoder, a self-attention mechanism, etc.

[0140] The encoder: receives a multi-temporal remote sensing image sequence (such as a time sequence composed of daily images covering the Shaguo Huang new energy base), uses a multi-head self-attention mechanism to mine the temporal dependence between images at different time steps (such as the change correlation of land cover in consecutive days), and simultaneously interacts and extracts the spatial features (such as the spatial distribution of vegetation and facilities in different regions) within a single image, encoding the image sequence into a feature representation containing spatiotemporal dynamic information.

[0141] Decoder: Based on the features output by the encoder, the decoder generates ground parameters corresponding to the calculation grid through a self-attention mechanism (including cross-attention, focusing on key features of the encoder). For static ground boundary parameters (such as average ground roughness, average thermal conductivity), the decoder directly outputs the parameter value of each grid; for dynamic ground source term parameters (such as the time series of CO2 absorption / release rate), the decoder further learns its time variation pattern and outputs the parameter sequence of continuous time steps (such as the daily CO2 release rate for the next 7 days).

[0142] Supervised learning can be used for training: use historical multi-temporal remote sensing image sequences as input, and collect ground boundary parameters (measured or high-fidelity simulation values such as average blackbody radiation, ground roughness, etc.) and ground source term parameters (measured data of CO2 absorption / release rate, sensible heat flux, etc.) corresponding to the calculation grid during the corresponding period as labels. By defining a loss function (such as mean square error for continuous parameters, and cross-entropy if classification parameters are involved), minimize the error between the model output and the label, thereby optimizing the parameters of the Transformer encoder, decoder, and self-attention mechanism.

[0143] In one specific implementation, taking the prediction of CO2 release rate on the ground of Shaguo Barren New Energy Base as an example: Input: Past 30 days, once a day, multi-temporal remote sensing image sequence covering the base (capturing the dynamic changes of vegetation, soil, new energy facilities, etc. within 30 days).

[0144] Output: Future 7 days, time series of ground CO2 release rate for each calculation grid in the base (i.e. output dynamic pattern over time), and output ground boundary parameters corresponding to the grid (such as average ground roughness, average blackbody radiation).

[0145] Effect display: Draw a line chart of "time and CO2 release rate", compare the results of the Transformer model and traditional methods (such as prediction models based on physical equations), so as to intuitively show its advantages over traditional methods in spatio-temporal sequence prediction.

[0146] In one specific implementation, the ground source term parameter is a dynamic parameter that varies with time, and the model simultaneously outputs the time variation pattern of the parameter; the ground boundary parameter includes but is not limited to the average ground roughness value, the average thermal conductivity, and the average blackbody radiation rate of the calculation grid; and the ground source term parameter includes but is not limited to the average carbon dioxide absorption / release rate source term, the oxygen absorption / release rate source term, the dust particle release rate source term, the sensible heat flux source term, the latent heat flux source term, the methane absorption / release rate source term, the nitrous oxide absorption / release rate source term, the water evaporation / transpiration rate source term, the volatile organic compound release rate source term, the artificial heat source term, and the device volatile organic compound release rate source term of the calculation grid.

[0147] Among them, the dust particle release rate source term: it is directly related to the "strong wind and sand" characteristics of the Shagehu Desert area, and the dust not only affects the ecology but also indirectly affects the climate by changing the radiation transmission, snow and ice albedo, etc. The sensible heat flux source term: the heat exchange between the ground and the atmosphere due to temperature difference. Strongly affects the stability and turbulence development of the atmospheric boundary layer, and directly restricts the eddy diffusion coefficient.

[0148] The latent heat flux source term (evaporation / transpiration source term): the energy consumed by water evaporation or plant transpiration. It is a key parameter connecting water cycle and energy cycle. Even in arid areas, the evaporation process after a small amount of vegetation and short rainfall is very important.

[0149] The methane absorption / release rate source term: although the Shagehu Desert area is not the main source or sink of methane, nearby livestock or wetlands may contribute. Consideration of comprehensiveness.

[0150] The nitrous oxide absorption / release rate source term: also from soil microbial processes.

[0151] The volatile organic compound release rate source term: vegetation (especially certain shrubs) releases VOCs, which participate in atmospheric chemical reactions.

[0152] The water evaporation / transpiration rate source term: another form of latent heat flux, quantifying water loss.

[0153] The dew condensation rate source term: in arid areas, dew formed by condensation at night is an important water source for some plants and microorganisms.

[0154] The source term directly related to new energy facilities includes: The artificial heat source term: the heat emitted during the operation of power facilities such as photovoltaic inverters and box transformers, forming a local, stable artificial heat source that affects local microclimate.

[0155] The device volatile organic compound release rate source term: trace gases that may be released by materials such as transformer insulating oil and cable sheath at high temperatures.

[0156] In one specific implementation process, in order to realize high-precision supervised training of the ground parameter model, high-credibility ground true value data needs to be obtained as a sample label. The embodiment provides a set of sample label library construction process combining multi-source fusion, air-ground cooperation, and mechanism and data driving according to the environmental characteristics of the Shaguo Gobi new energy base.

[0157] S61, acquiring a label of a ground boundary parameter: The ground boundary parameter is mainly a static parameter representing the inherent physical properties of the underlying surface. The following method is adopted to acquire the ground boundary parameter: Average ground roughness length (z0): Main method: airborne LiDAR scanning. High-precision digital elevation model (DEM) and digital surface model (DSM) of the base are obtained through flight operation. Micro-terrain method is used to calculate the roughness length value of each calculation grid based on the standard deviation of the elevation data. This method can obtain full-coverage and high-precision spatial distribution data.

[0158] Auxiliary method: For typical underlying surface types (such as smooth photovoltaic panels, Gobi desert, and shrubs), an array of micro resistance anemometers is laid out. The roughness is directly calculated by measuring the vertical wind speed profile, which is used to verify the remote sensing inversion result.

[0159] Average thermal conductivity and average blackbody emissivity (ε): Main method: field sampling and laboratory measurement. Typical ground materials (such as sand, gravel, photovoltaic panel glass, and concrete) are sampled in different subzone types (photovoltaic area, desert area, etc.) in the base. The thermal conductivity is measured in the laboratory by using the thermal probe method. The emissivity spectrum is measured by using the Fourier transform infrared spectrometer (FTIR) instrument, and the integral value is calculated to obtain the blackbody emissivity.

[0160] Spatial expansion: the laboratory measurement value is used as prior knowledge, which is cross-verified with the emissivity result of the hyperspectral remote sensing image. Finally, a weighted average parameter value based on the proportion of the ground cover type of each calculation grid is assigned.

[0161] S62, acquiring a label of a ground source parameter: The ground source parameter is a dynamic parameter representing the energy and mass exchange flux, and its acquisition needs to rely on high-frequency observation.

[0162] Sensible heat flux (H), latent heat flux (LE), carbon dioxide (CO2) flux, and methane (CH4) flux: Gold standard: Eddy Covariance (EC) system. Deploy multiple EC systems in representative areas (e.g. concentrated photovoltaic area, original desert area, green area) within the base. This system measures the pulsation of three-dimensional wind speed, temperature, water vapor and CO2 concentration at high frequency (10-20Hz) synchronously, and directly calculates the hourly or half-hourly average sensible heat, latent heat, CO2 and CH4 flux true value. This is the international standard method for flux observation.

[0163] Spatial representation: Deploy mobile flux observation systems (e.g. installed on vehicles) for short-term observation in different locations to make up for the lack of spatial coverage of fixed station observation, and to verify the spatial extrapolation ability of the model.

[0164] Source term of dust particle release rate: Main method: Large-scale automatic weighing dust cylinder network and laser dust instrument. Deploy a dense network of dust cylinders within the base, collect and weigh the settled dust material regularly, and convert it into time-integrated flux. At the same time, use the laser dust instrument to monitor the real-time PM10 / PM2.5 concentration changes in the air, combined with wind speed and direction data, to obtain the local dust release flux through the inverse diffusion model.

[0165] Control experiment: Set up a wind tunnel experiment device in a typical sand source area, and conduct wind erosion on surface soil with different water contents, directly measure the sand raising flux curve, and provide mechanism-based label data for the model.

[0166] Anthropogenic heat source and equipment VOCs release rate source: Anthropogenic heat source: Deploy heat flux plate sensors around power facilities such as booster stations and box transformers to directly measure the heat flux emitted by their outer surfaces to the environment. At the same time, obtain the real-time power, load rate and operating efficiency of the equipment from the Supervisory Control and Data Acquisition (SCADA) system, and calculate the theoretical heat dissipation through the heat balance equation, which is verified by the measured value.

[0167] Equipment VOCs release rate: Use portable gas chromatograph-mass spectrometer (GC-MS) to sample and analyze the air around running transformers and cable trenches, qualitatively and quantitatively identify VOCs species and concentration. Combined with computational fluid dynamics (CFD) simulation, the source intensity release rate is calculated.

[0168] Nitrous oxide (N2O), volatile organic compounds (VOCs) and other trace gas fluxes: Main method: static box-gas chromatography. Closed box is placed at typical locations (such as green belt after irrigation, near livestock farms) in the base, and the gas in the box is extracted regularly, and the concentration changes of N2O, VOCs and other gases are analyzed by high-precision gas chromatograph to calculate the flux. Although this method cannot be continuously observed, it can obtain high-precision data at key points and key periods.

[0169] S63, construct a sample label database: The labels of the ground boundary parameters and the labels of the ground source term parameters are all high-credibility data obtained by field measurement, remote sensing inversion and model calculation. Time synchronization and space gridding processing are performed, and the corresponding multi-temporal remote sensing image sequence input data are strictly matched to construct a sample label database for training and verifying the ground parameter model. The construction of the database ensures the scientificity and effectiveness of the supervised learning process, and is the basis for the high-precision prediction of the Transformer model.

[0170] In some embodiments, the present application further proposes to monitor cloud information by a full-sky imager, combine a high-precision solar trajectory tracking model, and use a Monte Carlo method to calculate the ground irradiance distribution under the condition of cloud cover, including: deploying a full-sky imager to continuously collect sky images of the monitoring area; pre-processing the sky images to obtain cloud information, the pre-processing including eliminating image blur and color distortion caused by dust aerosols based on the characteristics of the Shagehuang environment, and separating cloud and dust information; based on the cloud information, using image processing and inversion algorithms to determine the specific position and shape of the cloud in space; introducing a typical aerosol optical thickness model in the Shagehuang area as an input parameter for Monte Carlo ray tracing together with the optical properties of the cloud; using a solar trajectory tracking method to calculate the real-time position of the sun in the sky according to the geographical location, date and time; using a Buie distribution model as the shape model of the sun to describe the characteristics of sunlight with non-uniform intensity distribution; using the Monte Carlo method for ray tracing simulation to calculate the solar irradiance distribution at different positions on the ground of the Shagehuang new energy base under the condition of cloud cover.

[0171] The all-sky imager refers to a device capable of continuously shooting sky images with a hemispherical field of view, which can be realized by a CCD camera equipped with a fisheye lens, and dynamic change data of cloud coverage is obtained by shooting at regular intervals. The cloud information inversion refers to extracting cloud height, thickness and spatial distribution parameters from all-sky images based on image processing algorithms, which can be realized by threshold segmentation combined with a radiation transfer model, and is used to determine the range of cloud cover to solar radiation. The solar trajectory tracking method refers to a mathematical model for accurately calculating the azimuth and elevation angles of the sun according to geographic coordinates, date and time, which can be realized by the SOLPOS algorithm, and is used to simulate the real-time movement trajectory of the sun in the sky. The Buie distribution model refers to a mathematical model describing the non-uniform distribution of solar intensity on the sun, which can be represented by a ring Gaussian function to characterize the light intensity attenuation characteristics, and is used to more realistically simulate the intensity distribution of direct solar radiation. The Monte Carlo method light ray tracing refers to randomly sampling light paths and statistically analyzing the probability distribution of their interaction with the cloud layer, which can be realized by backward ray tracing combined with importance sampling, and is used to calculate the irradiance intensity of each point on the ground under the cloud cover.

[0172] Specifically, in the Shaguo Huang new energy base, the all-sky imager collects sky images at a frequency of once per minute, and the cloud information is obtained by preprocessing the sky images, which includes eliminating image blur and color distortion caused by dust aerosols based on the environmental characteristics of Shaguo Huang, and separating cloud and dust information; the cloud layer area is identified by gray threshold segmentation and morphological processing, and the optical thickness and spatial position of the cloud layer are inverted by combining the atmospheric radiation transfer equation, and the typical aerosol optical thickness model in the Shaguo Huang area is introduced as an input parameter of the Monte Carlo light ray tracing together with the optical properties of the cloud layer. The solar trajectory tracking model calculates the real-time azimuth and elevation angles of the sun according to the latitude and longitude of the base, the date and UTC time, and determines its projection position in the celestial coordinate system. The Buie distribution model regards the sun as a non-uniform light source composed of a central bright spot and a surrounding attenuated light ring, and uses a radial distribution function to simulate the light intensity gradient change. In the simulation process of the Monte Carlo light ray tracing, the cloud cover and scattering are simulated, and the absorption and scattering effects of dust aerosols on solar radiation are simulated, and the solar irradiance intensity distribution at different positions on the ground of the Shaguo Huang new energy base under the condition of cloud cover is calculated.

[0173] In some specific embodiments, the all-sky imager can be installed on the top of the flux tower at the center of the base at a height of 10 meters to avoid obstruction by ground obstacles. The cloud layer inversion algorithm can use a semantic segmentation model based on a convolutional neural network, and the training data includes all-sky images of the Shagehuang region in different seasons and cloud height data measured by a laser radar synchronously. When calculating the sun trajectory, the declination angle correction and atmospheric refraction compensation can be introduced to improve the calculation accuracy of the sun position to within 0.1 degrees. The parameter setting of the Buie distribution model can be adjusted according to the season, and a narrower light ring width is used in summer to match the strong sunlight conditions, and the light ring range is increased in winter to reflect the characteristics of enhanced diffuse light.

[0174] The present scheme can more realistically restore the spatial distribution difference of irradiance under complex meteorological conditions by dynamically monitoring the cloud layer morphology through the all-sky imager, accurately describing the solar intensity distribution characteristics by combining the Buie model, and simulating the interaction process of light and cloud layer by using the Monte Carlo method.

[0175] Through the above technical scheme, the present application solves the problem of insufficient calculation accuracy of ground irradiance in the Shagehuang region caused by rapid movement of clouds, and significantly improves the accuracy of setting the ground heat radiation source term in the LBM model by fusing dynamic cloud monitoring and non-uniform light source model. The present scheme can effectively capture the local shadow effect caused by cloud shading in the photovoltaic panel array area, provide high-precision irradiation input data for subsequent carbon flux simulation, adapt to the variable meteorological conditions in the Shagehuang region, and ensure the reliability of all-weather simulation.

[0176] Based on the same overall inventive concept, the present application also protects a carbon flux measurement inversion system for a Shagehuang new energy base. The carbon flux measurement inversion system for a Shagehuang new energy base provided by the present application is described below, and the carbon flux measurement inversion system for a Shagehuang new energy base described below can be mutually corresponding and referred to the carbon flux measurement inversion method for a Shagehuang new energy base described above.

[0177] Figure 2 is a structural schematic diagram of the carbon flux measurement inversion system for a Shagehuang new energy base provided by an embodiment of the present application, as Figure 2 indicated, the carbon flux measurement inversion system for a Shagehuang new energy base of the present embodiment includes a construction module 21, a generation module 22, a first setting module 23, a second setting module 24, a third setting module 25 and a solving module 26.

[0178] The construction module 21 is configured to obtain physical parameters, terrain data and image data of the Shagehuang new energy base, and establish a three-dimensional geometric model of the Shagehuang new energy base by using a three-dimensional reconstruction technology; The generation module 22 is configured to perform spatial lattice processing on the region represented by the three-dimensional geometric model to generate a calculation grid required by the LBM model. The first setting module 23 is configured to input the image data into a pre-trained ground parameter model for identification, and output ground boundary parameters and ground source term parameters corresponding to each grid unit in the calculation grid, so as to set the ground boundary condition of the LBM model and the ground material source term of the LBM model. The second setting module 24 is configured to monitor cloud information by using a full-sky imager, combine a high-precision sun trajectory tracking model, and calculate the ground irradiance intensity distribution under the condition of cloud layer shielding by using a Monte Carlo method, so as to set the ground thermal radiation source term of the LBM model. The third setting module 25 is configured to obtain input boundary parameters and input source term parameters of the atmosphere in the monitoring area, so as to set the atmospheric boundary condition of the LBM model and the atmospheric material source term of the LBM model. The solving module 26 is configured to solve the LBM model, obtain detailed simulation results of the carbon dioxide concentration distribution, flow velocity distribution, temperature distribution and pollutant concentration distribution in the Shagehuang new energy base, and calculate the carbon flux of the Shagehuang new energy base based on the detailed simulation results.

[0179] It should be noted that the related information involved in each embodiment of the present application is strictly in accordance with the requirements of laws and regulations, and follows the principles of legality, legitimacy and necessity, and is based on the reasonable purpose of the business scene, and processes the information provided by the user in the process of using the product / service or generated due to the use of the product / service, and the information authorized by the user.

[0180] The related information processed by the present application may vary depending on the specific product / service scene, and should be based on the specific scene of the user using the product / service. It may involve the user's account information, device information or other related information. The present application will treat the related information and its processing with a high degree of diligence.

[0181] The present application attaches great importance to the security of related information, and has taken security protection measures in accordance with industry standards, which are reasonable and feasible to protect related information from unauthorized access, public disclosure, use, modification, damage or loss.

[0182] The device embodiments described above are only schematic, and the units described as separate components can or can not be physically separated, and the components shown as units can or can not be physical units, i.e. they can be located in one place or distributed on multiple network units. Part or all of the modules can be selected to achieve the purpose of the present embodiment scheme according to actual needs. Those skilled in the art can understand and implement it without creative labor.

[0183] Those skilled in the art can clearly understand the technical solutions of the various embodiments from the above description of the embodiments, and the various embodiments can be implemented by means of software with the necessary general hardware platforms, and of course, can also be implemented by hardware. Based on such understanding, the above technical solutions, essentially or in other words, the part of the prior art that makes a contribution, can be embodied in the form of a software product, which can be stored in a computer readable storage medium, such as a ROM / RAM, a magnetic disk, an optical disk, and the like, and includes a number of instructions for causing a computer device (which can be a personal computer, a server, or a network device, etc.) to execute the methods described in the various embodiments or some parts of the embodiments.

[0184] Finally, it should be noted that: the above embodiments are only used to illustrate the technical solutions of the present application, and not to limit them; although the present application has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that: it can still modify the technical solutions recorded in the foregoing embodiments, or make equivalent replacement for some technical features therein; and these modifications or replacements do not make the essence of the corresponding technical solutions deviate from the spirit and scope of the technical solutions of the embodiments of the present application.

Claims

1. A method for carbon flux measurement and inversion in a desert renewable energy base, characterized in that, include: The physical parameters, terrain data, and image data of the Shagohuang New Energy Base were obtained, and a three-dimensional geometric model of the Shagohuang New Energy Base was established using three-dimensional reconstruction technology. The region represented by the three-dimensional geometric model is spatially latticized to generate the computational grid required for the LBM model; The image data is input into a pre-trained ground parameter model for identification, and the ground boundary parameters and ground source term parameters corresponding to each grid cell in the computational grid are output to set the ground boundary conditions and ground material source terms of the LBM model. By monitoring cloud information with an all-sky imager and combining it with a high-precision solar trajectory tracking model, the Monte Carlo method is used to calculate the ground irradiance distribution under cloud cover conditions, so as to set the ground thermal radiation source term of the LBM model. The input boundary parameters and input source term parameters of the atmosphere within the monitoring area are obtained in order to set the atmospheric boundary conditions and atmospheric material source terms of the LBM model. Solve the LBM model to obtain detailed simulation results of carbon dioxide concentration distribution, flow velocity distribution, temperature distribution and pollutant concentration distribution in the Shagohuang New Energy Base, and calculate the carbon flux of the Shagohuang New Energy Base based on the detailed simulation results.

2. The carbon flux measurement and inversion method for the Shagohuang new energy base according to claim 1, characterized in that, Based on the detailed simulation results, the carbon flux of the Shagohuang new energy base is calculated, including: The Shagohuang New Energy Base was divided into several zones, resulting in multiple zone types. For each partition type, the first carbon flux is calculated using the gradient diffusion method based on the carbon dioxide concentration distribution and the flow velocity distribution, and the second carbon flux is calculated using the eddy covariance method. The first carbon flux and the second carbon flux are weighted and summed to obtain the basic carbon flux corresponding to each partition type; Based on the correction factor for each partition type, the basic carbon flux is corrected to obtain the corrected carbon flux for each partition type; The carbon flux of the Shagohuang New Energy Base is obtained by summing the area-weighted values ​​of the modified carbon flux for all partition types and then dividing the sum by the total area.

3. The carbon flux measurement and inversion method for the Shagohuang new energy base according to claim 2, characterized in that, The Shagohuang New Energy Base was divided into several zones, resulting in multiple zone types, including: Acquire multi-source data covering the Shagohuang New Energy Base; wherein, the multi-source data includes at least one of digital elevation model data, multispectral remote sensing imagery, hyperspectral remote sensing imagery, wind turbine coordinates and wind resource meteorological data, and equipment deployment drawings; The digital elevation model data is preprocessed, the terrain complexity index of each computational grid is calculated, and the computational grid is initially partitioned based on a preset terrain threshold to obtain the first partition result. Radiometric calibration and atmospheric correction are performed on the multispectral remote sensing image. The normalized vegetation index value is calculated pixel by pixel. The vegetation cover level is divided according to the preset vegetation cover level. The calculation grid is initially partitioned based on the vegetation cover level to obtain the second partition result. Radiometric calibration and atmospheric correction are performed on the hyperspectral remote sensing image. Characteristic high reflectance peaks of photovoltaic panels in the near-infrared band are extracted. Feature enhancement is performed on the characteristic high reflectance peaks. Photovoltaic panel pixels are identified by combining the preset reflectance peak threshold. The preliminary range of photovoltaic power generation area is formed by post-processing. The computational grid is preliminarily partitioned to obtain the third partition result. Using the wind turbine coordinates, the wind resource meteorological data, and the digital elevation model data, a customized wake model combined with a wind speed loss threshold is used to calculate the wind turbine wake influence range, and the computational grid is initially partitioned to obtain the fourth partition result. The coordinates of the facility elements in the equipment deployment drawings are transformed to a coordinate system consistent with the multi-source data, a facility element vector layer is generated, and the computational grid is initially partitioned to obtain the fifth partition result; Based on the results of the first to the fifth partitioning, a comprehensive logical judgment is performed on each computing grid to determine the final partition type of each computing grid.

4. The carbon flux measurement and inversion method for the Shagohuang new energy base according to claim 3, characterized in that, Calculate the terrain complexity index for each computational grid, and perform preliminary partitioning of the computational grid based on a preset terrain threshold to obtain the first partitioning result, including: The slope value is calculated within a specified first moving window using the maximum elevation difference method or the Horn algorithm; the aspect value is calculated within the first moving window using a preset aspect calculation formula; the terrain undulation is calculated within a specified second moving window using a preset undulation calculation formula; and the first partition result is determined based on the slope value, the aspect value, and the terrain undulation. The multispectral remote sensing image is radiometrically calibrated and atmospherically corrected. The normalized vegetation index (NDI) is calculated pixel-by-pixel. Vegetation cover levels are then defined based on preset vegetation cover thresholds. Combined with these vegetation cover levels, the computational grid is preliminarily partitioned to obtain a second partitioning result, including: Using a preset normalized vegetation index (NVI) calculation formula, the NVI value is calculated, and based on the range to which the NVI value belongs, the vegetation cover level and the second zoning result corresponding to the vegetation cover level are determined.

5. The carbon flux measurement and inversion method for the Shagohuang new energy base according to claim 3, characterized in that, Using the wind turbine coordinates, the wind resource meteorological data, and the digital elevation model data, a customized wake model combined with a wind speed deficit threshold is employed to calculate the wind turbine wake influence range. The computational grid is then preliminarily partitioned to obtain the fourth partition result, including: Centered on each wind turbine, along the dominant and secondary wind directions, the farthest distance that can be reached when the wind speed loss reaches the wind speed loss threshold is calculated according to the customized wake model for each wind direction sector as the dynamic wake range. Based on the dynamic wake range, a corresponding wind direction sector buffer is generated for each wind turbine; Spatially overlay all wind direction sector buffer zones of all wind turbines to generate a comprehensive wake influence zone layer for the entire base; The results of the fourth partition are obtained by overlaying the comprehensive wake influence zone layer with the computational grid. The function corresponding to the customized wake model is: in, This represents the wind speed at a distance x downwind. Indicates the incoming wind speed. This indicates the thrust coefficient of the wind turbine. This represents the rotor diameter of the wind turbine, and x represents the downwind distance. This represents the wake attenuation coefficient, which is dynamically adjusted based on atmospheric stability.

6. The carbon flux measurement and inversion method for the Shagohuang new energy base according to claim 3, characterized in that, Also includes: Based on the differentiated settings for each zone type and meteorological interference factors, the probability distribution models of the eddy diffusion coefficient, the vertical concentration gradient of carbon dioxide, and the correction coefficient are obtained; wherein, the meteorological interference factors include at least aerosol optical interference caused by strong winds and sandstorms, and intense near-surface thermal turbulence caused by strong solar radiation; Monte Carlo random perturbation iteration is performed based on the probability distribution model with the aforementioned differential settings to obtain multiple candidate carbon fluxes; After sorting the multiple candidate carbon fluxes, the top n and bottom m candidate carbon fluxes are removed. The minimum and maximum candidate carbon fluxes among the remaining candidate carbon fluxes constitute the confidence interval of the carbon flux of the Shagohuang New Energy Base.

7. The carbon flux measurement and inversion method for the Shagohuang New Energy Base according to any one of claims 2-6, characterized in that, The second carbon flux was calculated using the eddy covariance method, including: Based on the atmospheric turbulence fluctuation characteristics, high-frequency data from multiple data points within a preset time period are extracted; wherein, the sampling time interval of the high-frequency data is a preset time interval, and the high-frequency data includes high-frequency vertical flow velocity extracted from the flow velocity distribution, and carbon dioxide concentration extracted from the carbon dioxide concentration distribution; The high-frequency data sequence is quality marked and preprocessed, including at least removing abnormal values ​​of carbon dioxide concentration signal caused by aerosol concentration abnormalities due to strong winds and sandstorms in the desert environment. The high-frequency vertical velocity of each data point in the sequence is subtracted from the mean vertical velocity to obtain the pulsation value of the high-frequency vertical velocity of each data point. The carbon dioxide concentration of each data point is subtracted from the mean carbon dioxide concentration of all data points to obtain the pulsation value of the carbon dioxide concentration of each data point. The covariance is calculated as the second carbon flux based on the high-frequency vertical velocity fluctuation value and the carbon dioxide concentration fluctuation value of each data point.

8. The carbon flux measurement and inversion method for the Shagohuang new energy base according to claim 1, characterized in that, The image data is a multi-temporal remote sensing image sequence covering the Shagohuang New Energy Base; the ground parameter model is a spatiotemporal sequence prediction model built based on the Transformer architecture. The image data is input into a pre-trained ground parameter model for identification, and the model outputs ground boundary parameters and ground source term parameters corresponding to each grid cell in the computational grid, including: By processing the multi-temporal remote sensing image sequence, the dynamic change characteristics of surface cover over time in each region are extracted, and the ground boundary parameters and ground source term parameters corresponding to each computational grid are output. The ground source term parameter is a dynamic parameter that changes over time, and the model also outputs the time variation pattern of this parameter. The process of constructing the sample label library for the ground parameter model includes: Acquire the labels for the ground boundary parameters; the ground boundary parameters are static parameters that characterize the inherent physical properties of the underlying surface. The labels for obtaining surface source term parameters are dynamic parameters characterizing energy and mass exchange fluxes, and their acquisition depends on high-frequency observations. Constructing a sample label database: The data with a confidence level greater than the preset level obtained from field measurement, remote sensing inversion and model calculation of the labels of the ground boundary parameters and the labels of the ground source parameters are time-synchronized and spatially gridded, and matched with the input data of the corresponding multi-temporal remote sensing image sequence to form a sample label database for training and validation of the ground parameter model.

9. The carbon flux measurement and inversion method for the Shagohuang new energy base according to claim 1, characterized in that, By monitoring cloud information using an all-sky imager and combining it with a high-precision solar trajectory tracking model, the Monte Carlo method was employed to calculate the ground irradiance distribution under cloud cover conditions, including: By deploying the all-sky imager, sky images of the monitored area are continuously acquired; The sky image is preprocessed to obtain cloud information. The preprocessing includes eliminating image blur and color distortion caused by sand and dust aerosols based on the characteristics of the desert environment, and separating cloud and sand and dust information. Based on the cloud information, image processing and inversion algorithms are used to determine the specific location and shape of the cloud in space. A typical aerosol optical thickness model for the Shago desert region is introduced and used together with the cloud optical properties as input parameters for Monte Carlo ray tracing. The solar trajectory tracking method is used to calculate the real-time position of the sun in the sky based on geographical location, date, and time; The Buie distribution model is used as the shape model of the sun to describe the characteristics of sunlight with a non-uniform light intensity distribution; The Monte Carlo method was used to simulate ray tracing, which simulated the obstruction and scattering of clouds, as well as the absorption and scattering effect of sand and dust aerosols on solar radiation. The distribution of solar irradiance intensity at different locations on the ground of the Shagohuang New Energy Base was calculated under the presence of clouds.

10. A carbon flux measurement and inversion system for a desert-based new energy base, characterized in that, include: The construction module is used to acquire the physical parameters, terrain data and image data of the Shagohuang New Energy Base, so as to establish a three-dimensional geometric model of the Shagohuang New Energy Base using three-dimensional reconstruction technology. The generation module is used to perform spatial lattice processing on the region represented by the three-dimensional geometric model to generate the computational grid required for the LBM model; The first setting module is used to input the image data into a pre-trained ground parameter model for identification, and output ground boundary parameters and ground source term parameters corresponding to each grid cell in the computational grid, so as to set the ground boundary conditions and ground material source terms of the LBM model. The second setting module is used to monitor cloud information through an all-sky imager, combine it with a high-precision solar trajectory tracking model, and use the Monte Carlo method to calculate the ground irradiance distribution under cloud cover conditions, so as to set the ground thermal radiation source item of the LBM model. The third setting module is used to obtain the input boundary parameters and input source term parameters of the atmosphere within the monitoring area, so as to set the atmospheric boundary conditions and atmospheric material source terms of the LBM model. The solution module is used to solve the LBM model, obtain detailed simulation results of carbon dioxide concentration distribution, flow velocity distribution, temperature distribution and pollutant concentration distribution in the Shagohuang New Energy Base, and calculate the carbon flux of the Shagohuang New Energy Base based on the detailed simulation results.

Images