A method and system for predicting vegetation vulnerability constraint under dry heat compound stress

Abstract

The application discloses a kind of vegetation vulnerability constraint estimation method and system under dry-hot compound stress, comprising: 1. obtaining observed vegetation total primary productivity data, meteorological data and the same kind of meteorological data of each grid point under global climate model;2. determine the standardized precipitation evapotranspiration index time scale of each grid point and determine the dominant stress type of different grid points growing season;3. respectively build the two-dimensional Copula model corresponding to the total primary productivity of corresponding dominant stress, determine the stress threshold corresponding to the vegetation loss induced;4. based on climate model, build emergent constraint model for future estimation of dominant stress;5. based on the mean value and uncertainty of the future estimation of dry-hot stress index corrected by emergent constraint relationship;6. the mean value and uncertainty range of the corrected future dry-hot stress index are input into Copula model, and the corresponding loss probability is calculated.The application considers the response difference of different vegetation ecosystems to dry-hot stress while improving the prediction accuracy of drought heat wave and its compound event.

Description

Technical Field

[0001] This invention belongs to the field of drought and heat waves and their ecological impacts, and proposes a method and system for predicting vegetation vulnerability under combined dry and hot stress. Background Technology

[0002] Against the backdrop of global warming, frequent droughts and heat waves have severely impacted terrestrial vegetation ecosystems. When these two extreme events occur simultaneously, they create a positive feedback loop through soil-water-atmosphere coupling, exacerbating the intensity of both droughts and heat waves. This leads to reduced crop yields, forest fires, mass mortality, further weakening the carbon absorption capacity of ecosystems, reducing vegetation productivity, and causing irreversible damage to the structure and function of terrestrial ecosystems. Therefore, assessing vegetation responses to drought and heat stress is crucial for ecological environmental protection and healthy sustainable development.

[0003] Some scholars have assessed the impacts of drought, heat waves, and combined dry-heat events on total primary productivity of vegetation. However, these studies often fail to consider the significant differences in the responses of different vegetation ecosystems to various dry-heat stresses due to variations in geographical location, hydrological and climatic conditions, and primarily focus on historical periods. Currently, research on future climate change and its impacts mainly relies on global climate models. Although climate model-based studies generally indicate an increasing trend in future extreme dry-heat events, and a few studies suggest further increases in vegetation loss during future droughts and heat waves, these studies all overlook the significant uncertainties in climate model projections.

[0004] Given the uncertainty of the future changes and impacts of drought, heat waves and combined dry and heat stresses, as well as the differences in the responses of various vegetation ecosystems to climate change, there are still many challenges in how to further accurately quantify the potential dominant dry and heat stress risks faced by different vegetation ecosystems in the future, and take targeted adaptation measures, vegetation restoration strategies and disaster prevention and mitigation policies based on the forecast results. Summary of the Invention

[0005] The purpose of this invention is to take into account the uncertainty of existing global climate models in predicting drought, heat waves and their combined events, quantify the potential dry and hot stress risks faced by different vegetation ecosystems in the future, and provide a method for predicting vegetation vulnerability constraints under dry and hot combined stress.

[0006] To solve the above-mentioned technical problems, the present invention provides the following technical solution: A method for predicting vegetation vulnerability constraints under combined dry and hot stress includes the following steps: Step 1. Obtain the observed total primary productivity of vegetation and meteorological data, and obtain similar meteorological data for historical and future growing seasons for each grid point under the global climate model; Step 2. Determine the timescale of the standardized precipitation evapotranspiration index for each grid point, and identify the dominant stress type during the growing season for different grid points; Step 3. Construct a two-dimensional Copula model of dominant stress and total primary productivity of vegetation to determine the stress threshold corresponding to induced vegetation loss; Step 4. Calculate future estimates of dominant stresses based on climate models and exclude outlier climate models; construct emergence constraint models for future projections of historical growing season total primary productivity, daily average temperature trends, and hot and dry stress indices for different dominant stress types. Step 5. Based on the emergent constraint model, use the total primary productivity of vegetation and the daily average temperature trend observed in the historical growing season to correct the estimated mean and uncertainty range of future dry and hot stress indicators. Step 6. Input the corrected future dry and hot stress indexes, along with their estimated mean and uncertainty range, into the two-dimensional Copula model to estimate the future vegetation loss in each region and calculate the probability of vegetation loss.

[0007] Furthermore, in step 1, the meteorological data includes at least the daily maximum and minimum temperatures, precipitation, and other meteorological data used to calculate the standardized precipitation evapotranspiration index, and the observed meteorological data is collected in at least three categories.

[0008] Furthermore, in step 2, different time scales for each grid point are calculated based on the observed total primary productivity of vegetation. t Corresponding Standardized Precipitation Evapotranspiration Index Pearson correlation coefficient with monthly total primary productivity of vegetation, based on the maximum correlation coefficient r Determine the time scale t .

[0009] Furthermore, in step 2, the dominant stress types include drought stress, heat wave stress, and combined dry-heat stress, wherein the combined dry-heat stress is when a drought event occurs simultaneously with a heat wave event. Among them, the indicators of drought stress are time scales. t Corresponding Standardized Precipitation Evapotranspiration Index The indicator of heat wave stress is the daily maximum temperature. The index of combined dry heat stress is the comprehensive intensity of combined dry heat events. :

[0010] In the formula, n This refers to the month corresponding to combined dry and hot stress. It is the highest daily temperature during a complex dry and hot event. The heat wave threshold is determined based on the percentile of the baseline period. NIt is the total number of days corresponding to the combined hot and dry period during the growing season of the year. for n Monthly timescale t The corresponding standardized precipitation evapotranspiration index; The regression coefficients between total primary productivity (GPP) and each indicator during the growing season at each grid point were calculated using standard multiple linear regression.

[0011] In the formula, ， , These are regression coefficients, representing the degree of impact of dry heat stress on GPP. The stress index with the largest absolute value of the regression coefficient is identified as the dominant stress. It is the intercept. This is the error term.

[0012] Furthermore, in step 3, a two-dimensional Copula model of dominant stress and total primary productivity (GPP) is constructed at the grid scale based on historical observations, yielding the conditional probabilities of GPP falling below different percentiles as follows:

[0013] In the formula, Indicates the different percentiles of GPP; Indicator The dominant stress indicator for each grid point; u1 and u2 These are the upper and lower bounds of the dominant coercion, respectively. The joint distribution function of dominant stress and GPP, The distribution function of the dominant stress; This indicates that when the dominant coercion is less than or equal to u1 And GPP is less than or equal to The joint probability at time, This indicates that when the dominant coercion is less than or equal to u2 And GPP is less than or equal to The joint probability at time, , These respectively indicate that the dominant stress is less than or equal to The probability value at that time.

[0014] Furthermore, in step 3, the two-dimensional Copula model is combined with the experimental algorithm to... Indicator Iterate at a certain step size to estimate the conditional probability corresponding to the iteration. When, take the corresponding Indicator Interval average (u1+u2) / 2 To trigger different vegetation losses The stress thresholds include: When drought is the dominant stress, 0.1 is used as the benchmark. The iteration step size; When heat waves are the dominant stress, 0.5 is used as... The iteration step size; When combined dry heat is the dominant stress, 0.5 is used as... The iteration step size.

[0015] Furthermore, in step 4, the emergence constraint model includes a univariate emergence constraint model and a bivariate emergence constraint model; For grid points dominated by drought or heat wave stress, the univariate emergent constraint model is constructed as follows:

[0016] In the formula, It is the regression coefficient. It is the intercept; and In drought-stress-dominated areas, respectively the future and history Trends; future trends in areas dominated by heat wave stress and history trend; At the grid points dominated by combined dry heat stress, the constructed bivariate emergent constraint model is as follows:

[0017] In the formula, It is the intercept. , ， It is the regression coefficient. It is an error term; , These are the historically long seasons and daily average temperature trend, It is the combined intensity of future complex dry and hot events.

[0018] Furthermore, in step 5, based on the univariate emergent constraint model for each grid point, the future predicted mean of the corrected dry heat stress index is obtained. and variance as follows:

[0019]

[0020] In the formula, x It is the historical trend of the dominant stress indicator in the model output. It is its corresponding mean. It is its corresponding variance; z It is the future forecast of the dominant stress indicator output by the model. It is its corresponding mean. It is its corresponding variance; y 0 represents the observed mean of the corresponding indicator. It is the variance between the observed data of the corresponding indicator; ρ yes x and z The Pearson correlation coefficient; Based on the bivariate emergent constraint model at each grid point, the corrected composite dry heat stress index is obtained. Future probability distribution as follows:

[0021]

[0022] In the formula, m Represents the original multi-mode output , x 1, x 2 represents and average daily temperature Simulated historical trends It is given x 1. x 2 of m probability density distribution; ( x 1 ,x 2) is a given x 1. x 2 multi-mode output Prediction variance f ( x 1 ,x 2) is x 1 and x The joint probability density function of 2; Represent and average daily temperature Historical observation trends yes The binary joint Gaussian distribution; Represents the future after constraint correction , Represents a given observation x 1 and x After constraint correction 2 Probability distribution.

[0023] Furthermore, in step 6, the future projected mean of the corrected dry heat stress index is compared. Different types of vegetation loss The corresponding dry heat stress threshold; the estimated future vegetation loss due to dry heat stress at each grid point; and the further adjustment of the uncertainty range of the corrected future dry heat stress index. < Indicator < Input the Copula model from step 2 to obtain the probability that the GPP is below a certain historical percentile. This refers to the probability of different degrees of vegetation loss.

[0024] On the other hand, the present invention provides a vegetation vulnerability constraint prediction system under combined dry and hot stress, comprising: The data acquisition module is used to acquire observed total primary productivity of vegetation and meteorological data, as well as historical and future growing season similar meteorological data for each grid point under the global climate model. The dominant stress type determination module is used to determine the precipitation evapotranspiration index timescale for each grid point and to determine the dominant stress type for the growing season at different grid points. The stress threshold determination module is used to construct a two-dimensional Copula model of dominant stress and total primary productivity of vegetation to determine the stress threshold corresponding to induced vegetation loss. Emergent constraint model building module. It is used to calculate the future estimate of dominant stress based on climate models and exclude outlier climate models; it constructs emergent constraint models for the future prediction of historical growing season total primary productivity of vegetation, daily average temperature trend and dry and hot stress index corresponding to different dominant stress types. The correction module is used to correct the estimated mean and uncertainty range of future dry and hot stress indicators based on the emergence constraint model, using historical growing season observations of total primary productivity of vegetation and daily average temperature trends. The vegetation loss calculation module is used to input the estimated mean and uncertainty range of the corrected future dry and hot stress index into the two-dimensional Copula model to estimate the future vegetation loss faced by each region and calculate the probability of vegetation loss.

[0025] Compared with the prior art, the present invention has the following beneficial effects: (1) Considering the differences in vegetation ecosystems: Considering the differences in geographical location, hydro-climatic conditions, and biochemical characteristics among different vegetation ecosystems, their responses to drought, heat waves, and their combined events vary. Therefore, this method refines the identification of locally dominant dry and heat stresses at the grid scale and conducts big data analysis. This enables the method to provide more accurate predictions and assessments for vegetation ecosystems in different regions, helping relevant departments to take targeted adaptive measures, whereas existing technologies usually do not fully consider these differences.

[0026] (2) Reduce the uncertainty of future forecasts: This invention takes into account the error of climate model simulation forecasts. By introducing the emergence constraint method and solving the constraint, combined with multi-source observation data, it can effectively correct the large uncertainty in climate model forecasts, provide more accurate forecasts of future climate change and its impacts, enhance the reliability and usability of forecast results, and provide more robust climate adaptability analysis support for decision-makers.

[0027] (3) Accurately quantify stress risks and vegetation vulnerability: This invention integrates emergent constraints, multiple regression and Copula model methods to accurately quantify the potential dominant dry and hot stress risks faced by different vegetation ecosystems in the future, provide more accurate vegetation loss probability prediction to assess future vegetation vulnerability, and support the formulation of vegetation restoration strategies and the optimization of disaster prevention and mitigation policies. Attached Figure Description

[0028] To more clearly illustrate the technical solutions in this invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are some embodiments of this invention. For those skilled in the art, other drawings can be obtained from these drawings without creative effort.

[0029] Figure 1 This is a flowchart of a method for predicting vegetation vulnerability under combined dry and hot stress according to Embodiment 1 of the present invention.

[0030] Figure 2 This is a schematic diagram illustrating the identification of drought, heat waves, and combined dry heat events in Embodiment 1 of the present invention. Detailed Implementation

[0031] To make the objectives, technical solutions, and advantages of this invention clearer, the technical solutions of this invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some, not all, of the embodiments of this invention. All other embodiments obtained by those skilled in the art based on the embodiments of this invention without creative effort are within the scope of protection of this invention.

[0032] Example 1 The invention will now be further described with reference to the accompanying drawings.

[0033] like Figure 1 As shown in the figure, this embodiment discloses a method for predicting vegetation vulnerability constraints under combined dry and hot stress, including the following steps: Step 1. Obtain observed total primary productivity of vegetation, temperature, precipitation, and other parameters used to calculate the standardized precipitation evapotranspiration index. SPEIMeteorological data, and at the same time, historical and future similar meteorological data for each grid point under the global climate model.

[0034] In one specific implementation, the Gross Primary Productivity (GPP) data for vegetation came from MODIS; the observational data for precipitation, daily maximum temperature, and average temperature came from five datasets: CRU TS4.08, ERA5 reanalysis, Berkeley Surface Temperature Project, and CPC-Unified, to prepare for calculating observational uncertainties in step 5; 24 CMIP6 global climate models were used, whose outputs included daily maximum and minimum temperatures and precipitation data; future scenarios included SSP1-2.6, SSP2-4.5, and SSP5-8.5, and all data were unified to a spatial resolution of 0.25° using bilinear interpolation.

[0035] Step 2. Determine each grid point SPEI Time scale t We used standard multiple linear regression to determine the dominant dry and hot stresses during the growing season in different vegetation zones.

[0036] In one specific implementation, the Thornthwaite method is first used to calculate potential evapotranspiration, and then the potential evapotranspiration at different time scales is calculated. The optimal value for each grid point is determined based on the maximum Pearson correlation coefficient. SPEI Time scale t Recorded as drought stress index ， when A drought event is considered to occur when the temperature is below -1°C. The day's highest temperature... Exceeding the 90th percentile of the baseline period (1980-2010) At that time, it was considered that a heat wave event had occurred. For example... Figure 2 As shown, a combined dry-heat event is considered to have occurred when a heat wave occurs within a dry month. The stress index of a combined dry-heat event is its overall intensity. : (1) In the formula, n It refers to the month corresponding to the complex dry and hot event. It is the highest daily temperature during a complex dry and hot event. N It is the total number of days corresponding to the combined hot and dry period during the growing season of the year.

[0037] The regression coefficients between total primary productivity (GPP) and dry-heat stress indices during historical growing seasons were calculated using standard multiple linear regression at a grid scale: (2) In the formula, ， , These are regression coefficients, representing the degree of impact of dry heat stress on GPP. The stress index with the largest absolute value of the regression coefficient is identified as the dominant stress. It is the intercept. This is the error term.

[0038] Step 3. Construct two-dimensional Copula models of the dominant stress and total primary productivity of vegetation for each grid point, and determine the stress threshold corresponding to induced vegetation loss.

[0039] In one specific implementation, in order to select a suitable bivariate Copula model for each grid point, the marginal distributions of each variable are first transformed.

[0040] Optionally, the gamma distribution function can be used to... Fit the data to transform it into a uniform marginal distribution; use a Gaussian distribution to... The data is fitted and transformed into a uniform marginal distribution. For GPP data, different distribution models (such as log-normal distribution, gamma distribution, etc.) are tried to fit the data based on its distribution characteristics, and the optimal distribution model is selected. After obtaining the uniform marginal distributions of all variables, the dominant stress is fitted at each grid point. or A bivariate Copula model between GPP and GPP is used, considering different types of Copula families for fitting. The optimal Copula function is selected by calculating the Bayesian Information Criterion (BIC), and the conditional probabilities (probabilities of different degrees of vegetation loss) of GPP below different percentiles are obtained as follows: (3) In the formula, Indicates the different percentiles of GPP; Indicator The dominant stress indicator for each grid point; u1 and u2 These are the upper and lower bounds of the dominant coercion, respectively. The joint distribution function of dominant stress and GPP, The distribution function of the dominant stress; This indicates that when the dominant coercion is less than or equal to u1 And GPP is less than or equal to The joint probability at time, This indicates that when the dominant coercion is less than or equal to u2 And GPP is less than or equal to The joint probability at time, , These respectively indicate that the dominant stress is less than or equal to The probability value at that time.

[0041] Furthermore, the constructed Copula model is combined with experimental algorithms to... Indicator Iterate at a certain step size to estimate the conditional probability corresponding to the iteration. When, take the corresponding Indicator Interval average (u1+u2) / 2 To trigger different vegetation losses ( The stress thresholds include: When drought is the dominant stress, 0.1 is used as the benchmark. The iteration step size; When heat waves are the dominant stress, 0.5 is used as... The iteration step size; When combined dry heat is the dominant stress, 0.5 is used as... The iteration step size.

[0042] Step 4. Calculate future estimates of dominant stresses for each grid point based on climate models, and use the Z-score method to exclude outlier models; construct historical growing season data for each vegetation zone. 、 Emergent constraint model for daily average temperature trends and future predictions of dominant dry and hot stress indicators.

[0043] In one specific implementation, future estimates of dominant stresses are calculated based on various climate models, and outlier models are excluded using the Z-score method: (4) In the formula, X These are the original estimates from various climate models. It is the average value. S It is the standard deviation, subscript g Representing grid points g If the simulated value of a particular model exceeds three times the standard deviation of the multi-model average (i.e., Z>3), it is classified as an outlier and excluded. The remaining models are used to construct an emergent constraint model.

[0044] Based on the selected set of climate models, the average daily temperature during the historical growing season for each grid point was constructed. ), Emergent constraint models for predicting future trends and dominant dry-heat stress indicators include: When drought or heat wave stress dominates, the univariate emergent constraint model is constructed as follows: (5) In the formula, a It is the regression coefficient. b It is the intercept; Indicator future and Trend history Under drought stress as the dominant factor, they represent the future. and history Trends; future trends when heat wave stress dominates and history trend; Under combined dry heat stress, the constructed bivariate emergent constraint model is as follows: (6) In the formula, a 0 is the intercept. a 1, a 2 ，a 3 is the regression coefficient, and ε is the error term; , These are the historically long seasons and daily average temperature trend, It is the combined intensity of future complex dry and hot events.

[0045] Step 5. Based on the above emergent constraint model, utilize historical growing season observations... The future estimated mean and uncertainty of the daily average temperature trend-corrected dry heat stress index.

[0046] In one specific implementation, the future predicted mean of the corrected dry heat stress index is obtained based on the univariate emergent constraint model at each grid point. and variance as follows: (7) (8) In the formula, x This is the historical trend of the dominant stress indicator in multi-mode output. It is its corresponding mean. It is its corresponding variance; z It is the future forecast of the dominant stress indicator output by the model. It is its corresponding mean. It is its corresponding variance; y 0 represents the observed mean of the corresponding indicator. It is the variance between the observed data of the corresponding indicator; ρ yes x and z The Pearson correlation coefficient; The corrected composite dry heat stress index is obtained based on the bivariate emergent constraint model at each grid point. Future probability distribution as follows: (9) (10) In the formula, mRepresents the original multi-mode output , x 1, x 2 represents and average daily temperature Simulated historical trends It is given x 1. x 2 of m probability density distribution; ( x 1 ,x 2) is a given x 1. x 2 multi-mode output Prediction variance f ( x 1 ,x 2) is x 1 and x The joint probability density function of 2; Represent and average daily temperature Historical observation trends yes The binary joint Gaussian distribution; Represents the future after constraint correction , Represents a given observation x 1 and x After constraint correction 2 Probability distribution.

[0047] Step 6. Input the mean and uncertainty range of the corrected future dry and hot stress index into the Copula model to predict the future vegetation loss in each region and calculate the probability of vegetation loss.

[0048] In one specific implementation, the mean value of the future dry heat stress index after grid-scale comparison correction. With different vegetation losses ( The corresponding dry heat stress threshold, specifically, Below the threshold, and Exceeding a threshold leads to vegetation loss; the spatial distribution of future vegetation loss, predicted by the constrained multi-mode mean, can be obtained. This further expands the uncertainty range of the corrected future dry-heat stress index. < Indicator < Input the Copula model from step 2 and calculate the probability P that the GPP is below a certain historical percentile. ); This allows us to obtain the spatial distribution of the probability of vegetation loss of different degrees when considering the uncertainty of multi-mode constraint prediction.

[0049] In summary, the vegetation vulnerability estimation method under combined dry and hot stress proposed in this invention combines emergence constraints, multiple regression, and Copula models. It considers the differences in responses to drought, heat waves, and their combined events among different vegetation ecosystems due to geographical location, hydro-climate, and biochemical characteristics. First, the dominant local dry and hot stresses are identified at a grid scale. Then, the future predictions and uncertainties of climate models are constrained and corrected, providing a more accurate quantification of the potential dominant dry and hot stresses faced by different vegetation ecosystems in the future. Based on the estimation results, targeted adaptive measures, vegetation restoration strategies, and disaster prevention and mitigation policies can be adopted. Simultaneously, it can provide important reference for the calculation and assessment of carbon cycle under climate change. Example 2 This embodiment provides a vegetation vulnerability constraint prediction system under combined dry and hot stress, including: The data acquisition module is used to acquire observed total primary productivity of vegetation and meteorological data, as well as historical and future growing season similar meteorological data for each grid point under the global climate model. The dominant stress type determination module is used to determine the precipitation evapotranspiration index timescale for each grid point and to determine the dominant stress type for the growing season at different grid points. The stress threshold determination module is used to construct a two-dimensional Copula model of dominant stress and total primary productivity of vegetation to determine the stress threshold corresponding to induced vegetation loss. Emergent constraint model building module. It is used to calculate the future estimate of dominant stress based on climate models and exclude outlier climate models; it constructs emergent constraint models for the future prediction of historical growing season total primary productivity of vegetation, daily average temperature trend and dry and hot stress index corresponding to different dominant stress types. The correction module is used to correct the estimated mean and uncertainty range of future dry and hot stress indicators based on the emergence constraint model, using historical growing season observations of total primary productivity of vegetation and daily average temperature trends. The vegetation loss calculation module is used to input the estimated mean and uncertainty range of the corrected future dry and hot stress index into the two-dimensional Copula model to estimate the future vegetation loss faced by each region and calculate the probability of vegetation loss.

[0050] It should be understood that any parts not described in detail in this specification belong to the prior art.

[0051] It should be understood that the above description of the preferred embodiments is quite detailed, but this should not be construed as limiting the scope of protection of this invention. It is neither necessary nor possible to exhaustively describe all possible implementations. Those skilled in the art, guided by this invention, can make substitutions or modifications without departing from the scope of the claims, all of which fall within the scope of protection of this invention. The scope of protection of this invention should be determined by the appended claims.

Claims

1. A method for predicting vegetation vulnerability constraints under combined dry and hot stress, characterized in that, Includes the following steps: Step 1. Obtain the observed total primary productivity of vegetation and meteorological data, and obtain similar meteorological data for historical and future growing seasons for each grid point under the global climate model; Step 2. Determine the timescale of the standardized precipitation evapotranspiration index for each grid point, and determine the dominant stress type for the growing season at different grid points; the dominant stress types include drought stress, heat wave stress, and combined dry-heat stress, wherein the combined dry-heat stress is when a drought event occurs simultaneously with a heat wave event; Among them, the indicators of drought stress are time scales. t Corresponding Standardized Precipitation Evapotranspiration Index The indicator of heat wave stress is the daily maximum temperature. The index of combined dry heat stress is the comprehensive intensity of combined dry heat events. : In the formula, n This refers to the month corresponding to combined dry and hot stress. It is the highest daily temperature during a complex dry and hot event. The heat wave threshold is determined based on the percentile of the baseline period. N It is the total number of days corresponding to complex hot and dry events during the growing season of the year. for n Monthly timescale t The corresponding standardized precipitation evapotranspiration index; The regression coefficients between total primary productivity (GPP) and each indicator during the growing season at each grid point were calculated using standard multiple linear regression. In the formula, ， , These are regression coefficients, representing the degree of impact of dry heat stress on GPP. The stress index with the largest absolute value of the regression coefficient is identified as the dominant stress. It is the intercept. It is an error term; Step 3. Construct a two-dimensional Copula model of dominant stress and total primary productivity (GPP) to determine the stress threshold corresponding to induced vegetation loss; based on historical observations, construct a two-dimensional Copula model of dominant stress and GPP at the grid scale to obtain the conditional probabilities of GPP falling below different percentiles as follows: In the formula, Indicates the different percentiles of GPP; Indicator The dominant stress indicator for each grid point; u1 and u2 These are the upper and lower bounds of the dominant coercion, respectively. The joint distribution function of dominant stress and GPP, The distribution function of the dominant stress; This indicates that when the dominant coercion is less than or equal to u1 And GPP is less than or equal to The joint probability at time, This indicates that when the dominant coercion is less than or equal to u2 And GPP is less than or equal to The joint probability at time, , These respectively indicate that the dominant stress is less than or equal to The probability value at that time; Step 4. Calculate future estimates of dominant stresses based on climate models and exclude outlier climate models; construct emergence constraint models for future projections of historical growing season total primary productivity, daily average temperature trends, and hot and dry stress indices for different dominant stress types. Step 5. Based on the emergent constraint model, use the total primary productivity of vegetation and the daily average temperature trend observed in the historical growing season to correct the estimated mean and uncertainty range of future dry and hot stress indicators. Step 6. Input the corrected future dry and hot stress indexes, along with their estimated mean and uncertainty range, into the two-dimensional Copula model to estimate the future vegetation loss in each region and calculate the probability of vegetation loss.

2. The method for predicting vegetation vulnerability constraints under combined dry and hot stress according to claim 1, characterized in that: In step 1, the meteorological data shall include at least the daily maximum and minimum temperatures, precipitation, and other meteorological data used to calculate the standardized precipitation evapotranspiration index, and at least three types of meteorological data shall be collected.

3. The method for predicting vegetation vulnerability constraints under combined dry and hot stress according to claim 1, characterized in that: In step 2, different time scales for each grid point are calculated based on the observed total primary productivity of vegetation. t Corresponding Standardized Precipitation Evapotranspiration Index Pearson correlation coefficient with monthly total primary productivity of vegetation, based on the maximum correlation coefficient r Determine the time scale t .

4. The method for predicting vegetation vulnerability constraints under combined dry and hot stress according to claim 1, characterized in that: In step 3, the two-dimensional Copula model is combined with the experimental algorithm to... Indicator Iterate at a certain step size to estimate the conditional probability corresponding to the iteration. When, take the corresponding Indicator Interval average (u1+u2) / 2 To trigger different vegetation losses The stress thresholds include: When drought is the dominant stress, 0.1 is used as the baseline. The iteration step size; When heat waves are the dominant stress, 0.5 is used as... The iteration step size; When combined dry heat is the dominant stress, 0.5 is used as... The iteration step size.

5. The method for predicting vegetation vulnerability constraints under combined dry and hot stress according to claim 1, characterized in that: In step 4, the emergence constraint model includes a univariate emergence constraint model and a bivariate emergence constraint model; For grid points dominated by drought or heat wave stress, the univariate emergent constraint model is constructed as follows: In the formula, It is the regression coefficient. It is the intercept; and In drought-stress-dominated areas, respectively the future and history Trends; future trends in areas dominated by heat wave stress and history trend; At the grid points dominated by combined dry heat stress, the constructed bivariate emergent constraint model is as follows: In the formula, It is the intercept. , ， It is the regression coefficient. It is an error term; , These are the historical growing seasons and daily average temperature trend, It is the combined intensity of future complex dry and hot events.

6. The method for predicting vegetation vulnerability constraints under combined dry and hot stress according to claim 5, characterized in that: In step 5, based on the univariate emergent constraint model for each grid point, the future predicted mean of the corrected dry heat stress index is obtained. and variance as follows: In the formula, x It is the historical trend of the dominant stress indicator in the model output. It is its corresponding mean. It is its corresponding variance; z It is the future forecast of the dominant stress indicator output by the model. It is its corresponding mean. It is its corresponding variance; y 0 represents the observed mean of the corresponding indicator. It is the variance between the observed data of the corresponding indicator; ρ yes x and z The Pearson correlation coefficient; Based on the bivariate emergent constraint model at each grid point, the corrected composite dry heat stress index is obtained. Future probability distribution as follows: In the formula, m Represents the original multi-mode output , x 1, x 2 represents and average daily temperature Simulated historical trends It is given x 1. x 2 of m probability density distribution; ( x 1 ,x 2) is a given x 1. x 2 multi-mode output Prediction variance f ( x 1 ,x 2) is x 1 and x The joint probability density function of 2; Represent and average daily temperature Historical observation trends yes The binary joint Gaussian distribution; Represents the future after constraint correction , Represents a given observation x 1 and x After constraint correction 2 Probability distribution.

7. The method for predicting vegetation vulnerability constraints under combined dry and hot stress according to claim 6, characterized in that: In step 6, the future estimated mean of the corrected dry heat stress index is compared. Different types of vegetation loss The corresponding dry heat stress threshold; Estimate future vegetation loss due to heat and dryness stress at each grid point; further narrow down the uncertainty range of the corrected future heat and dryness stress index. < Indicator < Input the Copula model from step 2 to obtain the probability that the GPP is below a certain historical percentile. This refers to the probability of different degrees of vegetation loss.

8. A vegetation vulnerability constraint prediction system under combined dry and hot stress, characterized in that, include: The data acquisition module is used to acquire observed total primary productivity of vegetation and meteorological data, as well as historical and future growing season similar meteorological data for each grid point under the global climate model. The dominant stress type determination module is used to determine the time scale of the standardized precipitation evapotranspiration index for each grid point and to determine the dominant stress type for the growing season at different grid points. The stress threshold determination module is used to construct a two-dimensional Copula model of dominant stress and total primary productivity of vegetation to determine the stress threshold corresponding to induced vegetation loss. Emergent constraint model building module. It is used to calculate the future estimate of dominant stress based on climate models and exclude outlier climate models; it constructs emergent constraint models for the future prediction of historical growing season total primary productivity of vegetation, daily average temperature trend and dry and hot stress index corresponding to different dominant stress types. The correction module is used to correct the estimated mean and uncertainty range of future dry and hot stress indicators based on the emergence constraint model, using historical growing season observations of total primary productivity of vegetation and daily average temperature trends. The vegetation loss calculation module is used to input the estimated mean and uncertainty range of the corrected future dry and hot stress index into the two-dimensional Copula model to estimate the future vegetation loss faced by each region and calculate the probability of vegetation loss. The vegetation vulnerability constraint prediction system under dry and hot combined stress is used to perform the steps in the vegetation vulnerability constraint prediction method under dry and hot combined stress as described in any one of claims 1-7.

Images