Evaluation method for grassland degradation recovery condition by integrating ecological structure and service indexes

Abstract

The invention provides a grassland degradation and recovery condition evaluation method integrating an ecological structure and a service index, and the method comprises the steps: constructing a grassland ecological index data set based on multi-source remote sensing data, soil data and meteorological data, and carrying out the inversion of a single-factor monitoring index comprising an ecological system structure index and an ecological system service index; identifying an ecosystem service beam, and performing ecological partitioning on the research area by using the ecosystem service beam; determining the weight of the single-factor monitoring index of each ecological partition, and calculating a grassland health and service comprehensive index according to the weight of the single-factor monitoring index; and evaluating grassland degradation and recovery grades by using the grassland health and service comprehensive indexes. According to the method, the ecological system structure and the service indexes are comprehensively considered, and the comprehensiveness of the indexes is ensured; meanwhile, the difference of grassland ecosystem functions of different regions is fully considered, the index weight is adjusted according to the region characteristics, and the scientificity and reliability of the evaluation result are remarkably improved.

Description

Technical Field

[0001] The present invention relates to the technical field of ecological remote sensing and environmental monitoring, and in particular to an evaluation method for grassland degradation recovery status based on comprehensive ecological structure and service indicators. Background Art

[0002] Remote sensing technology, as a core tool for monitoring grassland degradation, can provide extensive spatial coverage and regular monitoring data, offering solid technical support for assessing grassland ecological status. Common methods for monitoring grassland degradation include visual interpretation and machine learning classification. Visual interpretation relies on the interpreter's experience, is susceptible to human interference, and is inefficient in large-scale data processing. While machine learning classification offers the advantage of automation, its effectiveness is limited by classifier selection and the quality of training samples. Therefore, the accuracy and applicability of remote sensing data processing and analysis methods still need to be further optimized.

[0003] In recent years, researchers have gradually recognized that constructing a multi-indicator comprehensive index to monitor grassland degradation is an effective solution, particularly in the indicator screening and scoring process. Traditional single ecological structural indicators, such as vegetation cover, net primary productivity, and aboveground biomass, while able to reflect grassland growth conditions, often fail to fully capture the entire process of grassland degradation. These structural indicators may have limitations. For example, an increase in vegetation cover may simply represent an expansion of degraded plant species rather than an improvement in ecosystem health. Therefore, assessment methods that rely solely on these indicators are prone to bias. As understanding of grassland ecosystem degradation deepens, researchers have begun to focus on the degradation of ecosystem services. Ecological services such as water conservation, soil conservation, and windbreak and sand fixation directly impact the functional integrity and stability of grassland ecosystems. Therefore, the current research trend is to incorporate ecological service indicators into grassland degradation assessment frameworks. This integration will help provide more comprehensive and accurate assessments of grassland degradation and provide a scientific basis for ecological restoration.

[0004] In the process of constructing comprehensive indicators, scientifically determining the weights of various indicators becomes a key issue. Traditional weight calculation methods, such as the Analytic Hierarchy Process (AHP), entropy weighting, and principal component analysis (PCA), have been widely used in grassland degradation assessments. However, these methods fail to fully account for the spatial heterogeneity and interrelationships between ecosystem services, resulting in a lack of regional diversity in assessment results. To address this issue, the Ecosystem Services Bundle (ESB) theory has emerged. It can identify the aggregated characteristics of ecosystem services in different regions and assign differentiated weights to each region based on these characteristics, thereby improving the scientific nature and representativeness of assessment results.

[0005] Existing technologies for grassland degradation assessment suffer from two major shortcomings. First, indicator selection relies too heavily on a single or limited number of ecological structural indicators, failing to fully encompass ecosystem services and resulting in incomplete assessment results. Second, traditional weighting methods fail to adequately account for the spatial heterogeneity and interrelationships of ecosystem services, impacting the accuracy of assessment results and the reflection of regional differences. Therefore, optimizing indicator selection and weighting methods, and exploring a more scientific and comprehensive assessment framework, will help improve the accuracy and practicality of grassland degradation monitoring. Summary of the Invention

[0006] In response to the technical problem that the existing grassland degradation assessment methods have unreasonable indicator selection, resulting in incomplete assessment results, the present invention proposes an assessment method for grassland degradation recovery status that integrates ecological structure and service indicators. The grassland degradation recovery status is assessed based on multi-source remote sensing and meteorological data, and the integration of ecological structure and ecological service indicators. The assessment results are more comprehensive and accurate, and are suitable for dynamic monitoring of the health status of grassland ecosystems at the regional scale.

[0007] In order to achieve the above-mentioned purpose, the technical solution of the present invention is implemented as follows: the main technical ideas of the present invention are: based on remote sensing data and meteorological data, invert single-factor monitoring indicators; use ecosystem service bundles to ecologically zone the study area; within each zone, determine the weight of each single-factor monitoring indicator, and then construct a comprehensive evaluation index; finally, evaluate the grassland degradation and restoration status through the comprehensive evaluation index.

[0008] A method for assessing grassland degradation and restoration status based on integrated ecological structure and service indicators is developed, with the following steps:

[0009] Step 1: Construct a grassland ecological indicator dataset based on multi-source remote sensing data, soil data, and meteorological data. Based on the data in the grassland ecological indicator dataset, invert single-factor monitoring indicators including ecosystem structure indicators and ecosystem service indicators;

[0010] Step 2: Identify ecosystem service bundles and use them to ecologically zonate the study area;

[0011] Step 3: Determine the weight of the single-factor monitoring indicator for each ecological zone, and calculate the grassland health and service comprehensive index based on the weight of the single-factor monitoring indicator;

[0012] Step 4: Use the grassland health and service comprehensive index to evaluate grassland degradation and restoration levels.

[0013] Preferably, the ecosystem structure indicators include vegetation coverage, net primary productivity and sunlight-induced chlorophyll fluorescence, and the ecosystem service indicators include water production, soil conservation and windbreak and sand fixation.

[0014] Preferably, the pixel dichotomy method is used to calculate the vegetation cover of the study area;

[0015] The CASA model was used to calculate net primary productivity;

[0016] The sunlight-induced chlorophyll fluorescence is used to evaluate the chlorophyll content and nitrogen level of the grassland, reflecting the health status and nutritional needs of the grassland. The maximum value of the monthly data product of the existing solar-induced chlorophyll fluorescence data is used to synthesize the annual solar-induced chlorophyll fluorescence value.

[0017] Water conservation services reflect the ability of grassland ecosystems to maintain and regulate water resources and are an important indicator of grassland health and function. Water conservation services are quantified and calculated using the InVEST water production model.

[0018] Soil conservation services reflect the protective effect of grassland vegetation on soil erosion and are used to monitor grassland degradation; the universal soil loss equation calculates soil erosion.

[0019] Windbreak and sand fixation services are an important manifestation of grassland ecological functions, which directly reflect the effectiveness of grassland in protecting soil and resisting wind erosion. The modified wind erosion equation is used to estimate windbreak and sand fixation services.

[0020] Preferably, the vegetation coverage is:

[0021]

[0022] Where, NDVI soil is the NDVI value of the bare soil pixel, NDVI veg is the NDVI value of the pure vegetation cover pixel, NDVI is the NDVI value of the study area pixel; the 1% and 99% probability of the annual NDVI value range distribution are used as NDVI soil and NDVI veg The value of

[0023] The method for calculating net primary productivity using the CASA model is:

[0024] NPP=APAR×LUE

[0025] APAR=0.5×SOL×FPAR

[0026] LUE=T ε1 ×T ε2 ×W ε ×LUE max

[0027] Where APAR represents the photosynthetically active radiation absorbed by vegetation; LUE represents the actual light energy utilization efficiency. FPAR represents the ratio of incident photosynthetically active radiation absorbed by the vegetation layer; the constant 0.5 represents the ratio of solar active radiation that can be used by vegetation, that is, the ratio of solar active radiation to the total solar radiation SOL; T ε1 and T ε2 represents temperature stress factor; W ε Indicates water stress factor; LUE max Indicates the maximum light energy utilization under ideal conditions;

[0028] The calculation method of the InVEST water production model is:

[0029]

[0030] PET=Kc×ET0

[0031]

[0032] Where Y is water yield; AET is actual evapotranspiration; P is annual precipitation; PET is annual potential evapotranspiration; Kc is crop evapotranspiration coefficient; ET0 is reference crop evapotranspiration; W is the ratio of annual vegetation available water to precipitation; Z represents Zhang coefficient; AWC is soil available water content;

[0033] The soil conservation service is the difference between the potential soil erosion and the actual soil erosion. The soil conservation service is:

[0034] SC=R×K×L×S×(1-C×E)

[0035] Where R represents the precipitation erosivity factor, which is calculated using the monthly formula proposed by Wischmeier and Smith; K represents the soil erodibility factor, which is calculated using soil organic carbon and particle composition data proposed by Williams and Arnold; L and S represent the slope length factor and slope factor, respectively; C represents the vegetation cover management factor, which is calculated using the relationship equation between the C factor and vegetation cover fraction (FVC); and E represents the engineering measure factor.

[0036] The wind and sand fixation service is reflected by the difference between the potential wind erosion amount and the actual wind erosion amount.

[0037] SR=SL r -SL

[0038]

[0039] Q rmax =109.8×(WF×EF×SCF×K')

[0040] S r =150.71×(WF×EF×SCF×K') -0.3711

[0041]

[0042] Q max =109.8×(WF×EF×SCF×K'×COG)

[0043] S=150.71×(WF×EF×SCF×K'×COG) -0.3711

[0044] Where SL r is the potential wind erosion amount, SL is the actual wind erosion amount; Q rmax Indicates the maximum sediment transport capacity of potential wind force; S r represents the length of potential key plot; Q max represents the maximum sediment transport capacity of wind; S represents the length of the key plot; Z represents the distance where maximum wind erosion occurs; WF represents the climate factor; EF represents the percentage of soil erodibility; SCF represents the soil crust factor; K' represents the soil roughness factor; COG represents the vegetation cover factor.

[0045] Preferably, the method for identifying the ecosystem service bundle includes:

[0046] A. Using the unsupervised clustering algorithm of self-organizing maps, we clustered the spatial pixels of the study area based on multiple single-factor monitoring indicators of ecosystem service indicators and automatically divided the areas with similar ecosystem service functions;

[0047] B. Based on the zoning identification strategy of the degree of ecosystem service contribution, the study area is functionally zoned according to the spatial dominance of service functions to identify the dominant areas of ecosystem service functions.

[0048] Preferably, the implementation method of the zoning identification strategy based on the degree of ecosystem service contribution is as follows: (1) at the pixel scale, the annual average values ​​of the single-factor monitoring indicators of the three ecosystem service indicators are normalized to 0-1; (2) the total amount of the ecosystem service indicator of each pixel is calculated; (3) according to the size of the three ecosystem service functions, the total amount of ecosystem services is ranked from high to low, and the single-factor monitoring indicators of the key ecosystem service indicators that account for 75% of the total ecosystem service amount are identified as the ecosystem service bundle.

[0049] Preferably, the method for determining the weight of the single factor monitoring indicator of each ecological zone is:

[0050] Standardize single factor monitoring indicators;

[0051] Using principal component analysis or entropy weight method, principal component transformation is performed on the standardized single factor result layer of each ecosystem service bundle to obtain the variance contribution rate of each principal component and the weight coefficient of the single factor monitoring indicator; the weight coefficient is:

[0052]

[0053] Where W i is the weight of the i-th single factor monitoring indicator, m is the number of single factor monitoring indicators, n is the number of principal components with a cumulative contribution rate exceeding 90%, α ij is the weight of the i-th single factor in the j-th principal component, β j is the contribution rate of the jth principal component;

[0054] The single factor result layer includes three ecosystem structure indicators and the dominant ecosystem service indicator.

[0055] Preferably, the calculation formula of the grassland health and service comprehensive index GHSI is:

[0056]

[0057] Where W i is the weight of the i-th single factor monitoring indicator, N i is the i-th standardized single factor monitoring indicator.

[0058] Preferably, the annual average change rate of the grassland health and service comprehensive index GHSI over many years is calculated pixel by pixel, and the average value of the annual average change rate is used as the benchmark value to characterize the long-term change trend of the grassland ecological status;

[0059] The ordered sample optimal segmentation method in cluster analysis was used to classify the annual average change rate of the grassland health and service comprehensive index (GHSI), and the grassland was divided into five levels: severely degraded, moderately degraded, slightly degraded, non-degraded, and restored.

[0060] Preferably, the annual average change rate of the grassland health and service comprehensive index GHSI of all pixels is arranged in ascending order; the ecological status is divided according to the positive and negative values ​​of the change rate: when the change rate is greater than zero, it indicates that the grassland ecological status is stable or recovered, and is subdivided into two levels: "non-degraded" and "recovered"; when the change rate is less than zero, it indicates the degradation process, and is divided into four levels: "non-degraded", "mildly degraded", "moderately degraded" and "severely degraded".

[0061] Compared with the existing technology, the beneficial effects of the present invention are as follows: the present invention includes the steps of inverting single-factor monitoring indicators, identifying ecosystem service bundles, determining single-factor indicator weights, and evaluating grassland degradation and restoration levels. Through comprehensive analysis of remote sensing data and meteorological data, key ecological monitoring indicators are inverted, and the study area is divided using ecosystem service bundles to determine the indicator weights of each partition, thereby constructing a comprehensive evaluation index to more accurately evaluate grassland degradation and restoration status. The present invention not only comprehensively considers ecosystem structure and service indicators, but can also evaluate grassland health from multiple perspectives, ensuring the comprehensiveness of the indicators; at the same time, it fully considers the differences in grassland ecosystem functions in different regions, adjusts indicator weights according to regional characteristics, significantly improves the scientific nature and reliability of the evaluation results, and provides effective support for the dynamic monitoring of grassland ecosystem health. The present invention is not only suitable for grassland degradation and restoration assessment, but also provides effective support for the dynamic monitoring of other ecosystem health conditions, and can promote the scientific and refined management of ecological environmental protection and restoration work. BRIEF DESCRIPTION OF THE DRAWINGS

[0062] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the following briefly introduces the drawings required for use in the embodiments or the description of the prior art. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative work.

[0063] Figure 1 It is the overall technical framework diagram of the present invention;

[0064] Figure 2 This is an overview of the study area of ​​the present invention;

[0065] Figure 3 This is the spatial distribution map of the multi-year average value of the single-factor monitoring indicator of the present invention; among them, (a) is vegetation coverage, (b) is water conservation service, (c) is vegetation productivity, (d) is soil conservation service, (e) is sunlight-induced chlorophyll fluorescence, and (f) is windbreak and sand fixation service.

[0066] Figure 4 A distribution pattern diagram of the ecosystem service bundles of the present invention;

[0067] Figure 5 This is the spatial distribution map of the multi-year average value of the grassland health and service comprehensive index (GHSI) of Xinjiang grassland from 2000 to 2020;

[0068] Figure 6 This is the grassland degradation and restoration assessment results in Xinjiang from 2000 to 2020. DETAILED DESCRIPTION

[0069] The following will clearly and completely describe the technical solutions in the embodiments of the present invention in conjunction with the accompanying drawings. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. All other embodiments obtained by ordinary technicians in this field based on the embodiments of the present invention without creative work are within the scope of protection of the present invention.

[0070] Before introducing the specific embodiments, the sources of the satellite images in the present invention are explained as follows: (1) Normalized Difference Vegetation Index data are from the sixth-generation MOD13A3 NDVI dataset released by NASA, with a spatial resolution of 1 km. (2) Topographic data are from the China 1km resolution digital elevation model dataset released by the National Center for Glaciers, Frozen Soils and Desertification. (3) Meteorological data, including monthly precipitation, temperature and potential evapotranspiration data, are from the National Tibetan Plateau Data Center, with a spatial resolution of 1 km; daily precipitation and wind speed data are from the ERA5-Land reanalysis dataset of the Copernicus Climate Data Repository, with a spatial resolution of 11 km; solar radiation data are from the Climate Forecast System Version 2 6-hour product provided by NOAA NWS and the National Centers for Environmental Prediction, with a spatial resolution of 22 km. (4) Soil texture data are from the World Soil Coordinated Database, with a spatial resolution of 1 km. (5) Solar-induced chlorophyll fluorescence data are from the global 0.05°×0.05° solar-induced chlorophyll fluorescence product provided by the National Tibetan Plateau Data Center. (6) Snow depth data were obtained from the China Snow Depth Long-Term Series Dataset from the Spatiotemporal Environmental Big Data Platform, with a spatial resolution of 25 km. (7) Root-Restricted Layer Depth data were obtained from the 100-meter-resolution bedrock depth map of China inverted by Yan Fapeng et al. Before using all of the aforementioned spatial raster data, preprocessing operations such as spatial registration, cropping the study area, and unifying the spatial reference system were performed. All raster data were then resampled to a 1-km resolution.

[0071] This example takes Xinjiang as the research area. Figure 2 As shown, Xinjiang boasts complex topography and a wide range of elevation differences. Its overall landscape can be summarized as "three mountain ranges and two basins." From north to south, these are the Altai Mountains, the Junggar Basin, the Tianshan Mountains, the Tarim Basin, and the Kunlun Mountains. Using this method, we evaluated the recovery of grassland degradation in this region from 2000 to 2020, as briefly described below.

[0072] The main technical ideas of the present invention are as follows Figure 1As shown in the figure: First, based on multi-source remote sensing data and meteorological data, a grassland ecological indicator dataset from 2000 to 2020 was constructed, including three ecosystem structure indicators: vegetation cover, net primary productivity, and sunlight-induced chlorophyll fluorescence, and three ecosystem service indicators: water production, soil conservation, and windbreak and sand fixation. Second, based on the ecosystem service indicators, ecosystem service bundles were identified. Third, based on the ecological zoning, the principal component analysis (PCA) method was used to determine the weights of single-factor indicators in each zone, thereby constructing the grassland health and service comprehensive index GHSI. Finally, based on the construction of the GHSI model, scientific and objective benchmark values ​​and classification thresholds were determined, and the grassland status was divided into five levels: recovery, non-degraded, slightly degraded, moderately degraded, and severely degraded, to obtain the grassland degradation level status in Xinjiang from 2000 to 2020.

[0073] The present invention is specifically described as follows: A method for evaluating grassland degradation and restoration status by integrating ecological structure and service indicators comprises the following steps:

[0074] Step 1: Construct a grassland ecological indicator dataset based on multi-source remote sensing data, soil data, and meteorological data. Based on the data in the grassland ecological indicator dataset, invert single-factor monitoring indicators including ecosystem structure indicators and ecosystem service indicators.

[0075] A. Calculate vegetation cover using pixel dichotomy: Fractional vegetation cover (FVC) is the percentage of the vertically projected area of ​​vegetation to the area of ​​the statistical region, reflecting the greenness and growth of the grass. Pixel dichotomy is the most mature method in FVC research. Therefore, the pixel dichotomy method was used to calculate the FVC of the study area. The formula is as follows:

[0076]

[0077] Where, NDVI soil is the NDVI value of the bare soil pixel, NDVI veg is the NDVI value of the pure vegetation cover pixel, NDVI is the NDVI value of the study area pixel, and the NDVI value is from the MODIS NDVI data. In order to exclude the influence of abnormal pixel values, the 1% and 99% probability of the annual NDVI value range distribution are used as NDVI. soil Value and NDVI veg value.

[0078] Figure 3 (a) shows the average status of grassland vegetation coverage in Xinjiang from 2000 to 2020. It can be seen that the spatial distribution of grassland vegetation coverage in Xinjiang is significantly different. The high-value areas are mainly concentrated in the Ili River Valley and the northern Altai Mountains; while the low-value areas are concentrated in the Kunlun Mountains and the southern Tianshan Mountains.

[0079] B. Calculating Net Primary Productivity Using the CASA Model: Grassland biomass is an important indicator of the ecological quality and production function of grassland ecosystems. Directly monitoring grassland biomass using remote sensing data is difficult. However, Net Primary Productivity (NPP) has been demonstrated by numerous researchers to be a useful indicator for measuring grassland biomass and monitoring grassland degradation and improvement. The CASA (Carnegie-Ames-Stanford Approach) model, which comprehensively considers environmental conditions and vegetation characteristics for light energy utilization, is widely used to simulate NPP. Its formula is as follows:

[0080] NPP=APAR×LUE

[0081] APAR=0.5×SOL×FPAR

[0082] LUE=T ε1 ×T ε2 ×W ε ×LUE max

[0083] Where APAR is the photosynthetically active radiation absorbed by vegetation (unit: MJ / m 2 ); LUE represents the actual light energy utilization efficiency (unit: gC / MJ). FPAR represents the ratio of incident photosynthetically active radiation absorbed by the vegetation layer; the constant 0.5 represents the ratio of solar effective radiation that can be utilized by vegetation, that is, the ratio of the total solar radiation SOL. ε1 and T ε2 represents temperature stress factor; W ε Indicates water stress factor; LUE max It represents the maximum light energy utilization rate under ideal conditions. The maximum light energy utilization rate in this study adopted the maximum light energy utilization rate simulation result of Zhu Wenquan et al., that is, the maximum light energy utilization rate of grassland is 0.542gC / MJ.

[0084] The calculation formula for FPAR is as follows:

[0085]

[0086] Where, NDVI min The minimum value of NDVI is taken as a constant of 0.023; NDVI max is the maximum value of NDVI, which is a constant of 0.634; SR is the vegetation index ratio; SR min is the minimum value of SR, taking the constant as 1.05; SR max is the maximum value of SR, taking a constant of 4.46; FPAR NDVI It represents the proportion of photosynthetically active radiation absorbed by vegetation estimated by NDVI; FPAR SRFPAR is the proportion of photosynthetically active radiation absorbed by vegetation estimated by SR; min is the minimum value of the absorption ratio FPAR, which is taken as a constant of 0.001; FPAR max is the maximum value of the absorption ratio FPAR, which is taken as a constant of 0.95.

[0087] The calculation formula of temperature stress factor is as follows:

[0088]

[0089] Where, T represents the monthly mean temperature; T opt It is the suitable temperature for plant growth. <T—T opt <13℃, the temperature stress factor T for that month ε2 Equal to half of the value when the monthly mean temperature is the optimum temperature.

[0090] The calculation formula of water stress factor is as follows:

[0091]

[0092]

[0093] R n =(E po +P) 0.5 ×[0.369+0.598×(E po / P) 0.5 ]

[0094]

[0095] Where PET represents potential evaporation (mm); EET represents potential evaporation (mm), which is calculated by the regional actual evapotranspiration model established by Zhou Guangsheng and Zhang Xinshi in 1996; P represents annual precipitation (mm); R n is the total net radiation of the surface; E po is the local potential evapotranspiration (mm), calculated based on the Thornthwaite vegetation-climate relationship model.

[0096] The calculations used NDVI data from MDOIS, monthly precipitation / average temperature data from the National Tibetan Plateau Data Center, and solar radiation data from NOAA NWS and the National Centers for Environmental Prediction.

[0097] Figure 3 (c) shows the average status of net primary productivity of grassland in Xinjiang from 2000 to 2020. It can be seen that high-value areas are distributed in areas with good water and heat conditions such as the Ili River Valley, while low-value areas are concentrated in arid mountainous areas and basins.

[0098] C. Solar-induced chlorophyll fluorescence (SIF) can assess the chlorophyll content and nitrogen level of grassland, reflecting the health status and nutritional needs of grassland. The annual SIF value can be synthesized using the maximum value of the monthly data product of existing solar-induced chlorophyll fluorescence data.

[0099] Figure 3 (e) shows the average status of sunlight-induced chlorophyll fluorescence in Xinjiang grassland from 2000 to 2020. It can be seen that the SIF values ​​in the Ili River Valley and the northern Altai Mountains are higher, reflecting stronger photosynthetic capacity; while the SIF values ​​in the southern arid areas are lower, and the vegetation activity is weaker.

[0100] D. Water Conservation Service (AWC) reflects the ability of grassland ecosystems to maintain and regulate water resources and is a key indicator of grassland health and function. This paper quantifies AWC using the InVEST water production model to calculate water production. Annual water production is calculated based on the water production balance using the following formula:

[0101]

[0102] PET=Kc×ET0

[0103]

[0104] Where Y is water yield (mm); AET is actual evapotranspiration (mm); P is annual precipitation (mm); PET is annual potential evapotranspiration (mm); Kc is crop evapotranspiration coefficient; ET0 is reference crop evapotranspiration (mm); W is the ratio of the annual available water volume of the corrected vegetation to precipitation; Z represents the Zhang coefficient; AWC is the available soil water content (mm). The calculation formula for AWC is as follows:

[0105] AWC=Min(Rest_layer_depth,Root_depth)×PAWC

[0106] PAWC=54.509-0.132×sa-0.003×sa 2 -0.055×si-0.006×si 2 -0.738×cl+0.007×cl 2 -

[0107] 2.688×OM+0.501×OM 2

[0108] Where rest_layer_depth represents the maximum soil depth (mm); root_depth represents the root depth (mm); and PAWC represents vegetation available water, which is determined by the mechanical composition of the soil and the depth of the vegetation root system. It is related to soil texture and type, and reflects the soil's water storage capacity and strength. sa, si, and cl represent the sand, silt, and clay content (%), respectively, and OM represents the organic matter content (%).

[0109] The calculations used monthly precipitation / potential evapotranspiration data from the National Tibetan Plateau Data Center, root-restricting layer depth data from the inversion of Yan Fapeng et al., and soil texture data from the World Soil Harmonized Database. The crop evapotranspiration coefficient / reference crop evapotranspiration data / Zhang coefficient can be tabulated by referring to the corresponding papers for the study area.

[0110] Figure 3 (b) shows the average water production in Xinjiang grasslands from 2000 to 2020. It can be seen that the water production presents a spatial pattern of "high in the north and low in the south, high in the mountains and low in the basins". The high-value areas are concentrated in the Ili River Valley, Altai Mountains and the Central Tianshan Mountains, where water supply is relatively sufficient; while the southern deserts and Kunlun Mountains are short of water resources and have weak water conservation service capabilities.

[0111] E. Soil Conservation Service (SC) reflects the protective effect of grassland vegetation on soil erosion. Its decline usually indicates grassland degradation and reduced quality, and therefore can be used to monitor grassland degradation. Soil erosion is calculated using the universal soil loss equation, and soil conservation service is reflected by the difference between potential soil erosion and actual soil erosion, as shown in the following formula:

[0112] SC=R×K×L×S×(1-C×E)

[0113] Where SC is the soil conservation capacity (unit: t hm -2 a -1 ); R represents the precipitation erosivity factor (unit: MJ mm hm - 2 h -1 a -1 ), calculated by the monthly scale formula proposed by Wischmeier and Smith; K represents the soil erodibility factor (unit: th MJ -1 mm -1), which was calculated using soil organic carbon and particle composition data (EPIC model) proposed by Williams and Arnold; L and S represent the slope length factor (unit: mm) and slope factor (unit: °), respectively; C represents the vegetation cover management factor (dimensionless), which was calculated using the relationship equation between the C factor and vegetation cover FVC established by Cai Chong et al.; E represents the engineering measure factor (dimensionless), and E is taken as 1.

[0114] The calculation formula of precipitation erosion factor is as follows:

[0115]

[0116] Where, P i is the precipitation in the ith month (mm), and P is the annual precipitation (mm).

[0117] The calculation formula of soil erodibility factor is as follows:

[0118] K EIPC =(0.2+0.3×exp -0.0256×sa×(1-si / 100) )×[si / (cl+si)] 0.3 ×[1-0.25×OC

[0119] / (OC+exp 3.72-2.95×OC )]×{1-0.7×(1-sa / 100) / [(1-sa / 100)+exp -5.51+22.9×(1-sa / 100) ]}

[0120] K=0.1317×(0.51575×K EPIC -0.01383)

[0121] Where K EPIC is the soil erodibility factor calculated using the erosion-productivity evaluation model, where sa is the percentage of sand (%), si is the percentage of silt (%), cl is the percentage of clay (%), and OC is the percentage of organic carbon (%).

[0122] The calculation formulas for slope length factor and slope factor are as follows:

[0123]

[0124] Where m is the slope length index, θ is the slope (°), and λ is the slope length (m). The slope and length can be calculated using digital elevation model data in ArcGIS software.

[0125] The calculation formula of vegetation cover management factor is as follows:

[0126]

[0127] Where FVC is the vegetation coverage.

[0128] The calculation used NDVI data from MDOIS, monthly precipitation data from the National Tibetan Plateau Data Center, digital elevation model data released by the National Glacier, Permafrost and Desert Data Center, and soil texture data from the World Soil Harmonized Database.

[0129] Figure 3 (d) shows the average status of grassland soil retention in Xinjiang from 2000 to 2020. It can be seen that the high-value areas are mainly distributed in the northern slope of the Tianshan Mountains and the Altai Mountains. These areas have good vegetation coverage and low soil erosion risks; the low-value areas are concentrated in arid desert areas such as the Tarim Basin and the Turpan Basin, where the soil retention function is poor.

[0130] F. Windbreak and sand fixation (SR) is a key manifestation of grassland ecological functions, directly reflecting its effectiveness in protecting soil and resisting wind erosion. The modified wind erosion equation (RWEQ) is an empirical model used to estimate soil loss due to wind erosion. It reflects SR by measuring the difference between potential wind erosion and actual wind erosion. The formula is as follows:

[0131] SR=SL r -SL

[0132]

[0133] Q rmax =109.8×(WF×EF×SCF×K')

[0134] S r =150.71×(WF×EF×SCF×K') -0.3711

[0135]

[0136] Q max =109.8×(WF×EF×SCF×K'×COG)

[0137] S=150.71×(WF×EF×SCF×K'×COG) -0.3711

[0138] Where SL r is the potential wind erosion amount, SL is the actual wind erosion amount; Q rmax Indicates the maximum sediment transport capacity of potential wind force (kg / m 2 );S r Indicates the length of potential key plot (m); Q max Indicates the maximum sand transport capacity of wind (kg / m 2); S represents the length of the key plot (m); Z represents the distance at which maximum wind erosion occurs (m), which is taken as 50 m based on the relevant research of Xie Gaodi et al.; WF represents the climate factor (kg / m); EF represents the percentage of soil erodibility (%); SCF represents the soil crust factor (dimensionless); K' represents the soil roughness factor (dimensionless); COG represents the vegetation cover factor (dimensionless).

[0139] The calculation formula of climate factor is as follows:

[0140]

[0141] wf=(U2-U1) 2 ×U2×N d

[0142]

[0143] SD=1-P(SnowDepth>25.4mm)

[0144] Where wf represents the wind force factor (m / s 3 ); p is the air density (kg / m 3 ); g is the acceleration due to gravity (9.777m / s 2 ); SW represents the soil moisture factor (dimensionless); SD represents the snow cover factor. U1 is the dust-raising wind speed at 2 m (m / s), which is 5 m / s based on relevant research; U2 is the monthly average wind speed at 2 m (m / s); N d is the number of days with wind speed greater than 5 m / s at 2 m during the observation period; h is the altitude (km); t is the monthly average temperature (℃); ET0 is the potential evapotranspiration (mm); P is the precipitation (mm); I is the irrigation amount (mm) (taken as 0); R d is the number of rainfall events. P(SnowDepth>25.4mm) is the probability that the snow depth is greater than 25.4mm during the measurement period. The annual climate factor WF is obtained by summing the 12 monthly climate factors.

[0145] The calculation formulas for soil erodibility percentage and soil crust factor are as follows:

[0146]

[0147] Where sa is the soil coarse sand content (%); si is the soil silt content (%); cl is the soil clay content (%); OM is the soil organic matter content (%); CaCO3 is the calcium carbonate content (%).

[0148] The calculation formula of vegetation cover factor is as follows:

[0149] COG=e a×FVC

[0150] Where a represents the coefficient of different vegetation types, and the value for grassland is -0.1151.

[0151] The calculation formula of soil roughness factor is as follows:

[0152]

[0153] Where K r is the ridge roughness, calculated using the Smith-Carson equation; Crr is the random roughness factor, which is taken as 0; L is the terrain undulation parameter (which has a different meaning from the slope length factor and should be distinguished); ΔH is the altitude difference within the range of distance L.

[0154] The calculations used daily precipitation and wind speed data from the ERA5-Land reanalysis dataset of the Copernicus Climate Data Repository, monthly precipitation / potential evapotranspiration data from the National Tibetan Plateau Data Center, snow depth data from the Spatiotemporal Tripolar Environment Big Data Platform, soil texture data from the World Soil Coordinated Database, NDVI data from MDOIS, and digital elevation model data released by the National Center for Glacier, Frozen Soil and Desertification.

[0155] Figure 3 (f) shows the average status of grassland windbreak and sand fixation in Xinjiang from 2000 to 2020. It can be seen that high-value areas mainly appear in the Junggar Basin and the southern Kunlun Mountains. Although the vegetation is sparse, the wind is strong and the grassland sand fixation function is obvious; while the windbreak and sand fixation in the Altai Mountains and Tianshan Mountains are relatively low, mainly due to the low local wind speed and limited sand fixation capacity.

[0156] Step 2: Identify ecosystem service bundles and use them to ecologically zonate the study area.

[0157] Given the significant differences in grassland ecosystem service functions across regions, assigning the same weight to ecosystem service indicators across all regions in grassland degradation and restoration assessments can easily overlook regional differences in ecological function and fail to accurately reflect the true state of grassland quality. To address this, this paper identifies ecosystem service bundles and zoning the study area. This allows for regionalized indicator weightings and improves the accuracy and representativeness of assessment results.

[0158] This paper provides two technical approaches to identifying ecosystem service bundles:

[0159] A. Using unsupervised clustering algorithms such as self-organizing maps, we cluster the spatial pixels of the study area based on multiple ecosystem service indicators and automatically divide areas with similar ecosystem service functions.

[0160] B. Zoning identification strategy based on the degree of ecosystem service contribution: functional zoning is performed on the study area according to the spatial dominance of service functions, and the dominant areas of ecosystem service functions are identified. This involves three steps: (1) normalizing the annual average values ​​of the three ecosystem services to 0-1 at the pixel scale; (2) calculating the total amount of ecosystem services for each pixel; (3) sorting the total amount of ecosystem services from high to low according to the size of the three ecosystem service functions, and identifying the key service functions that account for 75% of the total ecosystem service as the ecosystem service bundle. Since 75% is the upper quartile and is often used to measure the degree of concentration of data distribution, this paper assumes that 75% of the total amount of key ecosystem services can reflect the overall state of the ecosystem.

[0161] Grassland ecosystem services in different regions of Xinjiang exhibit significant functional differences. Simply assigning the same weight to ecosystem service indicators across all regions when assessing grassland degradation ignores the differences in grassland ecosystem function across regions and fails to fully reveal the true state of grassland quality. Therefore, the present invention zons regions based on their ecosystem service functions to more accurately determine weights, employing a zoning identification strategy based on the degree of ecosystem service contribution (step B). Figure 4 The spatial distribution of four ecosystem service bundles in Xinjiang is shown. Among them, the water conservation-dominated service bundle accounts for 59.12% of the total area of ​​the study area, and is mainly distributed in the Altai Mountains, the edge of the Tianshan Mountains and the northern Kunlun Mountains; the water conservation-soil conservation composite-dominated service bundle accounts for 20.03%, and is mainly distributed in the Ili River Valley, the central Tianshan Mountains and the southwestern Kunlun Mountains; the water conservation-wind and sand fixation composite-dominated service bundle accounts for 16.26%, and is mainly distributed in the Junggar Basin and the southeastern Kunlun Mountains; the wind and sand fixation-dominated service bundle accounts for 4.59%, and is mainly distributed in the Junggar Basin and the southernmost Kunlun Mountains.

[0162] Step 3: Determine the single factor indicator weights for each ecological zone and calculate the grassland health and service comprehensive index based on the single factor indicator weights.

[0163] A. Because each individual factor has a different measurement unit and data range, it is necessary to standardize the individual factor indicators before constructing the grassland health and service comprehensive index. The standardization formula is as follows:

[0164]

[0165] Where N is the standardized result, X is the original value of the single factor obtained in step 1, and X max and X min are the maximum and minimum values ​​of the single factor index respectively.

[0166] B. Principal component analysis is a data dimension reduction method that converts related multiple variables into several unrelated comprehensive indicators through linear transformation. This method can extract the effective information of the variables to the greatest extent, and at the same time, objectively assign weights to each variable based on its contribution to the principal component. After standardization is completed, the present invention can use principal component analysis, entropy weight method and other methods to perform principal component transformation on the standardized single factor result layer (three ecosystem structure indicators and dominant ecosystem service indicators) of each ecosystem service bundle to obtain the variance contribution rate of each principal component and the weight coefficient of the single factor. Principal component analysis condenses original indicators such as vegetation coverage into a few principal components, which carry more than 90% of the effective information of the original data, thereby eliminating redundancy and simplifying calculations while retaining the core features. The weight calculation formula for a single factor is as follows:

[0167]

[0168] Where W i is the weight of the i-th single factor, m is the number of single factors (the number of single factors in different ecological zones is different), n is the number of principal components with a cumulative contribution rate exceeding 90%, α ij is the weight of the i-th single factor in the j-th principal component, β j is the contribution rate of the jth principal component.

[0169] The formula for the grassland health and service comprehensive index GHSI is as follows:

[0170]

[0171] Where W i is the weight of the i-th single factor, N i is the ith standardized indicator.

[0172] Figure 5 The spatial distribution of the Grassland Health and Services Index (GHSI) for Xinjiang grasslands from 2000 to 2020 is visualized through the multi-year average spatial distribution. The generation process involves three steps: first, inversion of six single-factor indicators, such as vegetation cover and net primary productivity, based on multi-source remote sensing and meteorological data; second, principal component analysis is used within the framework of ecosystem service bundle zoning to determine the weights of the indicators within each zone; and finally, the multi-year average GHSI for each pixel is calculated using a weighted summation formula and mapped. High values ​​are concentrated in the Ili River Valley, the central Tianshan Mountains, the Junggar Mountains, and the northern Altai Mountains, while low values ​​are concentrated in the Junggar Basin and the Kunlun Mountains.

[0173] Step 4: Use the grassland health and service comprehensive index to evaluate grassland degradation and restoration levels.

[0174] The present invention constructs a comprehensive index of grassland health and services, and uses its interannual change rate as the core indicator to measure the degradation and recovery status of grassland. Grassland degradation is a relative concept, and its grade evaluation requires the determination of a benchmark value and a grading threshold. The present invention quantitatively characterizes the grassland degradation status through the comprehensive index change rate, adopts the multi-year average comprehensive index GHSI as the benchmark value, and divides the grassland into five levels: severe degradation, moderate degradation, mild degradation, non-degradation and recovery point. Specifically, the average annual change rate of the comprehensive index GHSI from 2000 to 2020 is calculated pixel by pixel, and the average value from 2000 to 2020 is used as the benchmark value to characterize the long-term change trend of the grassland ecological status. In terms of dividing the degradation level, the present invention combines the existing degradation classification standards and relevant literature research results, and adopts the ordered sample optimal segmentation method in cluster analysis to classify the comprehensive index GHSI change rate, and quantitatively divides each degradation level to ensure the scientificity and objectivity of the degradation level division. The ordered sample optimal segmentation method is an optimization technology for cluster analysis of data with natural order, and is widely used to classify ordered data. In the present invention, the optimal segmentation method for ordered samples can automatically identify the intrinsic characteristics of the data and determine the optimal segmentation point according to the actual situation of grassland degradation, automatically identify the optimal segmentation point in the comprehensive index GHSI change rate, minimize the intra-class differences and enhance the inter-class distinction ability, thereby achieving effective distinction of the grassland ecological change state. The optimal segmentation method for ordered samples ensures that the distinction between different degradation levels is more accurate and representative by minimizing the variability within each level and maximizing the differences between the levels. In the specific implementation, the GHSI change rates of all pixels are first arranged in ascending order; then the ecological status is divided according to the positive and negative values ​​of the change rate: when the change rate is greater than zero, it indicates that the grassland ecological status is stable or recovered, and it is subdivided into two levels: "non-degraded" and "recovered"; when the change rate is less than zero, it characterizes the degradation process, and is specifically divided into four levels: "non-degraded", "mildly degraded", "moderately degraded" and "severely degraded". The specific grading thresholds are shown in the following table:

[0175]

[0176] Figure 6 Based on the change rate of the Grassland Health and Services Index (GHSI) at the pixel level across Xinjiang from 2000 to 2020, the region was categorized into five levels: severely degraded, moderately degraded, slightly degraded, not degraded, and recovered, using the ordered sample optimal segmentation method. This dynamic map reveals the overall characteristics of Xinjiang's grassland ecosystem transition from a degradation-dominated trend to recovery over the past two decades: The early period (2000-2009) showed alternating fluctuations between degradation and initial recovery, with a fragmented spatial distribution. The later period (2010-2020) saw an overall improvement, despite continued localized degradation pressures.

[0177] The present invention makes full use of multi-source remote sensing data and meteorological data, integrates key indicators of the two dimensions of ecological structure and ecological services, and breaks through the problems of traditional grassland degradation evaluation methods such as reliance on a single ecological factor, coarse spatial granularity, and strong subjectivity of classification standards. It has stronger scientificity and applicability. Especially in large-scale regional grassland monitoring and management, it can provide important technical support for grassland degradation prevention and control, ecological compensation assessment, and restoration project effectiveness monitoring. The present invention constructs a grassland degradation status identification and classification method with strong operability, good versatility, and high evaluation result accuracy. It can not only be applied to grassland ecological monitoring practices in Xinjiang, but also provides a new technical path and method reference for grassland health evaluation in other ecologically fragile areas.

[0178] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principles of the present invention should be included in the scope of protection of the present invention.

Claims

1. A method for evaluating grassland degradation and restoration status based on comprehensive ecological structure and service indicators, characterized in that: The steps are as follows: Step 1: Construct a grassland ecological indicator dataset based on multi-source remote sensing data, soil data, and meteorological data. Based on the data in the grassland ecological indicator dataset, invert single-factor monitoring indicators including ecosystem structure indicators and ecosystem service indicators; Step 2: Identify ecosystem service bundles and use them to ecologically zonate the study area; Step 3: Determine the weight of the single-factor monitoring indicator for each ecological zone, and calculate the grassland health and service comprehensive index based on the weight of the single-factor monitoring indicator; Step 4: Use the grassland health and service comprehensive index to evaluate grassland degradation and restoration levels.

2. The method for evaluating grassland degradation recovery status based on integrated ecological structure and service indicators according to claim 1 is characterized in that: The ecosystem structure indicators include vegetation cover, net primary productivity and sunlight-induced chlorophyll fluorescence, and the ecosystem service indicators include water production, soil conservation and wind and sand fixation.

3. The method for evaluating grassland degradation recovery status based on comprehensive ecological structure and service indicators according to claim 2 is characterized in that: The pixel dichotomy method was used to calculate the vegetation coverage of the study area; The CASA model was used to calculate net primary productivity; The sunlight-induced chlorophyll fluorescence is used to evaluate the chlorophyll content and nitrogen level of the grassland, reflecting the health status and nutritional needs of the grassland. The maximum value of the monthly data product of the existing solar-induced chlorophyll fluorescence data is used to synthesize the annual solar-induced chlorophyll fluorescence value. Water conservation services reflect the ability of grassland ecosystems to maintain and regulate water resources and are an important indicator of grassland health and function. Water conservation services are quantified and calculated using the InVEST water production model. Soil conservation services reflect the protective effect of grassland vegetation on soil erosion and are used to monitor grassland degradation; the universal soil loss equation calculates soil erosion. Windbreak and sand fixation services are an important manifestation of grassland ecological functions, which directly reflect the effectiveness of grassland in protecting soil and resisting wind erosion. The modified wind erosion equation is used to estimate windbreak and sand fixation services.

4. The method for evaluating grassland degradation recovery status based on integrated ecological structure and service indicators according to claim 3 is characterized in that: The vegetation coverage is: Where, NDVI soil is the NDVI value of the bare soil pixel, NDVI veg is the NDVI value of the pure vegetation cover pixel, NDVI is the NDVI value of the study area pixel; the 1% and 99% probability of the annual NDVI value range distribution are used as NDVI soil and NDVI veg The value of The method for calculating net primary productivity using the CASA model is: NPP=APAR×LUE APAR=0.5×SOL×FPAR LUE=T ε1 ×T ε2 ×W ε ×LUE max Where APAR represents the photosynthetically active radiation absorbed by vegetation; LUE represents the actual light energy utilization efficiency. FPAR represents the ratio of incident photosynthetically active radiation absorbed by the vegetation layer; the constant 0.5 represents the ratio of solar active radiation that can be used by vegetation, that is, the ratio of solar active radiation to the total solar radiation SOL; T ε1 and T ε2 represents temperature stress factor; W ε Represents water stress factor; LUE max Indicates the maximum light energy utilization under ideal conditions; The calculation method of the InVEST water production model is: PET=Kc×ET0 Where Y is water yield; AET is actual evapotranspiration; P is annual precipitation; PET is annual potential evapotranspiration; Kc is crop evapotranspiration coefficient; ET0 is reference crop evapotranspiration; W is the ratio of annual vegetation available water to precipitation; Z represents Zhang coefficient; AWC is soil available water content; The soil conservation service is the difference between the potential soil erosion and the actual soil erosion. The soil conservation service is: SC=R×K×L×S×(1-C×E) Where R represents the precipitation erosivity factor, which is calculated using the monthly formula proposed by Wischmeier and Smith; K represents the soil erodibility factor, which is calculated using soil organic carbon and particle composition data proposed by Williams and Arnold; L and S represent the slope length factor and slope factor, respectively; C represents the vegetation cover management factor, which is calculated using the relationship equation between the C factor and vegetation cover fraction (FVC); and E represents the engineering measure factor. The wind and sand fixation service is reflected by the difference between the potential wind erosion amount and the actual wind erosion amount. SR=SL r -SL Q rmax =109.8×(WF×EF×SCF×K') S r =150.71×(WF×EF×SCF×K') -0.3711 Q max =109.8×(WF×EF×SCF×K'×COG) S=150.71×(WF×EF×SCF×K'×COG) -0.3711 Where SL r is the potential wind erosion amount, SL is the actual wind erosion amount; Q rmax Indicates the maximum sediment transport capacity of potential wind force; S r represents the length of potential key plot; Q max represents the maximum sediment transport capacity of wind; S represents the length of the key plot; Z represents the distance where maximum wind erosion occurs; WF represents the climate factor; EF represents the percentage of soil erodibility; SCF represents the soil crust factor; K' represents the soil roughness factor; COG represents the vegetation cover factor.

5. The method for evaluating grassland degradation restoration status based on any one of the structural and service indicators of claims 1 to 4, characterized in that: The method for identifying the ecosystem service bundle includes: A. Using the unsupervised clustering algorithm of self-organizing maps, we clustered the spatial pixels of the study area based on multiple single-factor monitoring indicators of ecosystem service indicators and automatically divided the areas with similar ecosystem service functions; B. Based on the zoning identification strategy of the degree of ecosystem service contribution, the study area is functionally zoned according to the spatial dominance of service functions to identify the dominant areas of ecosystem service functions.

6. The method for evaluating grassland degradation recovery status based on integrated ecological structure and service indicators according to claim 5 is characterized in that: The implementation method of the zoning identification strategy based on the degree of ecosystem service contribution is as follows: (1) at the pixel scale, the annual average values ​​of the single-factor monitoring indicators of the three ecosystem service indicators are normalized to 0-1; (2) the total amount of the ecosystem service indicator of each pixel is calculated; (3) According to the size of the three ecosystem service functions, the total amount of ecosystem services is ranked from high to low, and the single factor monitoring indicators of key ecosystem service indicators that account for 75% of the total ecosystem service amount are identified as the ecosystem service bundle.

7. The method for evaluating grassland degradation recovery status based on integrated ecological structure and service indicators according to any one of claims 1 to 4 and 6, characterized in that: The method for determining the weight of the single factor monitoring indicator of each ecological zone is: Standardize single factor monitoring indicators; Using principal component analysis or entropy weight method, principal component transformation is performed on the standardized single factor result layer of each ecosystem service bundle to obtain the variance contribution rate of each principal component and the weight coefficient of the single factor monitoring indicator; The weight coefficient is: Where W i is the weight of the i-th single factor monitoring indicator, m is the number of single factor monitoring indicators, n is the number of principal components with a cumulative contribution rate exceeding 90%, α ij is the weight of the i-th single factor in the j-th principal component, β j is the contribution rate of the jth principal component; The single factor result layer includes three ecosystem structure indicators and the dominant ecosystem service indicator.

8. The method for evaluating grassland degradation recovery status based on integrated ecological structure and service indicators according to claim 7 is characterized in that: The calculation formula of the grassland health and service comprehensive index GHSI is: Where W i is the weight of the i-th single factor monitoring indicator, N i is the i-th standardized single factor monitoring indicator.

9. The method for evaluating grassland degradation recovery status based on integrated ecological structure and service indicators according to claim 8, characterized in that: The annual average change rate of the grassland health and service comprehensive index (GHSI) was calculated pixel by pixel, and the average annual change rate was used as the benchmark value to represent the long-term change trend of grassland ecological status. The ordered sample optimal segmentation method in cluster analysis was used to classify the annual average change rate of the grassland health and service comprehensive index (GHSI), and the grassland was divided into five levels: severely degraded, moderately degraded, slightly degraded, non-degraded, and restored.

10. The method for evaluating grassland degradation restoration status based on integrated ecological structure and service indicators according to claim 9, characterized in that: The annual average change rate of the grassland health and service comprehensive index (GHSI) of all pixels is arranged in ascending order; the ecological status is divided according to the positive and negative values ​​of the change rate: when the change rate is greater than zero, it indicates that the grassland ecological status is stable or recovered, and is subdivided into two levels: "non-degraded" and "recovered"; when the change rate is less than zero, it indicates the degradation process, and is divided into four levels: "non-degraded", "mildly degraded", "moderately degraded" and "severely degraded".

Images