Grassland ecological restoration project benefit evaluation system based on casa model

Abstract

This invention relates to the field of ecological restoration engineering management technology, and particularly to a grassland ecological restoration engineering benefit evaluation system based on the CASA model. The system includes the following modules: a data acquisition module, a vegetation index calculation module, a regulation factor calculation module, a productivity assessment module, and a benefit assessment module. The data acquisition module collects remote sensing images, meteorological observation data, and surface radiation data of the target area. The vegetation index calculation module calculates the normalized vegetation index for each dynamic observation unit. The regulation factor calculation module generates a set of vegetation growth parameters based on meteorological observation data, surface radiation data, and the normalized vegetation index. The productivity assessment module inputs the vegetation growth parameter set into the CASA model to generate net vegetation productivity. The benefit assessment module compares the net vegetation productivity to evaluate the benefits of the grassland ecological restoration project. This invention improves the accuracy of productivity estimation and the reliability of benefit evaluation results.

Description

Technical Field

[0001] This invention relates to the field of ecological restoration engineering management technology, and in particular to a grassland ecological restoration engineering benefit evaluation system based on the CASA model. Background Technology

[0002] Early evaluations of grassland ecological restoration projects typically employed single or limited statistical indicators, such as field surveys of vegetation cover, aboveground biomass, or grass yield changes, to assess the effectiveness of ecological restoration. While these methods can reflect localized ecological changes to some extent, their limited data collection scope, long timeframes, and high labor costs make it difficult to achieve continuous monitoring and comprehensive evaluation of large-scale grassland areas. Furthermore, these methods are usually based on static data, failing to reflect the dynamic changes in the ecosystem across different time scales.

[0003] With the development of remote sensing technology, geographic information systems, and ecological models, researchers are increasingly using remote sensing imagery to monitor grassland vegetation growth and estimate regional vegetation productivity through indicators such as vegetation indices. Among these, the Carnegie-Ames-Stanford Approach (CASA model) estimates net primary productivity by comprehensively utilizing vegetation indices, solar radiation data, and meteorological data obtained through remote sensing. This allows it to reflect changes in vegetation photosynthesis and ecosystem productivity at a regional scale. This model has already been applied to productivity estimation in forest, farmland, and grassland ecosystems, providing a new technical means for dynamic monitoring of grassland ecosystems.

[0004] However, in the actual management of grassland ecological restoration projects, existing evaluation methods mostly focus on data analysis of a single ecological indicator or a single time point, lacking the ability to comprehensively evaluate multiple factors such as ecosystem productivity, vegetation restoration degree and environmental changes, making it difficult to fully reflect the overall benefits of ecological restoration projects. Summary of the Invention

[0005] Therefore, it is necessary for the present invention to provide a grassland ecological restoration engineering benefit evaluation system based on the CASA model to solve at least one of the above-mentioned technical problems.

[0006] To achieve the above objectives, a grassland ecological restoration project benefit evaluation system based on the CASA model includes the following modules: The data acquisition module is used to acquire remote sensing images of the target area during the evaluation period through the remote sensing data platform, and to collect meteorological observation data and surface radiation data for the corresponding time period. The vegetation index calculation module is used to calculate the normalized vegetation index of each preset dynamic observation unit based on the near-infrared and red light band information in the remote sensing image. The regulation factor calculation module is used to generate effective environmental regulation factors for the target area based on meteorological observation data and surface radiation data, and to generate a set of vegetation growth parameters by combining the normalized vegetation index. The productivity assessment module is used to input the set of vegetation growth parameters into a preset CASA model to determine the net productivity of vegetation within the evaluation period. The benefit assessment module is used to compare the net vegetation productivity before and after the implementation of the grassland ecological restoration project within the evaluation period, determine the changes in productivity of each dynamic observation unit in different time periods, and evaluate the benefits of the grassland ecological restoration project.

[0007] This application integrates remote sensing imagery, meteorological observation data, and surface radiation data to achieve high-precision monitoring and analysis of vegetation dynamics in target areas. This allows for accurate acquisition of the normalized difference in vegetation index, environmental moderating factors, and photosynthetically active radiation absorption (PARA) for each dynamic observation unit within the evaluation period. Furthermore, the system calculates effective net primary productivity (NPMP) and cumulative NPV using the CASA model, providing a quantitative basis for the effectiveness of ecological restoration projects. Simultaneously, by comparing the productivity differences before and after restoration projects, and considering the spatial proportions of effective and favorable ecological restoration areas within the dynamic observation units, the system can scientifically calculate the productivity contribution of grassland ecological restoration projects. This provides quantifiable, comparable, and spatially applicable results for evaluating the benefits of ecological restoration projects. Furthermore, by utilizing a dynamic observation unit division and update mechanism, vegetation growth rules and growth stage determination, and the fusion optimization of photosynthetically active radiation absorption and environmental regulation parameters, the system enables the model to fully consider the growth characteristics of different vegetation types, changes in environmental conditions, and spatial heterogeneity. This improves the accuracy of productivity estimation and the reliability of evaluation results. Compared with traditional single-indicator or static evaluation methods, it can achieve intelligent, dynamic, and quantitative evaluation of grassland ecological restoration projects, providing technical support for scientific decision-making. At the same time, it has broad application value in ecological protection, land management, and optimization of restoration strategies. Attached Figure Description

[0008] Other features, objects, and advantages of the invention will become more apparent from the following detailed description of non-limiting embodiments with reference to the accompanying drawings: Figure 1 This is a schematic diagram of the execution process of the grassland ecological restoration engineering benefit evaluation system based on the CASA model of the present invention; Figure 2 This is a schematic diagram showing the distribution of dynamic observation units in an embodiment of the present invention; Figure 3This is a block diagram of the grassland ecological restoration engineering benefit evaluation system based on the CASA model of the present invention; The realization of the objective, functional features and advantages of the present invention will be further explained in conjunction with the embodiments and with reference to the accompanying drawings. Detailed Implementation

[0009] The technical method of the present invention will now be clearly and completely described with reference to the accompanying drawings. Obviously, the described embodiments are only some, not all, of the embodiments of the present invention. All other embodiments obtained by those skilled in the art based on the embodiments of the present invention without inventive effort are within the scope of protection of the present invention.

[0010] Furthermore, the accompanying drawings are merely illustrative of the invention and are not necessarily drawn to scale. The same reference numerals in the drawings denote the same or similar parts, and therefore repeated descriptions of them will be omitted. Some block diagrams shown in the drawings are functional entities and do not necessarily correspond to physically or logically independent entities. These functional entities can be implemented in software, in one or more hardware modules or integrated circuits, or in different network and / or processor methods and / or microcontroller methods.

[0011] It should be understood that although the terms "first," "second," etc., may be used herein to describe various units, these units should not be limited by these terms. These terms are used merely to distinguish one unit from another. For example, without departing from the scope of the exemplary embodiments, a first unit may be referred to as a second unit, and similarly, a second unit may be referred to as a first unit. The term "and / or" as used herein includes any and all combinations of one or more of the associated listed items.

[0012] To achieve the above objectives, please refer to Figures 1 to 3 This invention provides a grassland ecological restoration engineering benefit evaluation system based on the CASA model, the system comprising the following modules: The data acquisition module 101 is used to acquire remote sensing images of the target area during the evaluation period through the remote sensing data platform, and to collect meteorological observation data and surface radiation data for the corresponding time period. The vegetation index calculation module 102 is used to calculate the normalized vegetation index of each preset dynamic observation unit based on the near-infrared and red light band information in the remote sensing image. The adjustment factor calculation module 103 is used to generate effective environmental adjustment factors for the target area based on meteorological observation data and surface radiation data, and to generate a set of vegetation growth parameters by combining the normalized vegetation index. Productivity assessment module 104 is used to input the set of vegetation growth parameters into a preset CASA model to determine the net productivity of vegetation within the evaluation period. The benefit assessment module 105 is used to compare the net vegetation productivity before and after the implementation of the grassland ecological restoration project within the evaluation period, determine the changes in productivity of each dynamic observation unit in different time periods, and evaluate the benefits of the grassland ecological restoration project.

[0013] The grassland ecological restoration project benefit evaluation system based on the CASA model performs the following steps: S1: Obtain remote sensing images of the target area within the evaluation period through a remote sensing data platform, and collect meteorological observation data and surface radiation data for the corresponding time period; In one embodiment of the present invention, multiple remote sensing images of the target area within an evaluation period (1 year) are acquired by connecting to a remote sensing data platform. Multispectral remote sensing images with a spatial resolution of 30m can be selected as the basic data, and continuous image sequences are acquired at 5-day time intervals. At the same time, temperature data and precipitation data corresponding to each image time node are collected from meteorological observation stations. The temperature data is recorded in the form of daily average temperature, and the precipitation data is recorded in the form of daily cumulative precipitation. In addition, the total solar radiation data of the target area is acquired through surface radiation observation equipment, and the meteorological observation data and radiation data are time-synchronized according to the imaging time of the remote sensing images.

[0014] S2: Based on the near-infrared and red light band information in the remote sensing images, and to calculate the normalized vegetation index of each preset dynamic observation unit; In a further embodiment, a pixel-level normalized vegetation index is calculated based on near-infrared and red band information from the remote sensing image, wherein the near-infrared band reflectance and the red band reflectance are derived from the corresponding spectral channels of the remote sensing image, and are calculated according to... The calculation is performed in the following manner: Then, the target area is initially divided into grids with a basic grid scale of 60m × 60m, and the pixel affiliation is determined based on the pixel center coordinates within each grid cell; after completing the pixel aggregation, the pixel affiliation within each grid cell is determined... The values ​​are statistically processed, and the grid cells are split or merged according to the variance of the vegetation index to form dynamic observation cells. Then, the mean NDVI of the pixels in the formed dynamic observation cells is statistically analyzed to obtain the normalized vegetation index data of each dynamic observation cell at each time node.

[0015] S3: Generate effective environmental regulation factors for the target area based on meteorological observation data and surface radiation data, and generate a set of vegetation growth parameters by combining the normalized vegetation index. In a further embodiment, temperature and precipitation data corresponding to each time point in meteorological observation data are extracted, and the suitable temperature range (e.g., 10℃~25℃) for the main herbaceous vegetation in the target area is determined according to vegetation growth rules. The temperature regulation coefficient is calculated by comparing the actual temperature with the suitable temperature range. Subsequently, the soil moisture supply capacity is calculated based on precipitation data, and the moisture regulation coefficient is represented by a normalized coefficient between 0 and 1. At the same time, the total solar radiation is calculated based on surface radiation data and converted into the photosynthetically active radiation ratio at a ratio of approximately 0.5. Finally, the temperature regulation coefficient, moisture regulation coefficient, photosynthetically active radiation ratio, and normalized vegetation index of each dynamic observation unit are integrated to form a vegetation growth parameter set.

[0016] S4: Input the set of vegetation growth parameters into the preset CASA model to determine the net productivity of vegetation within the evaluation period; In a further embodiment, a set of vegetation growth parameters is input into a preset CASA model to calculate the net vegetation productivity within the evaluation period. The CASA model is constructed based on the relationship between net primary productivity and absorbed photosynthetically active radiation (PARF). Its calculation structure consists of two parts: PARF absorption and light energy utilization efficiency. PARF absorption is obtained by multiplying the PARF ratio calculated using the normalized vegetation index (NDI) with the PARF data. Light energy utilization efficiency is determined by the maximum light energy utilization capacity parameter, the temperature regulation coefficient, and the water regulation coefficient. The maximum light energy utilization capacity parameter is set according to the main grassland type in the target area. After completing the above parameter input, the productivity of each dynamic observation unit is calculated on a monthly scale to obtain the vegetation net productivity data of each dynamic observation unit within the evaluation period.

[0017] S5: Compare the net vegetation productivity before and after the implementation of the grassland ecological restoration project within the evaluation period, determine the changes in productivity of each dynamic observation unit in different time periods, and evaluate the benefits of the grassland ecological restoration project.

[0018] In a further embodiment, the restoration benefits are evaluated by comparing the net vegetation productivity before and after the implementation of the grassland ecological restoration project. Specifically, firstly, the implementation time period of the restoration project is identified based on the pixel grayscale changes in the remote sensing image sequence, and 12 months before and after the implementation time point are used as comparison time windows. Subsequently, the net vegetation productivity of each dynamic observation unit is averaged in the time periods before and after implementation, and the productivity values ​​before and after implementation are generated respectively. Based on this, the difference between the two is calculated to obtain the change in productivity, and spatial statistical processing is performed on each dynamic observation unit according to the change in productivity to obtain the productivity changes of different units in the target area, thereby forming the evaluation result of the grassland ecological restoration project benefits.

[0019] Optionally, the method for dividing the dynamic observation units in the vegetation index calculation module includes: The basic grid scale is determined based on the spatial resolution of the remote sensing image, and the target area is divided into several grid units in combination with the geographic coordinate system of the remote sensing image. In this embodiment, a multispectral image with a spatial resolution of 30 m is selected as input data, and the target area is divided into initial grid units of 30m×30m based on image pixels. Combined with the geographic coordinate system information of the image, each grid unit is mapped to the geographic coordinate plane to ensure that the grid units continuously cover the target area in geographic space. At the same time, the coordinates of the upper left corner and the center point of each grid unit are recorded.

[0020] The assigned pixels of each grid unit are determined based on the spatial coordinates of the center point of the remote sensing image pixels, and the assigned pixels of each grid unit are then collected. In a further embodiment, the assigned pixel of each grid cell is determined based on the spatial coordinates of the pixel center point in the remote sensing image. The Euclidean distance between the center point of each pixel and the center point of the initial grid cell can be calculated, and the nearest pixel is assigned to the corresponding grid cell. A pixel aggregation table is generated within the grid cell, recording the pixel number and its NDVI value contained in each grid cell. During the aggregation process, pixels that are missing or obscured by clouds or fog in the image are removed and marked as null values ​​in the aggregation table to ensure the effectiveness of subsequent statistical analysis.

[0021] Based on the vegetation index variance of each grid cell in the aggregation processing results, grid cell splitting and merging are performed to obtain dynamic observation cells.

[0022] In a further embodiment, the variance of NDVI of the pixels in each grid cell is calculated. When the variance is greater than 0.05, the grid cell is uniformly divided into four sub-cells along the X and Y directions. If the variance is less than or equal to 0.02, adjacent grid cells with variances less than the threshold are merged in a 3×3 neighborhood to form a dynamic observation cell. During the splitting and merging process, the pixel number, center point coordinates and mean NDVI of each newly generated dynamic observation cell are recorded to form a dynamic observation cell table that can be directly used for subsequent vegetation index statistical calculations.

[0023] Optionally, performing grid cell splitting and merging includes: The vegetation index within each grid cell is calculated based on the near-infrared and red light band information of each grid cell in the collected processing results. In this embodiment, the Normalized Difference Vegetation Index (NDVI) formula is used based on the near-infrared (NIR) and red (RED) reflectance within the aggregated grid cells. Calculate the average NDVI value for each grid cell; during the calculation, exclude band data from clouds or invalid pixels, and perform a weighted average on the remaining pixels to obtain a representative NDVI value within the grid cell.

[0024] Calculate the vegetation index variance within each grid cell. If the vegetation index variance is greater than a preset variance threshold, then split the corresponding grid cell into several grid cells to be merged. In a further embodiment, the variance of NDVI pixels in each grid cell is calculated, and a variance threshold is set to 0.05. When the NDVI variance of the grid cell is greater than the threshold, the grid cell is uniformly divided into four sub-cells to be merged along the X and Y directions, and each sub-cell inherits the pixel number and spatial coordinate information of the original grid cell.

[0025] If the variance of the vegetation index is less than or equal to the variance threshold, the corresponding grid cell will be treated as a dynamic observation cell. In another embodiment, when the NDVI variance of a grid cell is less than or equal to the threshold of 0.05, the grid cell is directly used as a dynamic observation cell, and its center point coordinates, NDVI mean and variance are recorded.

[0026] Calculate the spatial gradient of vegetation index of the grid cell to be merged within a preset neighborhood range, and merge adjacent grid cells to be merged with spatial gradients less than or equal to a preset spatial gradient threshold as a dynamic observation unit.

[0027] In a further embodiment, the NDVI spatial gradient of the grid cell to be merged within a 3×3 neighborhood is calculated, and the spatial gradient threshold is set to 0.02. When the NDVI spatial gradient of adjacent grid cells to be merged is less than or equal to the threshold, these grid cells are merged to form the final dynamic observation cell, and the cell list, center point coordinates and average NDVI value of the merged cell are updated to generate a dynamic observation cell table.

[0028] Figure 2 This is a schematic diagram of the distribution of dynamic observation units in an embodiment of the present invention; as shown below. Figure 2 As shown, the area circled by each red box is a dynamic observation unit; the un-circled areas in the remote sensing image frame are adjacent grid units to be merged where the spatial gradient is greater than the spatial gradient threshold.

[0029] It is worth noting that the dynamic observation unit also includes an update mechanism, specifically: periodically updating existing dynamic observation units based on the latest acquired remote sensing imagery and meteorological data. Specifically, firstly, the mean and variance of NDVI for each pixel within each dynamic observation unit are recalculated, and the uniformity within the unit is judged according to a preset variance threshold (e.g., 0.05). If the variance exceeds the threshold, the unit is split into several sub-units along the X and Y directions, and the center point coordinates and pixel list of each sub-unit are re-recorded. For units with variance below the threshold, their NDVI spatial gradient in a 3×3 neighborhood is calculated. If the gradient is less than the spatial gradient threshold (e.g., 0.02), adjacent units are merged to form a new dynamic observation unit. After the update, each dynamic observation unit contains the latest pixel set, center coordinates, and average NDVI value.

[0030] Optionally, the effective environmental regulation factors for the target area generated in the regulation factor calculation module include: Extract temperature and precipitation data from meteorological observation data corresponding to the time nodes of remote sensing images; calculate the photosynthetically active radiation ratio of the target area at each time node based on surface radiation data; In this embodiment, the daily average temperature (unit: °C) and daily cumulative precipitation (unit: mm) of the target area at the corresponding time points in the remote sensing image are obtained from the regional meteorological observation network via a data interface. For missing or abnormal data points, the moving average of the most recent three days is used to fill the gaps, and the data is mapped to the center point of the corresponding dynamic observation unit according to the pixel spatial coordinates. The total solar radiation (unit: mm) of the target area at each time point is obtained based on the surface radiation data. The system then obtains the corresponding geographic coordinates, solar altitude angle, and atmospheric transmittance information. Subsequently, the total solar radiation is multiplied by the spectral coefficient. (Weighted ratio of NIR and red light bands for vegetation photosynthetic activity) to calculate the photosynthetically active radiation (PAR) value for each pixel. For multiple pixels within the same dynamic observation unit, the average aggregation method is used to calculate the PAR absorption ratio of that unit at that time point, i.e., the ratio of PAR to total solar radiation, and a time series matrix is ​​generated. To reduce the impact of short-term cloud cover, a 3-point moving average of the PAR values ​​from two consecutive days is used to ensure a smooth and continuous PAR ratio.

[0031] The temperature data at each time point are compared with the suitable temperature range for the vegetation type in the target area determined according to the preset vegetation growth rules to obtain the temperature regulation coefficient corresponding to each time point. In a further embodiment, the temperature data at each time point is compared with the suitable temperature range (e.g., 15–25℃) of the dominant vegetation type within the dynamic observation unit. When the temperature falls within the suitable range, the temperature adjustment coefficient is set to 1; if it is higher or lower than the suitable range, it is linearly reduced according to the deviation, with a reduction ratio not exceeding 0.5, to reflect the inhibitory effect of temperature on photosynthesis.

[0032] The water supply capacity is calculated based on the precipitation data at each time point to determine the water regulation coefficient corresponding to each time point. In a further embodiment, the water supply capacity is calculated based on the precipitation at each time point and the soil moisture content of the dynamic observation unit (e.g., 200 mm). When the proportion of precipitation to soil moisture content is... When the ratio is less than 0.8, the moisture adjustment coefficient is taken as 1; when the ratio is less than 0.8, it is reduced proportionally. The interval linear distribution generates a time series moisture adjustment coefficient.

[0033] Temperature regulation coefficient, moisture regulation coefficient, and photosynthetically active radiation ratio are integrated into environmental regulation factors, and these environmental regulation factors are spatially mapped to dynamic observation units.

[0034] In a further embodiment, the temperature regulation coefficient, moisture regulation coefficient, and photosynthetically active radiation ratio are integrated into an environmental regulation factor by weighted averaging (the weights can be set to 0.4, 0.3, and 0.3), and mapped to each observation unit through the center coordinates of the dynamic observation unit to form a spatial distribution parameter matrix that can be directly input into the CASA model.

[0035] It is worth noting that vegetation growth rules are obtained through literature review, field surveys, and national or local grassland ecological databases to identify the growth characteristics of common vegetation types in the target area, including species classification, typical growth cycles (budding period, vigorous growth period, dormancy period), suitable temperature range (°C), soil moisture content range (mm), and maximum light energy utilization capacity (unit: Subsequently, the growth stages, suitable environmental conditions, and light energy utilization parameters of each vegetation type were compiled into a structured data table including vegetation type, growth stage, temperature range, moisture range, and light energy utilization capacity.

[0036] Optionally, the method for constructing the CASA model in the productivity assessment module includes: The normalized vegetation index was used to construct the vegetation absorption of photosynthetically active radiation ratio parameter, and the photosynthetically active radiation absorption of each dynamic observation unit at different time points was determined in combination with the photosynthetically active radiation ratio. In this embodiment, a vegetation absorption ratio parameter is constructed based on the empirical relationship between vegetation index and vegetation canopy absorption capacity. The vegetation absorption ratio can be calculated using the NDVI linear transformation relationship. First, the lower and upper limits of the vegetation index for the target area are statistically obtained from long-term ecological monitoring samples, where NDVI_min is set to 0.1 (corresponding to bare soil or very low cover surface) and NDVI_max is set to 0.8 (corresponding to high-coverage herbaceous vegetation). Then, interval normalization is performed on the NDVI values ​​of each dynamic observation unit, converting the NDVI values ​​to a vegetation absorption capacity coefficient in the range of 0 to 1. When the NDVI value is less than NDVI_min, the photosynthetically active radiation ratio is set to 0; when the NDVI value is greater than NDVI_max, the photosynthetically active radiation ratio is set to 1; when the NDVI is between the two, interpolation is performed according to a linear ratio to obtain the vegetation absorption ratio parameter. Subsequently, this ratio is multiplied by the photosynthetically active radiation ratio corresponding to the aforementioned time node to obtain the photosynthetically active radiation absorption of each dynamic observation unit at different time nodes. Extract the maximum light energy utilization parameters corresponding to the growth characteristics of different vegetation types in the vegetation growth rules, and merge the maximum light energy utilization parameters into the environmental regulation factors to generate a set of environmental regulation parameters. In a further embodiment, growth characteristic parameters corresponding to each vegetation type are extracted from vegetation growth rules. For the herbaceous vegetation type (such as typical grassland vegetation) of the target area, its corresponding maximum light energy utilization parameter is selected, for example, set as... Subsequently, this parameter was integrated with the temperature regulation coefficient, moisture regulation coefficient, and photosynthetically active radiation ratio, and an environmental regulation parameter set was constructed according to the data format of observation unit number—time node—temperature coefficient—moisture coefficient—maximum light energy utilization capacity parameter, so as to form an environmental regulation parameter matrix that can be directly used for productivity calculation.

[0037] The photosynthetically active radiation absorption and the set of environmental regulation parameters are used as model optimization parameters and input into a preset genetic optimization algorithm to construct a CASA model.

[0038] In a further embodiment, the photosynthetically active radiation absorption (PARA) data matrix and the environmental regulation parameter matrix are used as optimization input parameters, and a genetic optimization algorithm is introduced to correct key parameters in the CASA model. The genetic optimization algorithm consists of computational structures such as initial population generation, fitness assessment, selection, crossover, and mutation. The initial population size is set to 50 parameter combinations, the crossover probability is set to 0.8, and the mutation probability is set to 0.05. The algorithm uses minimizing the error between historical sample productivity data and model calculation results as the fitness evaluation criterion to iteratively optimize light energy utilization efficiency and environmental regulation parameters, thereby generating a CASA model parameter configuration that matches the ecological conditions of the target area and forming the final productivity calculation model structure.

[0039] Optionally, incorporating the maximum light energy utilization capability parameter into the environmental adjustment factor includes: Based on the proportion of vegetation types corresponding to each dynamic observation unit, the maximum light energy utilization capacity parameter corresponding to the vegetation type with the largest proportion is selected as the representative light energy utilization capacity parameter of the dynamic observation unit. In this embodiment, a classification table of pixel number and vegetation type is formed using the results of supervised classification. The vegetation types are then statistically analyzed according to the pixel affiliation of the dynamic observation unit to calculate the area percentage of each vegetation type within the corresponding observation unit. When the area percentage of a certain vegetation type exceeds 50%, it is identified as the dominant vegetation type of that dynamic observation unit. Subsequently, the maximum light energy utilization parameter corresponding to that vegetation type is extracted from a pre-established vegetation growth rule data table. For example, for typical herbaceous vegetation... shrub vegetation The extracted parameters are used as representative light energy utilization parameters for the dynamic observation unit.

[0040] The representative light energy utilization capability parameters of each dynamic observation unit are mapped to the environmental adjustment factors of the corresponding dynamic observation unit in order to perform parameter merging.

[0041] In a further embodiment, based on the dynamic observation unit's number and center coordinates, parameters representing light energy utilization capability are mapped to the corresponding environmental regulation factor records. A maximum light energy utilization capability parameter field is added to the existing data structure (observation unit number—time node—temperature coefficient—moisture coefficient), thus forming a new environmental regulation parameter structure table. To ensure parameter consistency, time nodes are uniformly verified during the merging process, and all parameters are stored in a matrix format (observation unit number—time node—parameter item). Optionally, the method for identifying the vegetation type in the target area includes: Spectral feature analysis was performed based on the differences in reflectance characteristics of pixels in the near-infrared and red bands of remote sensing images, and the normalized vegetation index was used to screen the pixel regions of vegetation growth. In this embodiment, reflectance data of each pixel in the remote sensing image in the near-infrared and red bands are read, and a pixel spectral feature table is constructed. Each record includes pixel number, near-infrared reflectance, red reflectance, and spatial coordinates. Then, the Normalized Difference Vegetation Index (NDVI) is calculated based on the reflectance difference between the two bands, and pixels with an NDVI value greater than 0.2 are identified as candidate pixels with vegetation growth characteristics. Simultaneously, a secondary screening is performed using the spectral discrimination criteria of near-infrared reflectance higher than 0.25 and red reflectance lower than 0.15 to exclude bare soil or built-up surfaces. After screening, a set of vegetation growth pixels is formed and recorded in the structure of pixel number—NDVI—spectral feature value.

[0042] Based on the differences in spectral reflectance characteristics and seasonal variation characteristics of pixels within the vegetation growth pixel area, the growth patterns of each vegetation type in the vegetation growth rules are compared to classify and identify the pixels, thereby determining the vegetation type of the target area.

[0043] In a further embodiment, the NDVI change sequence of each pixel at consecutive time points is statistically analyzed. Then, growth cycle characteristic parameters corresponding to each vegetation type are extracted from the vegetation growth rule data table. For example, the NDVI peak of herbaceous vegetation typically occurs from June to August with a peak range of 0.6-0.8, while the peak range of shrub vegetation is 0.5-0.7 with a longer duration. By comparing the similarity between the pixel NDVI time series and the growth cycle characteristics of each vegetation type, when the similarity is greater than 0.7, the pixel is determined to belong to the corresponding vegetation type, and a classification result table is generated, thereby determining the spatial distribution of vegetation types in the target area.

[0044] Optionally, the vegetation net productivity determined in the productivity assessment module during the evaluation period includes: Based on the output of the CASA model, the net primary productivity of vegetation in each dynamic observation unit at each time node is determined, and data points that exceed the vegetation growth rules in the net primary productivity of vegetation are removed, so as to obtain the effective net primary productivity of vegetation in each dynamic observation unit at each time node. In this embodiment, the CASA model output is read. The output is stored in a three-column data structure: dynamic observation unit number—time node—net primary productivity value of vegetation. The productivity is expressed in units. Then, pre-established vegetation growth rules are invoked. This data table contains the reasonable productivity ranges for different vegetation types at different growth stages; for example, the productivity range for herbaceous vegetation during its vigorous growth period is 2.0–6.0. The CASA model output values ​​are compared one by one with the reasonable range for the corresponding vegetation type. When the productivity value at a certain time point is lower than 0 or higher than the upper limit of the corresponding range (e.g., exceeding 6.0), the comparison is made accordingly. When the data point is an anomaly, it is marked as an anomaly and removed. A new productivity record table is formed from the removed data, thereby obtaining the effective net primary productivity data sequence of vegetation for each dynamic observation unit at each time node.

[0045] Using the evaluation period as the statistical time scale, the effective net primary productivity of vegetation at each time node of each dynamic observation unit is calculated cumulatively over the period to obtain the net vegetation productivity of each dynamic observation unit within the evaluation period.

[0046] In a further embodiment, the effective net primary productivity of vegetation is cumulatively calculated using the evaluation period as the statistical time scale. The evaluation period can be set to one year, and a productivity time series can be established with 10-day intervals. For the same dynamic observation unit, the effective net primary productivity of vegetation at each time point within the evaluation period is accumulated sequentially, and the cumulative results are recorded. To ensure statistical stability, missing nodes are compensated for using the mean of adjacent nodes during the accumulation process. The final result is a data structure of dynamic observation unit number—evaluation period—cumulative productivity, where the cumulative productivity is calculated using... This indicates that the net vegetation productivity of each dynamic observation unit during the evaluation period can be obtained.

[0047] Optionally, data points in net primary productivity that exceed the vegetation growth rules are excluded, including: Based on vegetation growth rules, determine the range of net primary productivity values ​​for each dynamic observation unit corresponding to the vegetation type at each growth stage. In this embodiment, the growth stage field in the vegetation growth rules is read, where the growth stages include the germination period, rapid growth period, and maturity / stabilization period. For example, for typical herbaceous vegetation in the target area, the net primary productivity range of the germination period is set to... The rapid growth period is The mature and stable period is Subsequently, based on the vegetation type corresponding to the dynamic observation unit, the range of net primary productivity values ​​for each growth stage was extracted from the data table.

[0048] The vegetation growth stage of each dynamic observation unit at each time point is determined, and the net primary productivity of vegetation at each time point of each dynamic observation unit is compared with the range of net primary productivity values ​​under each growth stage to eliminate data points that exceed the range of net primary productivity values.

[0049] In a further embodiment, the rate of change of NDVI is used as the criterion. When the NDVI value rises continuously and the change is greater than 0.05, it is determined to be in the rapid growth stage. When the NDVI is in the range of 0.6 to 0.8 and the change is less than 0.02, it is determined to be in the mature and stable stage. The remaining stages are determined to be the germination stage. Subsequently, the net primary productivity data calculated by the CASA model is read, and the productivity value is compared with the established stage productivity range according to the growth stage at the corresponding time node. When the productivity value is lower than the lower limit or higher than the upper limit, the data point is marked as an anomaly and removed, thereby retaining the valid net primary productivity data within a reasonable range.

[0050] Of particular importance is determining the vegetation growth stage at each time point for each dynamic observation unit, including: Based on the normalized vegetation index corresponding to each dynamic observation unit, calculate the magnitude of vegetation index change between adjacent time nodes. In this embodiment, differential calculation is performed on adjacent time nodes of the same dynamic observation unit, that is, the NDVI value of the later time node is subtracted from the NDVI value of the earlier time node to obtain the magnitude of vegetation index change. For example, when a certain observation unit is in and When the NDVI values ​​corresponding to the time points are 0.42 and 0.50 respectively, the change range is 0.08.

[0051] Based on the vegetation growth rules, the growth pattern of the vegetation type corresponding to each dynamic observation unit pixel is determined, and the growth pattern is numerically converted to obtain the vegetation index change threshold corresponding to the growth stage of each vegetation type. In a further embodiment, growth pattern characteristics of different vegetation types are extracted from vegetation growth rules and converted into numerical threshold parameters that can be used for calculation. For example, for typical herbaceous vegetation, based on long-term ecological monitoring samples, the NDVI variation range during the germination period is 0.01–0.04, during the rapid growth period it is 0.05–0.12, and during the maturity and stabilization period it is -0.02–0.02. These ranges are then converted into a vegetation type—growth stage— Threshold lower limit — A threshold matrix with upper threshold is generated and matched with the vegetation type corresponding to the dynamic observation unit to form a set of numerical thresholds that can be used for stage discrimination.

[0052] By comparing the vegetation index change threshold with the vegetation index change magnitude, the vegetation growth stage of each vegetation type in each dynamic observation unit can be determined.

[0053] In a further embodiment, the vegetation index change threshold and the vegetation index change magnitude are compared item by item. When a dynamic observation unit is within a certain time interval... When the value falls within the threshold range of the rapid growth period for the corresponding vegetation type (e.g., 0.05–0.12), the time point is determined to be in the rapid growth period; when… When the value is in the range of 0.01 to 0.04, it is considered to be in the germination period; when... lie in The range of 0.02 to 0.02 indicates a mature and stable period.

[0054] Optionally, the benefit assessment module may determine the productivity changes of each dynamic observation unit over different time periods, including: The time period for implementing the restoration project is determined based on the grayscale changes of the same pixel in the remote sensing image within the evaluation period; In this embodiment, the grayscale value sequence of the same pixel in the near-infrared band is extracted from the remote sensing image, with a time interval set to 10 days. Then, the grayscale change rate between adjacent time nodes is calculated. When the grayscale value undergoes a sudden change within three consecutive time nodes and the change amplitude exceeds 0.08, this time interval is marked as a period in which the surface vegetation structure undergoes significant changes. Combined with the engineering construction records of the target area, the time periods in which grayscale abrupt changes occur are cross-validated to determine the implementation time period of the ecological restoration project.

[0055] By comparing the net vegetation productivity of the same dynamic observation unit during the restoration project implementation period and before and after the implementation period, the productivity difference before implementation and the productivity difference after implementation can be obtained. In a further embodiment, the vegetation net productivity data table calculated by the CASA model is read. Then, based on the determined restoration project implementation period, the vegetation net productivity of the same dynamic observation unit is averaged in the pre-implementation time window (e.g., 6 months before implementation) and the post-implementation time window (6 months after implementation). Using the implementation period as the baseline node, the difference between the average productivity before implementation and the productivity of the baseline node, as well as the difference between the average productivity after implementation and the productivity of the baseline node, are calculated to obtain the productivity difference before implementation and the productivity difference after implementation.

[0056] If the difference in productivity after implementation of any dynamic observation unit is greater than the difference in productivity before implementation, the corresponding dynamic observation unit is determined to be an effective area for ecological restoration; if the difference in productivity after implementation of any dynamic observation unit is less than or equal to the difference in productivity before implementation, the corresponding dynamic observation unit is determined to be an area with good ecological restoration. In a further embodiment, the productivity changes of each dynamic observation unit are determined. When the productivity difference after implementation of an observation unit is greater than the productivity difference before implementation, it is determined that the unit has shown a significant increase in productivity after the project is implemented, and it is marked as an effective area for ecological restoration; when the productivity difference after implementation is less than or equal to the productivity difference before implementation, it indicates that the productivity change tends to be stable or the increase is limited, and it is marked as a good area for ecological restoration.

[0057] The spatial proportions of effective and well-restored ecological restoration areas within the target area are determined, and these spatial proportions are used as weights. The difference in productivity after implementation is used as the contribution value for productivity improvement to calculate the benefits of grassland ecological restoration projects.

[0058] In a further embodiment, the total area of ​​the dynamic observation units corresponding to the effective ecological restoration area and the well-restored ecological restoration area is statistically summarized and divided by the total area of ​​the target area to obtain the spatial proportion of the two types of areas, for example, the effective area accounts for 0.35 and the well-restored area accounts for 0.65. Then, the spatial proportion is used as a weighting parameter, and the difference in productivity after implementation is used as the contribution value for productivity improvement. A weighted cumulative calculation is then performed on all dynamic observation units to obtain the comprehensive benefit index of the grassland ecological restoration project within the evaluation period.

[0059] Of particular importance is determining the timeframe for the restoration work, including: The grayscale value of the same pixel in the remote sensing image is compared and detected at different time nodes in the hourly comparison. The grayscale change between adjacent time nodes is calculated, and the time node with grayscale change greater than the preset change threshold is identified. In this embodiment, a remote sensing image sequence within the evaluation period is read, and grayscale value data of the same pixel at consecutive time nodes are extracted. The grayscale values ​​can be derived from the near-infrared band, and the time node interval is set to 10 days. Subsequently, a difference calculation between adjacent time nodes is performed on the same pixel, that is, the grayscale value of the previous time node is subtracted from the grayscale value of the later time node to obtain the grayscale change. ;when When the change exceeds a preset threshold (e.g., 0.07), the corresponding time node is marked as a potential change node.

[0060] When more than 9 pixels have grayscale changes greater than the change threshold in a continuous time period, the starting time node of the continuous time period is determined as the start time of the restoration project for the corresponding pixel area. In a further embodiment, a 3×3 neighborhood pixel set is constructed according to the spatial location of pixels in the remote sensing image. At each time point, the number of pixels in the neighborhood whose grayscale change exceeds a threshold is counted. When more than 9 pixels in a neighborhood set simultaneously meet the threshold within a consecutive time period, the set is considered complete. In such cases, it is assumed that concentrated surface disturbance occurred in the area during that time period. The first time point of this continuous time period is then recorded as the start time of the restoration project for that pixel area.

[0061] The grayscale changes of pixels are monitored at subsequent time points after the start time of the restoration project. When the grayscale changes of pixels tend to stabilize at several consecutive monitoring time points, the start time of the several consecutive monitoring time points is taken as the end time of the restoration project for the corresponding pixel area.

[0062] In a further embodiment, after determining the start time of the project implementation, the grayscale changes in the area at subsequent time points are continuously monitored. The amount of grayscale change at subsequent time points is then calculated. and for continuous time nodes The value is used to perform sliding window statistics, for example, setting the monitoring window length to 4 time nodes. When the values ​​of each node within the window... When all values ​​are less than the stability threshold (e.g., 0.02) and the trend remains stable, the surface disturbance process is considered to have basically ended. The first time point among the consecutive time points that meet the stability conditions is recorded as the end time of the remediation project.

[0063] Therefore, the embodiments should be considered as exemplary and non-limiting in all respects, and the scope of the invention is defined by the appended claims rather than the foregoing description. Thus, all variations falling within the meaning and scope of the equivalents of the application are intended to be included within the invention.

[0064] The above description is merely a specific embodiment of the present invention, enabling those skilled in the art to understand or implement the invention. Various modifications to these embodiments will be readily apparent to those skilled in the art, and the general principles defined herein may be implemented in other embodiments without departing from the spirit or scope of the invention. Therefore, the present invention is not to be limited to the embodiments shown herein, but is to be accorded the widest scope consistent with the principles and novel features of the invention herein.

Claims

1. A grassland ecological restoration engineering benefit evaluation system based on the CASA model, characterized in that, Includes the following modules: The data acquisition module is used to acquire remote sensing images of the target area during the evaluation period through the remote sensing data platform, and to collect meteorological observation data and surface radiation data for the corresponding time period. The vegetation index calculation module is used to calculate the normalized vegetation index of each preset dynamic observation unit based on the near-infrared and red light band information in the remote sensing image. The regulation factor calculation module is used to generate effective environmental regulation factors for the target area based on meteorological observation data and surface radiation data, and to generate a set of vegetation growth parameters by combining the normalized vegetation index. The productivity assessment module is used to input the set of vegetation growth parameters into a preset CASA model to determine the net productivity of vegetation within the evaluation period. The benefit assessment module is used to compare the net vegetation productivity before and after the implementation of the grassland ecological restoration project within the evaluation period, determine the changes in productivity of each dynamic observation unit in different time periods, and evaluate the benefits of the grassland ecological restoration project.

2. The grassland ecological restoration engineering benefit evaluation system based on the CASA model according to claim 1, characterized in that, The methods for dividing dynamic observation units in the vegetation index calculation module include: The basic grid scale is determined based on the spatial resolution of the remote sensing image, and the target area is divided into several grid units in combination with the geographic coordinate system of the remote sensing image. The assigned pixels of each grid unit are determined based on the spatial coordinates of the center point of the remote sensing image pixels, and the assigned pixels of each grid unit are then collected. Based on the vegetation index variance of each grid cell in the aggregation processing results, grid cell splitting and merging are performed to obtain dynamic observation cells.

3. The grassland ecological restoration engineering benefit evaluation system based on the CASA model according to claim 2, characterized in that, Performing grid cell splitting and merging includes: The vegetation index within each grid cell is calculated based on the near-infrared and red light band information of each grid cell in the collected processing results. Calculate the vegetation index variance within each grid cell. If the vegetation index variance is greater than a preset variance threshold, then split the corresponding grid cell into several grid cells to be merged. If the variance of the vegetation index is less than or equal to the variance threshold, the corresponding grid cell will be treated as a dynamic observation cell. Calculate the spatial gradient of vegetation index of the grid cell to be merged within a preset neighborhood range, and merge adjacent grid cells to be merged with spatial gradients less than or equal to a preset spatial gradient threshold as a dynamic observation unit.

4. The grassland ecological restoration engineering benefit evaluation system based on the CASA model according to claim 1, characterized in that, The effective environmental regulation factors generated in the regulation factor calculation module for the target area include: Extract temperature and precipitation data from meteorological observation data corresponding to the time nodes of remote sensing images; calculate the photosynthetically active radiation ratio of the target area at each time node based on surface radiation data; The temperature data at each time point are compared with the suitable temperature range for the vegetation type in the target area determined according to the preset vegetation growth rules to obtain the temperature regulation coefficient corresponding to each time point. The water supply capacity is calculated based on the precipitation data at each time point to determine the water regulation coefficient corresponding to each time point. Temperature regulation coefficient, moisture regulation coefficient, and photosynthetically active radiation ratio are integrated into environmental regulation factors, and these environmental regulation factors are spatially mapped to dynamic observation units.

5. The grassland ecological restoration engineering benefit evaluation system based on the CASA model according to claim 1, characterized in that, The methods for constructing the CASA model in the productivity assessment module include: The normalized vegetation index was used to construct the vegetation absorption of photosynthetically active radiation ratio parameter, and the photosynthetically active radiation absorption of each dynamic observation unit at different time points was determined in combination with the photosynthetically active radiation ratio. Extract the maximum light energy utilization parameters corresponding to the growth characteristics of different vegetation types in the vegetation growth rules, and merge the maximum light energy utilization parameters into the environmental regulation factors to generate a set of environmental regulation parameters. The photosynthetically active radiation absorption and the set of environmental regulation parameters are used as model optimization parameters and input into a preset genetic optimization algorithm to construct a CASA model.

6. The grassland ecological restoration engineering benefit evaluation system based on the CASA model according to claim 5, characterized in that, The maximum light energy utilization capability parameter is incorporated into the environmental adjustment factor, including: Based on the proportion of vegetation types corresponding to each dynamic observation unit, the maximum light energy utilization capacity parameter corresponding to the vegetation type with the largest proportion is selected as the representative light energy utilization capacity parameter of the dynamic observation unit. The representative light energy utilization capability parameters of each dynamic observation unit are mapped to the environmental adjustment factors of the corresponding dynamic observation unit in order to perform parameter merging.

7. The grassland ecological restoration engineering benefit evaluation system based on the CASA model according to claim 4, characterized in that, Methods for identifying vegetation types in target areas include: Spectral feature analysis was performed based on the differences in reflectance characteristics of pixels in the near-infrared and red bands of remote sensing images, and the normalized vegetation index was used to screen the pixel regions of vegetation growth. Based on the differences in spectral reflectance characteristics and seasonal variation characteristics of pixels within the vegetation growth pixel area, the growth patterns of each vegetation type in the vegetation growth rules are compared to classify and identify the pixels, thereby determining the vegetation type of the target area.

8. The grassland ecological restoration engineering benefit evaluation system based on the CASA model according to claim 1, characterized in that, The productivity assessment module determines the net vegetation productivity during the evaluation period, including: Based on the output of the CASA model, the net primary productivity of vegetation in each dynamic observation unit at each time node is determined, and data points that exceed the vegetation growth rules in the net primary productivity of vegetation are removed, so as to obtain the effective net primary productivity of vegetation in each dynamic observation unit at each time node. Using the evaluation period as the statistical time scale, the effective net primary productivity of vegetation at each time node of each dynamic observation unit is calculated cumulatively over the period to obtain the net vegetation productivity of each dynamic observation unit within the evaluation period.

9. The grassland ecological restoration engineering benefit evaluation system based on the CASA model according to claim 1, characterized in that, Data points in net primary productivity that exceed the vegetation growth rules were removed, including: Based on vegetation growth rules, determine the range of net primary productivity values ​​for each dynamic observation unit corresponding to the vegetation type at each growth stage. The vegetation growth stage of each dynamic observation unit at each time point is determined, and the net primary productivity of vegetation at each time point of each dynamic observation unit is compared with the range of net primary productivity values ​​under each growth stage to eliminate data points that exceed the range of net primary productivity values.

10. The grassland ecological restoration engineering benefit evaluation system based on the CASA model according to claim 1, characterized in that, The benefit assessment module determines the productivity changes of each dynamic observation unit over different time periods, including: The time period for implementing the restoration project is determined based on the grayscale changes of the same pixel in the remote sensing image within the evaluation period; By comparing the net vegetation productivity of the same dynamic observation unit during the restoration project implementation period and before and after the implementation period, the productivity difference before implementation and the productivity difference after implementation can be obtained. If the difference in productivity after implementation of any dynamic observation unit is greater than the difference in productivity before implementation, the corresponding dynamic observation unit is determined to be an effective area for ecological restoration; if the difference in productivity after implementation of any dynamic observation unit is less than or equal to the difference in productivity before implementation, the corresponding dynamic observation unit is determined to be an area with good ecological restoration. The spatial proportions of effective and well-restored ecological restoration areas within the target area are determined, and these spatial proportions are used as weights. The difference in productivity after implementation is used as the contribution value for productivity improvement to calculate the benefits of grassland ecological restoration projects.

Images