Grassland ecosystem drought vulnerability dynamic assessment method based on multi-source remote sensing data and multi-dimensional indexes
By integrating multi-source remote sensing data and using a multi-dimensional indicator system, the problems of data discontinuity and limitations of single indicators in the assessment of drought sensitivity of grassland ecosystems have been solved. This has enabled high-precision assessment of grassland drought vulnerability and identification of priority restoration areas, thereby improving the scientific nature of the assessment and the operability of management decisions.
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2025-11-28
- Publication Date
- 2026-03-10
AI Technical Summary
Existing technologies for assessing drought sensitivity in grassland ecosystems suffer from data discontinuity and limitations of single indicators, resulting in insufficient assessment accuracy and difficulty in accurately reflecting the spatiotemporal heterogeneity of vegetation responses and generating spatial distribution maps to guide ecological restoration.
By employing multi-source remote sensing data fusion technology, a dataset of total primary productivity and leaf area index of grassland was constructed. Combined with a multi-dimensional indicator system, including resistance, resilience and sensitivity indicators, a spatial distribution map of grassland drought vulnerability was generated to achieve high-precision assessment and identification of priority restoration areas.
It enables high-precision assessment of the drought vulnerability of grassland ecosystems, provides scientific spatial distribution maps, enhances the scientific nature of the assessment and the operability of management decisions, and supports ecological restoration and adaptive management in arid areas.
Smart Images
Figure CN121638656A_ABST
Abstract
Description
Technical Field
[0001] This invention belongs to the field of global change ecology technology, and in particular relates to a method for dynamic assessment of drought vulnerability of grassland ecosystems based on multi-source remote sensing data and multi-dimensional indicators. Background Technology
[0002] Against the backdrop of global climate change, drought events are becoming more frequent and intense, profoundly impacting grassland ecosystems. Current research faces two main technical bottlenecks: Firstly, at the data level, existing studies generally rely on single data sources, whose spatiotemporal discontinuities lead to significant errors in monitoring results, making it difficult to accurately capture the spatiotemporal heterogeneity of vegetation responses, particularly the significant differences in drought sensitivity among different grassland types. Furthermore, the complex cumulative and lag effects of drought on vegetation further increase the difficulty of assessment. Secondly, in terms of analytical methods, current vulnerability assessments primarily employ single-indicator methods, which cannot comprehensively reflect the multidimensional characteristics of ecosystems, nor can they generate spatial distribution maps that directly guide ecological restoration practices, making it difficult for management departments to accurately identify priority protection areas. These technical limitations severely restrict the scientific understanding and precise management of the drought resistance capacity of grassland ecosystems. Therefore, to address the insufficient accuracy issues caused by data discontinuity and single-indicator limitations in current grassland drought sensitivity assessments, it is urgent to propose a dynamic assessment method for grassland ecosystem drought vulnerability based on multi-source remote sensing data and multidimensional indicators. Summary of the Invention
[0003] To address the aforementioned technical challenges, this invention proposes a dynamic assessment method for the drought vulnerability of grassland ecosystems based on multi-source remote sensing data and multi-dimensional indicators. By integrating multi-source data, it overcomes the bottleneck of discontinuous spatiotemporal data and achieves high-precision spatial assessment of grassland drought response. Simultaneously, by employing a multi-dimensional indicator system, it comprehensively reflects the drought vulnerability characteristics of the ecosystem and generates a spatial distribution map to guide ecological restoration, greatly enhancing the scientific rigor and practicality of the assessment.
[0004] To achieve the above objectives, this invention provides a method for dynamic assessment of drought vulnerability of grassland ecosystems based on multi-source remote sensing data and multi-dimensional indicators, comprising: By integrating multi-source remote sensing data, a dataset of total primary productivity and leaf area index of grassland were constructed. Based on the total primary productivity dataset of grassland and the leaf area index dataset, drought events are identified and drought features are extracted to quantify the cumulative and lagged effects of grassland ecosystems on drought. Based on the cumulative and lagged effects of drought, a multidimensional vulnerability index system including resistance, resilience, and sensitivity indicators was constructed to dynamically assess the drought vulnerability of grassland ecosystems and obtain assessment results. Based on the assessment results mapped to RGB three channels, a spatial distribution map of grassland drought vulnerability is generated by combining the values of the three channels, thereby realizing vulnerability classification and identification of priority restoration areas.
[0005] Optional, the fusion of multi-source remote sensing data includes: Acquire leaf area index data, total primary productivity data of grassland, and meteorological grid data; By integrating the leaf area index data and the total primary productivity data of grassland, a standardized precipitation evapotranspiration index is calculated, and all data are unified to the same spatial grid and time series.
[0006] Optionally, identifying drought events based on drought indices includes: Based on the pixel-by-pixel standardized precipitation evapotranspiration index time series, independent drought events are identified using run theory; For each identified drought event, calculate its drought duration, drought severity, drought intensity, and peak value.
[0007] Optionally, drought features can be extracted, including: The cumulative and lag effects of drought on vegetation were analyzed using the maximum correlation coefficient method. The cumulative effect is characterized by the maximum cumulative correlation coefficient and its corresponding optimal number of cumulative months; The lag effect is characterized by the maximum lag correlation coefficient and its corresponding optimal lag month.
[0008] Optional, quantified cumulative effects include: Calculate the Pearson correlation coefficient between total primary productivity or leaf area index of grassland during the growing season and SPEI on a 1-12 month scale, and extract the maximum cumulative correlation coefficient and the maximum cumulative time scale.
[0009] Optional, sensitivity metrics can be constructed as follows: The standardized maximum cumulative correlation coefficient, maximum lag correlation coefficient, cumulative time, and lag time are used to calculate the comprehensive sensitivity index using the arithmetic mean.
[0010] Optionally, constructing resilience indicators includes calculating the relative deviation of vegetation standardized outliers in the most severe month during the drought from the pre-drought baseline.
[0011] Optional resilience metrics include: Calculate the recovery rate of vegetation standardized outliers in the optimal recovery month after the drought ends and the outliers at the start of the drought.
[0012] Optionally, generating a spatial distribution map of grassland drought vulnerability includes: The resistance index, the sensitivity index, and the resilience index are mapped to the red, green, and blue channels, respectively. The values are standardized to a preset range, and a composite color spectrum is generated. Then, the Jenks natural breakpoint method is used to divide the five-level vulnerability gradient.
[0013] Technical advantages of this invention: This invention discloses a dynamic assessment method for drought vulnerability of grassland ecosystems based on multi-source remote sensing data and multi-dimensional indicators. It provides a high spatiotemporal resolution grassland dataset based on multi-source remote sensing data; and provides quantitative assessment results that comprehensively reflect the drought response characteristics of grassland ecosystems based on a constructed three-dimensional indicator system of resistance, sensitivity, and resilience. This enables the hierarchical classification of grassland vulnerability and the identification of priority restoration areas, thereby significantly improving the scientific nature of grassland drought vulnerability assessment and the operability of management decisions. This method provides reliable technical support for the ecological restoration and adaptive management of grasslands in arid regions. Attached Figure Description
[0014] The accompanying drawings, which form part of this application, are used to provide a further understanding of this application. The illustrative embodiments and descriptions of this application are used to explain this application and do not constitute an undue limitation of this application. In the drawings: Figure 1 This is a flowchart illustrating a dynamic assessment method for drought vulnerability of grassland ecosystems based on multi-source remote sensing data and multi-dimensional indicators, according to an embodiment of the present invention. Figure 2 This is a schematic diagram of the spatial distribution of grassland vulnerability according to an embodiment of the present invention, wherein (a) is the grassland leaf area index (LAI), (b) is the RGB composite index of grassland primary productivity (GPP) vulnerability, (c) is the spatial classification of grassland LAI vulnerability, and (d) is the spatial classification of GPP vulnerability. Detailed Implementation
[0015] It should be noted that, unless otherwise specified, the embodiments and features described in this application can be combined with each other. This application will now be described in detail with reference to the accompanying drawings and embodiments.
[0016] It should be noted that the steps shown in the flowchart in the accompanying drawings can be executed in a computer system such as a set of computer-executable instructions, and although a logical order is shown in the flowchart, in some cases the steps shown or described may be executed in a different order than that shown here.
[0017] like Figure 1 As shown, this embodiment provides a method for dynamic assessment of drought vulnerability of grassland ecosystems based on multi-source remote sensing data and multi-dimensional indicators, including: By integrating multi-source remote sensing data, a dataset of total primary productivity and leaf area index of grassland were constructed. Based on the total primary productivity dataset of grassland and the leaf area index dataset, drought events are identified and drought features are extracted to quantify the cumulative and lagged effects of grassland ecosystems on drought. Based on the cumulative and lagged effects of drought, a multidimensional vulnerability index system including resistance, resilience, and sensitivity indicators was constructed to dynamically assess the drought vulnerability of grassland ecosystems and obtain assessment results. Based on the assessment results mapped to RGB three channels, a spatial distribution map of grassland drought vulnerability is generated by combining the values of the three channels, thereby realizing vulnerability classification and identification of priority restoration areas.
[0018] Furthermore, the fusion of multi-source remote sensing data includes: Acquire leaf area index data, total primary productivity data of grassland, and meteorological grid data; By integrating the leaf area index data and the total primary productivity data of grassland, a standardized precipitation evapotranspiration index is calculated, and all data are unified to the same spatial grid and time series.
[0019] Specifically, one implementation of this embodiment is as follows: (1) Integrate four LAI datasets: GIMMS, GLOBMAP, GLASS, and MODIS. Integrate four GPP datasets: ECLUE, NIRv, MuSyQ, and MODIS. The temporal resolution is 1982-2020. (2) Unify the spatial resolution to 0.05°×0.05° and construct a monthly spatiotemporally continuous grassland LAI and GPP dataset from 1982 to 2020. A drought event database is established using CRU TS meteorological data and the SPEI drought index.
[0020] Specifically, the data preprocessing stage in this embodiment integrates multi-source remote sensing data to construct a high-precision, spatiotemporally continuous grassland monitoring dataset, laying the data foundation for subsequent vulnerability assessment. This stage includes three core steps: constructing a leaf area index dataset, constructing a total primary productivity dataset, and preparing drought index data.
[0021] LAI Dataset Construction and Fusion: First, four LAI datasets from different sources were acquired: GIMMS LAI dataset (time span 1982-2020, spatial resolution 8km, temporal resolution 15 days), GLOBMAP LAI dataset (time span 1982-2020, spatial resolution 8km, temporal resolution monthly), GLASS LAI dataset (time span 2000-2020, spatial resolution 1km, temporal resolution 8 days), and MODIS LAI dataset (time span 2000-2020, spatial resolution 500m, temporal resolution 4 days). To address the missing data in the GLASS LAI dataset for January-February 2000, interpolation was performed using the monthly average values from the same period 2000-2020. Subsequently, all LAI datasets were uniformly resampled to a spatial resolution of 0.05°×0.05°, and the temporal resolution was standardized to a monthly scale. Finally, the ensemble mean of the four datasets was calculated to generate a spatiotemporally continuous grassland LAI time series from 1982-2020.
[0022] GPP Dataset Construction and Fusion: Four GPP datasets were acquired and fused: ECLUE GPP dataset (time span 1982-2018, spatial resolution 0.5°, temporal resolution month), NIRv GPP dataset (time span 1982-2018, spatial resolution 0.05°, temporal resolution month), MuSyQ GPP dataset (time span 2001-2020, spatial resolution 0.05°, temporal resolution month), and MODIS GPP dataset (time span 2000-2020, spatial resolution 500m, temporal resolution 8 days). All GPP datasets were uniformly resampled to a spatial resolution of 0.05°×0.05°. By calculating the ensemble mean of the four datasets, a grassland GPP time series database from 1982 to 2020 was constructed.
[0023] Furthermore, identifying drought events based on drought indices includes: Based on the pixel-by-pixel standardized precipitation evapotranspiration index time series, independent drought events are identified using run theory; For each identified drought event, calculate its drought duration, drought severity, drought intensity, and peak value.
[0024] The characteristics of drought extracted include: The cumulative and lag effects of drought on vegetation were analyzed using the maximum correlation coefficient method. The cumulative effect is characterized by the maximum cumulative correlation coefficient and its corresponding optimal number of cumulative months; The lag effect is characterized by the maximum lag correlation coefficient and its corresponding optimal lag month.
[0025] Specifically, an implementation method of this embodiment: Obtain the CRU TS v4.06 meteorological dataset (time span from 1982 to 2020, spatial resolution 0.5°, temporal resolution monthly), and extract monthly precipitation (PRE) and potential evapotranspiration (PET) data. Based on the United Nations Environment Programme standards, calculate the drought index AI = PRE / PET, and divide the study area into four drought gradients: arid region (AR, AI < 0.20), semi-arid region (SAR, 0.20 < AI < 0.50), dry sub-humid region (DSH, 0.50 < AI < 0.65), and humid region (HU, AI > 0.65). At the same time, obtain the drought index data at the SPEI1-12 scale to provide basic data support for subsequent drought event identification and response feature analysis.
[0026] Quantification implementation of the drought response characteristics of the grassland ecosystem: Quantifying the response characteristics of the grassland ecosystem to drought is the core technical link of this embodiment. By systematically analyzing the temporal dynamic relationship between drought event characteristics and vegetation responses, the sensitivity mechanism of the grassland to drought stress is revealed. This stage includes drought event identification and feature extraction, cumulative effect analysis, and lag effect analysis.
[0027] Drought event identification and feature extraction stage: Based on the run theory, establish a drought event identification standard. Define a period with continuous negative SPEI-12 values for more than 3 months and the minimum SPEI value during this period less than or equal to -1 as an independent drought event. For each identified drought event, the system extracts four core feature indicators: drought duration (number of months from the start to the end of the drought event), drought severity (sum of the absolute values of all negative SPEI values during the drought), drought intensity (obtained by dividing the drought severity by the drought duration), and drought peak (the minimum value of SPEI during the drought). These indicators comprehensively characterize the temporal evolution characteristics and intensity distribution patterns of drought events.
[0028] Furthermore, the quantification of the cumulative effect includes: Calculate the Pearson correlation coefficient between the total primary productivity or leaf area index of the grassland in the growing season and SPEI at the 1-12 month scale, and extract the maximum cumulative correlation coefficient and the maximum cumulative time scale.
[0029] Specifically, an implementation method of this embodiment: Grassland LAI and GPP data from the growing season (April–October) were extracted as proxy variables for vegetation activity to establish the response relationship between vegetation indices and drought conditions at different time scales. Pearson correlation coefficients between grassland LAI and GPP and SPEI at 1–12-month time scales were calculated, and the maximum cumulative correlation coefficient and the maximum cumulative time scale were obtained by traversing all time scales. This analysis reveals the cumulative response sensitivity of grassland vegetation to drought conditions of different durations, providing a quantitative basis for understanding the temporal cumulative effects of drought stress.
[0030] Further analysis of the lag effect: Using lag correlation analysis, the correlation coefficients between grassland LAI and GPP in the current month and SPEI-1 values from the previous 1-12 months were calculated to identify the time lag characteristics of vegetation response to drought stress. By systematically analyzing the correlation strength at different lag times, the maximum lag correlation coefficient and its corresponding maximum lag time were extracted. This analysis quantifies the lag effect of drought stress on grassland vegetation, reveals the time lag mechanism of ecosystem response to drought disturbance, and provides a scientific basis for accurately assessing the persistence and recovery time of drought impacts.
[0031] Furthermore, the dynamic assessment of grassland ecosystem drought vulnerability using multidimensional indicators involves the following steps: Resistance: Calculate the resistance to LAI / GPP anomalies during drought.
[0032] Resilience: Quantifying the rate of vegetation function recovery after drought.
[0033] Sensitivity: A comprehensive sensitivity index is constructed based on the standardized maximum cumulative correlation coefficient, the standardized maximum lag correlation coefficient, the standardized cumulative time, and the standardized lag time.
[0034] A spatial visualization of grassland drought vulnerability was achieved using an RGB three-channel synthesis method: resistance (normalized from 0-255), sensitivity, and resilience were mapped to the red, green, and blue primary color channels, respectively, and a comprehensive vulnerability map was generated by combining the values from these three channels. Based on Jenks' natural breakpoint method, vulnerability was divided into five levels (insensitive, low-sensitive, medium-sensitive, high-sensitive, and extremely sensitive), and combined with the dichotomy (high / low) of the three-dimensional indicators to generate a combination map of eight vulnerability types. This visually identifies high-risk areas that require priority restoration, providing a scientific basis for the spatially targeted deployment of ecological restoration projects.
[0035] Construction and implementation of a multidimensional vulnerability indicator system: The core innovation of this embodiment lies in the construction of a multidimensional vulnerability index system. By integrating three dimensions—resistance, sensitivity, and resilience—a comprehensive assessment framework is established to fully reflect the drought vulnerability characteristics of grassland ecosystems. This system represents a technological breakthrough from single-indicator to multidimensional collaborative assessment and generates intuitive spatial decision maps using RGB three-channel visualization technology.
[0036] Furthermore, the construction of sensitivity indicators includes: The standardized maximum cumulative correlation coefficient, maximum lag correlation coefficient, cumulative time, and lag time are used to calculate the comprehensive sensitivity index using the arithmetic mean.
[0037] Specifically, one implementation of this embodiment is as follows: Based on the analysis results of early-stage cumulative and lag effects, a comprehensive sensitivity index (DSG) is constructed. First, the four core parameters are standardized: the standardized value of the maximum cumulative correlation coefficient (R0). max-cum-nor Standardized value of maximum lag correlation coefficient (R) max-lag-nor ), cumulative time standardized value (T) cum-nor ) and standardized value of lag time (T) lag-nor The standardization formula uses the min-max standardization method to unify the value range of each parameter to 0-1. The comprehensive sensitivity index is calculated by the arithmetic mean of the four standardized parameters, i.e., DSG = (R... max-cum-nor + R max-lag-nor + T cum-nor + T lag-nor ) / 4. This index comprehensively reflects the temporal response sensitivity and spatial heterogeneity of grasslands to drought stress.
[0038] Furthermore, constructing resilience indices includes: calculating the relative deviation between the standardized vegetation outliers of the most severe month during the drought and the baseline values before the drought. Constructing resilience indices includes: calculating the recovery rate between the standardized vegetation outliers of the best recovery month after the drought and the outliers at the start of the drought.
[0039] Specifically, one implementation of this embodiment is as follows: First, the standardized outliers of grassland LAI and GPP from 1982 to 2020 are calculated using the following formula: (1); Where SAt represents the standardized outlier of the grassland leaf area index (GPP) in month t. This represents the leaf area index (GPP) value for month t. and These represent the mean and standard deviation of grassland leaf area index (GPP) from 1982 to 2020, respectively.
[0040] The formulas for calculating resistance and resilience are as follows: (2); (3); (4); in, Outliers in the grassland leaf area index (GPP) for the month with the most severe drought. This indicates anomalies in the grassland leaf area index (GPP) prior to the occurrence of drought. This indicates anomalies in the grassland leaf area index (GPP) at the onset of drought; This indicates anomalies in the grassland leaf area index (GPP) one month before the onset of drought; This represents the standardized outliers of the grassland leaf area index (GPP) during the optimal recovery month; Indicates resistance; It indicates resilience.
[0041] Furthermore, generating a spatial distribution map of grassland drought vulnerability includes: The resistance index, the sensitivity index, and the resilience index are mapped to the red, green, and blue channels, respectively. The values are standardized to a preset range, and a composite color spectrum is generated. Then, the Jenks natural breakpoint method is used to divide the five-level vulnerability gradient.
[0042] Specifically, one implementation of this embodiment is as follows: RGB Three-Channel Visualization and Vulnerability Grading Implementation: RGB three-channel synthesis technology is used to achieve spatial visualization of multi-dimensional vulnerability. The three core indicators—resistance, sensitivity, and resilience—are standardized to a numerical range of 0-255 and assigned values to the red (R), green (G), and blue (B) color channels respectively. Different combinations of the three channel values generate a composite color spectrum reflecting the overall vulnerability characteristics. The median segmentation method is used to divide each dimension indicator into high and low levels, forming a spatial distribution pattern of 2³=8 vulnerability type combinations. Finally, the Jenks natural breakpoint method is used to classify overall vulnerability into five levels: insensitive, low sensitive, moderately sensitive, highly sensitive, and extremely sensitive, directly identifying high-risk areas requiring priority restoration and providing a scientific decision-making basis for the spatially targeted deployment of ecological restoration projects.
[0043] The experimental results of this embodiment: A grassland drought vulnerability assessment database with a resolution of 0.05° was constructed for the period 1982–2020. For example... Figure 2As shown, (a) represents the grassland leaf area index (LAI), (b) represents the RGB composite index of grassland primary productivity (GPP) vulnerability, (c) represents the spatial classification of grassland LAI vulnerability, and (d) represents the spatial classification of GPP vulnerability. The most vulnerable combination—low resistance-high sensitivity-low resilience (Tlow-Shigh-Rlow)—accounts for 6.41% and 5.55% of the grassland LAI and GPP assessment areas, respectively, and therefore requires priority protection. Figure 2 (c) and (b)). The low-resistance-high-sensitivity-high-toughness (Tlow-Shigh-Rhigh) and high-resistance-high-sensitivity-low-toughness (Thigh-Shigh-Rlow) regions also exhibited significant vulnerability. The total area of these three highly vulnerable types accounted for 40.84% and 40.46% of the LAI and GPP assessment areas, respectively, indicating widespread drought vulnerability risk in the Central Asian steppes. Furthermore, these three highly vulnerable types were found to be mainly concentrated in semi-arid regions, accounting for 60.06% (LAI) and 58.65% (GPP) of the vulnerable areas, respectively. This is consistent with the observed highest sensitivity in semi-arid grasslands, further indicating that the Central Asian semi-arid grasslands are extremely vulnerable and should be prioritized for protection measures. This embodiment verifies the effectiveness and practicality of this embodiment in large-scale grassland drought vulnerability assessment, providing a scientific decision support tool for grassland ecological protection and restoration in arid areas.
[0044] This invention discloses a dynamic assessment method for drought vulnerability of grassland ecosystems based on multi-source remote sensing data and multi-dimensional indicators. It provides a high spatiotemporal resolution grassland dataset based on multi-source remote sensing data; and offers quantitative assessment results that comprehensively reflect the drought response characteristics of grassland ecosystems based on a constructed three-dimensional indicator system of resistance, sensitivity, and resilience. This enables the hierarchical classification of grassland vulnerability and the identification of priority restoration areas, thereby significantly improving the scientific rigor of grassland drought vulnerability assessment and the operability of management decisions. This method provides reliable technical support for the ecological restoration and adaptive management of grasslands in arid regions.
[0045] The above are merely preferred embodiments of this application, but the scope of protection of this application is not limited thereto. Any variations or substitutions that can be easily conceived by those skilled in the art within the scope of the technology disclosed in this application should be included within the scope of protection of this application. Therefore, the scope of protection of this application should be determined by the scope of the claims.
Claims
1. A method for dynamic assessment of drought vulnerability of grassland ecosystems based on multi-source remote sensing data and multi-dimensional indicators, characterized in that, The application relates to a method for evaluating the drought vulnerability of a grassland ecosystem. The method comprises the following steps: Fusing multi-source remote sensing data to construct a total primary productivity dataset and a leaf area index dataset of the grassland; Identifying a drought event and extracting drought characteristics based on the total primary productivity dataset and the leaf area index dataset, and quantifying the cumulative effect and lag effect of the grassland ecosystem on drought; Constructing a multi-dimensional vulnerability index system comprising a resistance index, a resilience index and a sensitivity index based on the cumulative effect and lag effect of drought, dynamically evaluating the drought vulnerability of the grassland ecosystem, and obtaining an evaluation result; 2. The method for dynamic assessment of drought vulnerability of grassland ecosystem based on multi-source remote sensing data and multi-dimensional indicators according to claim 1, characterized in that, Mapping the evaluation result into an RGB three-channel, generating a spatial distribution map of the drought vulnerability of the grassland by combining the three-channel values, and realizing the classification of the vulnerability and the identification of the priority repair area. The step of fusing multi-source remote sensing data comprises the following steps: Obtaining leaf area index data, total primary productivity data and meteorological grid data; 3. The method for dynamic assessment of drought vulnerability of grassland ecosystem based on multi-source remote sensing data and multi-dimensional indicators according to claim 1, characterized in that, Fusing the leaf area index data and the total primary productivity data, calculating a standardized precipitation evapotranspiration index, and unifying all data into the same spatial grid and time sequence. The step of identifying a drought event based on a drought index comprises the following steps: Based on the time sequence of the pixel-by-pixel standardized precipitation evapotranspiration index, a run theory is applied to identify independent drought events; 4. The method for dynamic assessment of drought vulnerability of grassland ecosystem based on multi-source remote sensing data and multi-dimensional indicators according to claim 1, characterized in that, For each identified drought event, the duration, severity, intensity and peak value of the drought are calculated. The step of extracting drought characteristics comprises the following steps: A maximum correlation coefficient method is adopted to analyze the cumulative effect and lag effect of drought on vegetation; The cumulative effect is characterized by a maximum cumulative correlation coefficient and a corresponding optimal cumulative month; 5. The method for dynamic assessment of drought vulnerability of grassland ecosystem based on multi-source remote sensing data and multi-dimensional indicators according to claim 1, characterized in that, The lag effect is characterized by a maximum lag correlation coefficient and a corresponding optimal lag month. The step of quantifying the cumulative effect comprises the following steps:
6. The method for dynamic assessment of drought vulnerability of grassland ecosystem based on multi-source remote sensing data and multi-dimensional indicators according to claim 1, characterized in that, The Pearson correlation coefficient of the total primary productivity or the leaf area index of the grassland in the growing season and the SPEI in the scale of 1-12 months is calculated, and the maximum cumulative correlation coefficient and the maximum cumulative time scale are extracted. The step of constructing a sensitivity index comprises the following steps:
7. The method for dynamic assessment of drought vulnerability of grassland ecosystem based on multi-source remote sensing data and multi-dimensional indicators according to claim 1, characterized in that, The maximum cumulative correlation coefficient, the maximum lag correlation coefficient, the cumulative time and the lag time are standardized, and a comprehensive sensitivity index is calculated by using an arithmetic mean value.
8. The method for dynamic assessment of drought vulnerability of grassland ecosystem based on multi-source remote sensing data and multi-dimensional indicators according to claim 1, characterized in that, The step of constructing a resistance index comprises the following steps: The relative deviation of the standardized anomaly value of the vegetation in the most severe month during the drought and the baseline value before the drought is calculated.
9. The method for dynamic assessment of drought vulnerability of grassland ecosystem based on multi-source remote sensing data and multi-dimensional indicators according to claim 1, characterized in that, The step of constructing a resilience index comprises the following steps: The recovery rate of the standardized anomaly value of the vegetation in the optimal recovery month after the drought and the anomaly value at the beginning of the drought is calculated. The step of generating a spatial distribution map of the drought vulnerability of the grassland comprises the following steps: The resistance index, the sensitivity index and the resilience index are respectively mapped to the red, green and blue channels, the values are standardized to a preset range, a composite color spectrum is generated, and a Jenks natural breakpoint method is adopted to divide a five-level vulnerability gradient.